Datasets:

TaylorAI

/

pubmed_commercial

like 12

PMC005xxxxxx/PMC5002203.txt

==== Front J Transl MedJ Transl MedJournal of Translational Medicine1479-5876BioMed Central London 100910.1186/s12967-016-1009-3ResearchProton-pump inhibitor omeprazole attenuates hyperoxia induced lung injury Richter Jute +32 16 342288jute.richter@uzleuven.be 127Jimenez Julio juliojimenez@gmail.com 13Nagatomo Taro ctnagatomo@gmail.com 14Toelen Jaan jaan.toelen@uzleuven.be 12Brady Paul paul.brady@med.kuleuven.be 5Salaets Thomas salaetst@gmail.com 12Lesage Flore flore.lesage@med.kuleuven.be 1Vanoirbeek Jeroen jeroen.vanoirbeek@med.kuleuven.be 6Deprest Jan jan.deprest@uzleuven.be 121 Department of Development and Regeneration, Faculty of Medicine, KU Leuven, Leuven, Belgium 2 Division Woman and Child, University Hospitals Leuven, Leuven, Belgium 3 Departamento Ginecología y Obstetricia, Clínica Alemana, Santiago, Chile 4 Department of Neonatology, Ehime Prefectural Central Hospital, Matsuyama, Japan 5 Centre for Human Genetics, University Hospitals Leuven, KU Leuven, Leuven, Belgium 6 Laboratory of Occupational and Environmental Toxicology, Department of Public Health and Primary Care, KU Leuven, Leuven, Belgium 7 Clinical Department of Obstetrics and Gynaecology and Academic Department of Development and Regeneration, Organ System Cluster, University Hospitals of Leuven, Herestraat 49, 3000 Leuven, Belgium 27 8 2016 27 8 2016 2016 14 1 24721 5 2016 16 8 2016 © The Author(s) 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background The administration of supplemental oxygen to treat ventilatory insufficiency may lead to the formation of reactive oxygen species and subsequent tissue damage. Cytochrome P4501A1 (CYP1A1) can modulate hyperoxic lung injury by a currently unknown mechanism. Our objective was to evaluate the effect of administration of omeprazole on the induction of CYP1A1 and its influence on hyperoxic lung injury in an established preterm rabbit model. Methods Omeprazole was administered either (1) directly to the fetus, (2) to the mother or (3) after birth to the pups in different doses (2–10 or 20 mg/kg). Controls were injected with the same amount of saline. Pups were housed in normoxia (21 %) or hyperoxia (>95 %) for 5 days. Outcome parameters were induction of CYP1A1 measured by real-time polymerase chain reaction (RT-PCR) immediately after delivery, at day 3 and day 5 as well as lung function, morphometry and immunohistochemistry assessed at day 5 of life. Transcriptome analysis was used to define the targeted pathways. Results Daily neonatal injections demonstrated a dose-dependent increase in CYP1A1. Lung function tests showed a significant improvement in tissue damping, tissue elasticity, total lung capacity, static compliance and elastance. Morphometry revealed a more developed lung architecture with thinned septae in animals treated with the highest dose (20 mg/kg) of omeprazole. Surfactant protein B, vascular endothelial growth factor and its receptor were significantly increased on immunohistochemical stainings after omeprazole treatment. Conclusions Neonatal administration of omeprazole induces CYP1A1 in a dose-dependent matter and combined pre- and postnatal administration attenuates hyperoxic lung injury in preterm rabbits, even with the lowest dose of omeprazole without clear CYP1A1 induction. Electronic supplementary material The online version of this article (doi:10.1186/s12967-016-1009-3) contains supplementary material, which is available to authorized users. Keywords Bronchopulmonary dysplasiaOmeprazoleCYP1A1RabbitPretermMarie Curie Industria-Academia Partnership Program grantPIAP-GA-2009-251356Richter Jute Erasmus Mundi Doctoral grant2013-0040Jimenez Julio Fonds Wetenschappelijk Onderzoek Vlaanderen1801207Deprest Jan Klinische Opleidings- en Onderzoeks- Raadissue-copyright-statement© The Author(s) 2016 ==== Body Background Although the incidence of preterm birth is decreasing slightly over the last years, it still affects 12 % of pregnancies in the United States [1]. In other countries, like Belgium, the preterm delivery rate may be lower (7.6 % deliveries <37 weeks), though remains stable. Severe prematurity can affect many organ systems but respiratory insufficiency remains the major contributor to perinatal morbidity and mortality. The use of supplemental oxygen can be life-saving in preterm infants, but may also cause bronchopulmonary dysplasia (BPD). BPD is a multi-organ disorder, which affects up to 50 % of extremely low birth weight infants <1000 g [2]. Survivors with BPD are at increased risk for readmission during the 1st year of life, long-term pulmonary problems and abnormal neurodevelopmental outcome compared to those without BPD [3]. The use of excessive oxygen may lead to an increased production of reactive oxygen species (ROS) such as superoxide anion, hydroxyl radical and hydrogen peroxide [4–6] as well as the expression of pro-inflammatory cytokines [7]. ROS can oxidate or peroxide molecules like lipids, proteins, DNA or RNA and thereby changing their structure or function [8]. Hyperoxia-induced production of ROS is recognized as a major contributor to the development of BPD [9] because the tissue damage and inflammation leads to a developmental arrest of the lung. In recent years, the importance of cytochrome P450 enzymes as well as the aryl hydrocarbon receptor (AhR) has been demonstrated in the metabolism of a number of endogenous and exogenous compounds as well as for oxygen-induced toxicity [10–13]. When a ligand binds, AhR is translocated to the nucleus, where it dimerizes with the aryl hydrocarbon nuclear translocator (ARNT). This complex can thereafter bind to aryl hydrocardon reponse elements (AhRE) like the xenobiotic response elements (XRE) or dioxin reponse elements (DRE). Those response elements function as a cis-acting enhancer in the regulatory domains of genes known as the AhR gene battery. Included in this gene battery are many phase I and phase II detoxification enzymes like CYP1A1, CYP1A2, glutathione S-transferase-alpha, NAD(P)H-quinone reductase-1, UDP-glucuronosyl transferase and aldehyde dehydrogenase [8]. The cytochrome P450 enzymes are a superfamily of heme-containing proteins that are involved in the metabolism of a number of endogenous and exogenous compounds [14]. CYP1A1 is typically induced by planar aromatic hydrocarbons (PAH), but hyperoxia can also induce CYP1A1 [10, 15]. The exact mechanism of induction is unknown, but the AhR seems to play an important role as AhR [10, 12, 13] or CYP1A1 [16] deficient mice react differently on hyperoxic exposure. The induction by hyperoxia decreases if it continues for over 60 h in adult mice models [4, 17]. It has been suggested that CYP1A1 may modulate hyperoxic lung injury by scavenging of the ROS. Possible inducers besides PAH are beta-naphthoflavone (BNF) or 3-methylcholanthrene (3-MC). Also omeprazole (OM) induces CYP1A1 possibly through an AhR-mediated process. Omeprazole does not have the structural features of a typical AhR-ligand [18] and does not seem to bind AhR in a classical way [19, 20]. Possible explanations of transcription of CYP1A1 after administration of omeprazole are the decrease of the interaction forces which keep the AhR complex in a silencing state and thus a transformation of the AhR into a DNA binding form [19]. A nuclear accumulation of a DNA-binding form of the AhR activates CYP1A1 transcription [19]. Also, omeprazole seems to activate CYP1A1 transcription through common regulatory regions as typical AhR ligands [21]. Omeprazole is a proton-pump inhibitor, which is widely used to treat gastro-oesophageal reflux disease or gastric ulcers by inhibition of gastric acid secretion in both humans and animals. It is plasma-protein bound for 97 % and is metabolized through the CYP2C19 and CYP1A4. The half-life is less than 1 h in humans, and elimination occurs renal for 80 %. The drug is approved for administration to infants and neonates as well as pregnant women [22]. Transplacental transfer is present but low, and dependent on the maternal plasma levels [23]. To our knowledge the use of omeprazole in preventing hyperoxia induced lung injury has only been demonstrated so far in vitro [24] and in in vivo mice models [12, 25]. Herein, we test the hypothesis that in a hyperoxic induced lung injury preterm rabbit model, prenatal or neonatal administration of omeprazole (1) induces transcription of CYP1A1 in a dose-dependent way and (2) improves pulmonary outcome both on histological as well as functional level. In addition, we performed an unbiased pulmonary transcriptome analysis by RNA sequencing to assess the pathways targeted by omeprazole to explore its mechanism of action. We used our previously described model [26] of preterm rabbit pups exposed to hyperoxia to assess the induction of CYP1A1 as well as the functional outcome parameters. Methods Animal protocol All experiments were approved by the Ethics committee for Animal Experimentation of the Faculty of Medicine of the KU Leuven. Animals were treated according to current guidelines of animal well-being. Time-mated pregnant does (hybrid of New Zealand White and Dendermonde) were obtained from the animalium of the KU Leuven. All does were housed in separate cages prior to an intervention or the cesarean section with a light–dark cycle of 12 h, a normal room temperature and free access to water and chow. The first animals were used for a dose-finding study in which different administration routes, time points and doses of omeprazole were used to assess induction of CYP1A1 in the lung tissue. In the following animals both lung function as well as histological analysis were analyzed after pre- and postnatal administration of three different doses of omeprazole. For the prenatal fetal injections (dose finding study), the doe was anesthetized with an intra-muscular injection consisting of a mixture of ketamin 35 mg/kg (Ketamin 1000®; CEVA Sante Animale, Libourne, France) and xylazine 6 mg/kg (Vexylan®; CEVA Sante Animale). Anesthesia was maintained using a facemask with isoflurane 1.5–2 % (Isoba®Vet; Abbott Laboratories Ltd., Queenborough, Kent, UK) in oxygen at 2 L/min. Does received a single subcutaneous injection with penicillin G 300.000 IU (Kela Pharma, Hoogstraten, Belgium), medroxyprogesterone acetate 0.9 mg/kg (Depo-Provera, Pharmacia Upjohn, Puurs, Belgium) and buprenorphine 0.03 mg/kg (Vetergesic, Alstoe Limited, York, UK) before the surgical procedure. Surgery was performed in sterile conditions with the rabbit positioned in a supine position on a heating pad. A midline laparotomy was used to exteriorize the uterine horns and thereafter the fetal umbilical vein was punctured with a 24 G needle (Terumo® Surflo® Winged IV catheters, Terumo Europe, Leuven, Belgium) under ultrasound guidance (VEVO® 2100 system, Visualsonic, Toronto, Canada) to inject the fetus. To minimize the risks for miscarriage, maximum six fetuses per doe were injected. For the maternal injections (dose finding as well as functional outcome study), the doe was anaesthetized with an intra-muscular injection of ketamine (35 mg/kg) and xylazine (6 mg/kg) where after the marginal ear vein was punctured to inject omeprazole or saline. To deliver the pups, a cesarean section was performed at day 28 of gestation (term = 31 days) as previously described [26]. The pups used for the dose finding study at time point 0 were immediately euthanized with an intraperitoneal injection of T61® before the first breath and the lungs were removed and snap frozen for determination of CYP1A1 levels. All other pups were placed in an incubator (Dräger Incubator 7310, Dräger®, Lübeck, Germany) at 32 °C in either normoxia (21 % O2) or hyperoxia (>95 % O2). Pups were fed twice daily with a mixture of special formula milk containing 30 % of proteins and 50 % of lipids (FoxValley 30/50, Illinois, US), Bio-Lapis for electrolytes, vitamins and probiotics (Protexin Veterinary, Somerset, UK) and Col-o-cat with a high amount of immunoglobulins (Sanobest, ‘s Hertogenbosch, Netherlands). The amount of feeding increased daily as previously described in detail [26]. Pups were daily injected with prophylactic antibiotics (penicillin and amikacine) from day 2 onwards [26]. Furthermore, daily intraperitoneal injections with placebo (=saline) or omeprazole were performed. Pups were harvested immediately after delivery, at day 3 and 5 for the dose finding study, and at day 5 for the functional analysis. Experimental groups Dose finding study Fifteen does (65 pups) were used for pre- or postnatal maternal, fetal and neonatal injections of omeprazole to assess the level of induction of CYP1A1 (Table 1).Fetal injections: fetuses were administered omeprazole into the umbilical vein under ultrasound guidance at 2 different time points (day 26 or day 27 of gestation, = 24 or 48 h prior to delivery) and with 2 different doses of omeprazole (low = 2 mg/kg or high = 20 mg/kg). The dose was calculated with an estimated fetal weight described by our group previously [27]. This generated four groups, with a minimum of six pups per group (−48 h low/−48 h high/−24 h low/−24 h high). All injected pups were harvested by cesarean section at day 28 of gestation for immediate preservation of lung tissue. Prenatal maternal injections: under anesthesia other does were injected with omeprazole (low dose: 2 mg/kg) in the marginal ear vein 48, 24 or 8 h prior to delivery. All their pups (minimum 7) were harvested immediately after delivery. Neonatal injections: Pups held in hyperoxia were once daily injected intraperitoneally with one of three different doses of omeprazole (low: 2 mg/kg; medium: 10 mg/kg; high: 20 mg/kg) starting immediately after delivery. Lungs were harvested at day 3 and 5. Pups housed in normoxia were only injected with the highest dose of omeprazole (=20 mg/kg/day). Control animals were injected with a similar amount of saline and harvested at the same time points as treated animals. Untouched animals harvested immediately after delivery were used as the control group normalized to 1.Table 1 Dose finding study Administration Injections Harvest Fetal −48 h saline −48 h OM low −48 h OM high −24 h saline −24 h OM low −24 h OM high At delivery Maternal −48 h saline −48 h OM low −24 h saline −24 h OMlow −8 h saline −8 h OM low At delivery Neonatal Daily saline (normoxia) Daily OM high (normoxia) Daily saline (hyperoxia) Daily OM low (hyperoxia) Daily OM med (hyperoxia) Daily OM high (hypreoxia) Day 3 –day 5 Functional assessment Fifteen does (96 pups) were used for the assessment of lung function and histology (morphometry and immunohistochemistry) at day 5 of life Omeprazole was administered prenatally to the doe (2 mg/kg) in combination with daily neonatal intraperitoneal injections (low: 2 mg/kg, medium 10 mg/kg and high 20 mg/kg) to the pups held in hyperoxia. Control animals were injected the same amount of saline. This lead to the following groups: (1) saline-injected, housed in normoxia (normo-saline); (2) saline-injected, housed in hyperoxia (hyper-saline); (3) low dose of omeprazole, housed in hyperoxia (hyper-OMLow); (4) medium dose of omeprazole, housed in hyperoxia (hyper-OMmed) and finally (5) high dose of omeprazole, housed in hyperoxia (hyper-OMhigh). Survival of the pups was assessed on a daily basis. Quantification of CYP1A1 expression Total RNA from freshly harvested and snap frozen lungs was isolated using TriPure Isolation Reagent (Roche Diagnostics, Vilvoorde, Belgium) according to the manufacturer’s guidelines. A maximum of 100 mg tissue was used per sample. The concentration was evaluated with NanoDrop ND-10000 spectrophotometer (NanoDrop Technologies, Wilmington, US). RNA was reversed transcribed to cDNA using TaqMan Reverse Transcription Reagents (Applied Biosystems, Gent, Belgium). Real-time quantitative PCR analysis was performed with an ABI Prism 7000 detection system (Applied Biosystems) using Platinum SYBR Green qPCR Supermix-UDG (Invitrogen Life Technologies, Gent, Belgium). Primers were obtained from Integrated DNA Technologies (IDT, Heverlee, Beglium) and Actin beta (ActB) was used as a housekeeping gene to normalize mRNA levels. The following primer sequences were: FWD 5′- GCACCGCAAGTGCTTCTA -3′ and REV 5′- GCCAATCTCGTCTCGTTTCT -3′ for ActB and FWD 5′- CATCTGTGCCATGTGCTTTG -3′ and REV 5′- TAGCGGAGGATGAGGAAGAA -3′ for CYP1A1. The relative mRNA levels for CYP1A1 were normalized to their ActB content. The ΔΔCt method was used to calculate the fold change in mRNA expression, where ΔCt = Ct(CYP1A1 gene)−Ct(ActB gene) and ΔΔCt = ΔCt (Omeprazole)−ΔCt(saline); fold change = 2(−ΔΔCt). Lung function testing Pups were anesthetized with ketamine (35 mg/kg) and xylazin (6 mg/kg) as previously described [26]. A FlexiVent analysis [28], an invasive measurement which is the gold standard for lung function testing in vivo, was performed on anesthetized pups and the following parameters were assessed: airway resistance (Rn), tissue damping (resistance, G) and tissue elasticity (H) using Primewave-8 forced oscillation and the total lung capacity (A), static compliance (Cst) and static elastance (Est) using the pressure–volume perturbation. All measurements were performed until three consistent measurements were obtained, with a coefficient of determination of >0.95 as the limit to accept the measurement. The average of these three measurements was calculated for further reporting. After the measurements, pups were euthanized using an intracardiac injection of 0.1 mL of T61®. Morphometry After lung function assessment, pups were euthanized, a thoracotomy was performed to remove the lungs and trachea ‘en bloc’. The left bronchus was ligated following which the left lung was removed and snap frozen for determination of CYP1A1 levels. A 20G catheter was inserted into the trachea to pressure-fix the right lung with 4 % paraformaldehyde by immersion and a constant hydrostatic pressure of 25 cm H2O for 24 h. After embedding, 5 µm paraffin sections were stained with hematoxylin and eosin (HE). Morphometric measurements consisted of (1) the linear intercept (Lm), which is a measure of alveolar size, (2) the mean terminal bronchiolar density (MTBD) which is inversely correlated to the number of alveoli supplied by each bronchiole and finally (3) the mean wall transection length (Lmw), a measure of the interalveolar septal thickness [29]. Vascular morphometry was performed on sections stained with Miller’s elastic staining of lungs harvested in pups held in normoxia, or in hyperoxia treated with saline or OMhigh. The external diameter and internal diameter were measured along the shortest axis of peripheral muscularized vessels with less than 100 µm external diameter. These parameters were used to calculate the proportionate medial thickness (% MT = ED−ID/ED × 100) as previously described [29]. Immunohistochemistry Immunohistochemical staining was performed for surfactant protein B (SP-B) to determine airway maturity and for vascular endothelial growth factor (VEGF) and its receptor fetal liver kinase 1 (Flk-1) to assess vascular markers of lung maturity. Slides were incubated with (1) goat anti-mouse polyclonal anti-SP-B (Santa Cruz Biotechnology, Heidelberg, Germany), (2) mouse anti-human monoclonal anti-VEGF (NeoMarkers, Fremont, CA, US) and finally (3) mouse anti-human monoclonal anti-Flk-1 (Santa Cruz Biotechnology). Secondary antibody incubation for SP-B slides was mouse anti-goat biotin (Santa Cruz Biotechnology) plus normal rabbit serum (DakoCytomation) followed by rinsing and incubation with streptavidin alkaline phosphatase (Roche Diagnostics, Vilvoorde, Belgium), rinsing, incubation with NBT solution and counterstaining with Methyl green (DakoCytomation). After rinsing the slides stained for VEGF and Flk-1 they were incubated with peroxidase-conjugated EnVisionTM and reagent (DakoCytomation), rinsed with PBS, incubated with peroxidase substrate solution containing DAB, rinsed with distilled water, counterstained with hematoxylin, dehydrated and mounted. Quantification of positive cells was performed semi-automatically using ImageJ software (1.47v, NIH, Bethesda, Maryland, USA). Ten random images from each slide were processed using the Axioskop platform (Carl Zeiss, Oberkochen, Germany). Transcriptome analysis and validation RT-PCR Transcriptome analysis was done on snap frozen whole left lungs of four saline- and four omeprazole-treated (high dose) rabbits held in hyperoxia. Snap frozen whole left lungs were homogenized using the TissueLyser (Qiagen) and total RNA was isolated with the RNeasy mini-kit (Qiagen), RNA concentration was measured using the Nanodrop 1000 spectrophotometer (Thermo Scientific) and RNA integrity was assessed using the RNA 6000 Nano Kit and the Bioanalyser (Agilent Technologies). mRNA isolation, cDNA conversion and sequencing library preparation was performed using the TruSeq RNA library preparation kit (Illumina). Fastq files were thereafter imported into Array Studio (Omicsoft) and mapped against the Ensembl reference rabbit genome and transcriptome (Build: Oryctolagus cuniculus.oryCun2.64). Expression values were calculated per gene and normalized to ‘reads per kilobase per million reads’ (RPKM) values as described by Mortazavi [30]. Genes with an RPKM value of <1 in all samples were excluded. RPKM values were Log transformed for downstream analysis using the general linear model function in Array Studio for group comparison. Fold changes (FC) were calculated, indicating the ratio of change in gene expression. We applied a fold change cut-off of >1.5 or <−1.5, in order to filter out non-dysregulated molecules. Due to multiple testing, false discovery rates (FDR—Benjamini-Hochberg procedure) were calculated as a measure for statistical significance of the fold change difference observed between the two groups. We considered a transcript change significant if FDR was <0.05. Further analysis was performed using the ingenuity pathway analysis (IPA) software. The IPA ‘upstream regulator analysis’ (URA) predicts upstream regulators by combining the directional expression changes from our mRNA-sequencing, and knowledge from prior experimental reports on causal effects between molecules (endogenous and exogenous), compiled in IPA Knowledge Base. URA calculates a z-score based on the edge of dysregulation of all the downstream molecules and the uniformity of the existing evidence about the upstream–downstream relation, for every upstream regulator known to have a causal effect on at least four dysregulated transcripts. Z-scores <−2 and >2 respectively predict a significant inhibition and activation state of the upstream regulator, regardless of the actual expression level of these molecules. The fresh frozen lung tissue of those animals was furthermore used for RT-PCR analysis to validate three genes: CA4, SCGB1A1 and VEGF. To perform the analysis, the same methodology was used as described earlier, and ActB was used as housekeeping gene. The primer sequences for the genes were: FWD 5′-GGAGTTCTCGAGCAAACTCTAC-3′ and REV 5′-CTGCGGCCTGTGACTTAAA-3′ for CA4; FW 5′-GATGCAGGGATGCAGATGAA-3′ and REV 5′-CACAGTGGGCTCTTCACTATTT-3′ for SCGB1A1 and FWD 5-ATCATGCGGATCAAACCTCA-3′ and REV 5′-CAAGGCCCACAGGGATTTTC-3′ for VEGF. Statistical analysis Kaplan–Meier curves with posthoc testing were used to quantify postnatal survival of rabbit pups using GraphPad Prism 5.0 software (GraphPad, La Jolla, California, USA). For the dose finding study as well as immunohistochemical and RT-PCR analysis, a Kruskal–Wallis test followed by Dunn’s multiple comparison test for post hoc analysis was performed. All other parameters were assessed in a regression-modeling framework, using PROC MIXED with the repeated statement in SAS (Statistical Analysis Software, Cary, USA). This method was chosen because several pups from the same doe were used, such that data are clustered, and the mother can be considered a random effect that is nested within the group. For those parameters, all three omeprazole exposed groups (low, med and high) were compared as a group against the saline treated animals held in normoxia and hyperoxia, and against each other to define significant changes. A value of p < 0.05 was considered statistically significant. All values are expressed as mean ± standard deviation. Results Omeprazole dose finding study Direct fetal injections with omeprazole did not increase CYP1A1 expression as measured at delivery (data not shown). Increase in expression of CYP1A1 after a single maternal dose of omeprazole (2 mg/kg IV) administered 8 h prior to delivery (relative expression of CYP1A1 1.139 ± 0.4105) did not reach statistical significance compared to controls (p = 0.872). Daily neonatal injections in hyperoxia exposed pups showed a dose-dependent rise in expression of CYP1A1 (Fig. 1) both at postnatal day 3 and 5. The induction of CYP1A1 expression at day 3 was significantly higher in pups treated with the medium and high OM dose compared to saline treated animals (p < 0.001 for both). This was not the case for the low dose of omeprazole (p = 0.399). At day 5, the effect persisted, significant for the medium and high dose, yet less pronounced than on day 3 (p = 0.022 and p = 0.002 resp.). In normoxic animals, the administration of omeprazole (high dose) showed a significant induction of CYP1A1 both on day 3 and 5 (p = 0.005 and p < 0.001 resp.). Furthermore, at day 5 the CYP1A1 expression in normoxic saline-treated pups was significantly higher compared to hyperoxic saline treated pups (p = 0.027).Fig. 1 Induction of CYP1A1. Relative expression of CYP1A1 using the ΔΔCt method, measured by RT-PCR in freshly harvested lung tissue at day 3 and day 5 of life. Animals held in normoxia (injected with saline or high dose of omeprazole) or housed in hyperoxia (injected with saline or low/medium/high dose of omeprazole). Bars are mean ± SEM. *p < 0.05 compared to normoxia, saline treated. #p < 0.05 compared to hyperoxia, saline treated All further experiments were performed after a maternal injection of saline or omeprazole (2 mg/kg IV) 8 h prior to delivery, and daily neonatal injections with either saline or omeprazole (low, med, high). Effect of omeprazole on postnatal survival Survival rates were 79.2 % (normo-saline), 57.6 % (hyper-saline), 67.7 % (hyper-OMlow), 65.2 % (hyper-OMmed) and 82.1 % (hyper-OMhigh). These differences did not reach statistical significance (p = 0.229). Omeprazole improves lung function The results of the lung function tests are displayed in Fig. 2. There was no difference in the airway resistance between the different groups. Hyperoxia exposure (hyper-saline) caused a significant increase in both tissue damping and elasticity compared to normoxic controls (p = 0.0011 and p = 0.0032 resp.). Administration of variable doses of omeprazole was associated with a decreased tissue damping as well as elasticity of pups exposed to hyperoxia compared to saline-treated controls (p = 0.0007 and p = 0.0034 resp.). pressure–volume perturbation analysis revealed a significantly improved total lung capacity, static compliance and static elastance compared to saline treated animals held in hyperoxia (p < 0.001, p = 0.0001 and p = 0.0016 resp.)Fig. 2 Lung function tests using forced oscillation technique. a Primewave-8 measurement for Rn airway resistance, G tissue damping and H tissue elasticity. b Pressure–volume perturbation for A total lung capacity, Cst static compliance and Est static elastance. Animals housed in normoxia (injected with saline) or hyperoxia (injected with saline or low/medium/high dose of omeprazole) and harvested day 5. Bars are mean ± SEM. *p < 0.05 compared to normoxia, saline treated. #p < 0.05 compared to hyperoxia, saline treated Omeprazole attenuates prematurity induced lung-developmental arrest Morphometry results are displayed in Fig. 3. There were no significant differences between saline treated animals held in normoxia or hyperoxia for Lm (p = 0.83) neither MTBD (p = 0.23). Hyperoxia however did increase Lmw significantly (p = 0.01). Comparing all treated animals together as one group against saline treated hyperoxic animals, no significant differences were found for Lm (p = 0.08), MTBD (p = 0.34), nor Lmw (p = 0.05). Again, there were significant differences for the highest dose of OM for Lm (p = 0.02) as well as Lmw (p = 0.03). There was no obvious effect on Lm and Lmw following administration of the low or medium dose.Fig. 3 Lung morphometry. Analysis of the Lm linear intercept, MTBD mean terminal bronchiolar density and Lmw mean wall transection length. Animals housed in normoxia (injected with saline) or hyperoxia (injected with saline or low/medium/high dose of omeprazole) and harvested day 5. Bars are mean ± SEM, 1–5 HE staining of the right lung, animals held in normoxia (1), hyperoxia saline treated (2) or hyperoxia treated with a low (3), medium (4) or high (5) dose of omeprazole. *p < 0.05 compared to normoxia, saline treated. #p < 0.05 compared to hyperoxia, saline treated The lungs of hyperoxic animals treated with the highest dose of omeprazole were thereafter compared with saline treated controls held in normoxia and hyperoxia. These results demonstrated a significant increase of the medial thickness of saline treated animals held in hyperoxia compared to normoxic controls. This effect was attenuated after administration of omeprazole (p = 0.007, Fig. 4).Fig. 4 Vascular morphometry. vascular morphometry with the medial thickness (%MT) and Miller staining of a blood vessel <100 μm diameter; animals housed in normoxia (1), hyperoxia treated with saline (2) or hyperoxia treated with OM high (3). *p < 0.05 compared to normoxia, saline treated. #p < 0.05 compared to hyperoxia, saline treated The results of the immunohistochemical stains are shown in Fig. 5. There were no differences between both saline treated groups. Following administration of the medium and high dose of omeprazole, SP B, as well as VEGF and its receptor Flk-1 were significantly increased (p < 0.001 for all three). The lowest dose of omeprazole did not have an effect on any of the measured variables.Fig. 5 Immunohistochemstry. Immunohistochemical staining for Flk-1, VEGF and SP-B performed on lung tissue harvested day 5 of life. Animals housed in normoxia (injected with saline) or hyperoxia (injected with saline or low/medium/high dose of omeprazole) and harvested day 5. Bars are mean ± SEM. *p < 0.05 compared to normoxia, saline treated. #p < 0.05 compared to hyperoxia, saline treated Transcriptome analysis A total number of 315 transcripts were significantly dysregulated applying a filter on fold-change of >1.5 and <−1.5 with a FDR of <0.05. Expression data from all 315 dysregulated genes are displayed in a heat map (Fig. 6) where color intensity reflects the Log2 transformed RPKM gene expression values. Of these 315 transcripts, 271 had known human homologues that are recognized by IPA. Further analysis was performed using these 271 transcripts. Table 2 shows the 10 most up- and most down-regulated transcripts. The most upregulated gene in our dataset is CYP1A1 (FC 96,782; FDR 1,80 × 10−3), the most downregulated gene is PLEKHB2 (FC-55,950; FDR 3,79 × 10−2). By performing URA, we predict that 13 endogenous and exogenous molecules (Table 3) are significant upstream regulators of the transcription changes observed in our dataset.Fig. 6 Heatmap. All transcripts with a fold change of >1.5 or <−1.5 and false discovery rate of <0.05 are shown. Color intensity displayed in the heatmap are the Log2 transformed RPKM gene expression value. These are normalised to relative low (green) and high (red) signal intensities shown in the heatmap key. OM omeprazole, P placebo Table 2 Most up- and down regulated transcripts in omeprazole treated animals Name Full name FC FDR CYP1A1 cytochrome P450, family 1, subfamily A, polypeptide 1 96.782 1.80E−03 CA4 carbonic anhydrase IV 4.507 2.34E−02 S100A1 S100 calcium binding protein A1 3.081 2.87E−02 SCGB1A1 secretoglobin, family 1A, member 1 (uteroglobin) 2.438 3.79E−02 ASB9 ankyrin repeat and SOCS box containing 9 2.326 2.22E−02 SRGN Serglycin 2.259 1.65E−02 PSMC3IP PSMC3 interacting protein (HOP2) 2.087 1.64E−02 UNC13D unc-13 homolog D (C. elegans) 2.059 4.59E−02 PTRHD1 peptidyl-tRNA hydrolase domain containing 1 1.967 4.00E−02 ID2 inhibitor of DNA binding 2, dominant negative helix-loop−helix protein 1.902 1.52E−02 PLEKHB2 pleckstrin homology domain containing, family B (evectins) member 2 −55.950 3.79E−02 S100A12 S100 calcium binding protein A12 −4.497 4.84E−02 CALB2 calbindin 2 −4.056 2.90E−02 KIF5C kinesin family member 5C −2.541 3.08E−02 ANO5 anoctamin 5 −2.428 1.10E−03 PAPPA pregnancy-associated plasma protein A, pappalysin 1 −2.416 1.33E−02 SLC1A3 solute carrier family 1 (glial high affinity glutamate transporter), member 3 −2.352 4.97E−02 OMA1 OMA1 zinc metallopeptidase −2.188 4.22E−02 EDA2R ectodysplasin A2 receptor −2.119 3.75E−02 LIMA1 LIM domain and actin binding 1 −2.034 4.94E−02 Table 3 Endogenous and exogenous molecules predicted to be upstream regulators of the observed transcription changes IPA Entrez gene name z-score P of overlap Upstream regulators predicted to be activated  miR-155-5p (miRNAs w/seed UAAUGCU) microRNA 155-5p 2.611 2.08 × 10−3  miR-16-5p (and other miRNAs w/seed AGCAGCA) microRNA 16-5p 2.425 2.58 × 10−2  miR-30c-5p (and other miRNAs w/seed GUAAACA) microRNA 30c-5p 2.000 3.46 × 10−2  MAPK9 mitogen-activated protein kinase 9 2.189 1.18 × 10−2 Upstream regulators predicted to be inhibited  VEGF vascular endothelial growth factor (as a group) −2.229 1.23 × 10−3  Gentamicin / −2.236 1.68 × 10−1  STAT3 signal transducer and activator of transcription 3 −2.613 8.60 × 10−3  EPAS1 endothelial PAS domain protein 1 −2.219 3.03 × 10−3  NKX2-3 NK2 homeobox 3 −2.091 3.65 × 10−3  ERG v-ets avian erythroblastosis virus E26 oncogene homologue −2.000 2.47 × 10−2  OSM oncostatin M −2.687 7.06 × 10−4  SP600125 / −2.236 5.61 × 10−2  Lipopolysaccharide / −3.072 2.63 × 10−2 Based on current knowledge on the pathophysiological mechanisms involved in BPD and hyperoxia induced lung injury, and by using the analysis tools in IPA, we identified several dysregulated molecules that potentially play a role in the observed beneficiary effect of omeprazole. We demonstrate that omeprazole affects molecules involved in inflammation, reactive oxygen species (ROS) metabolism, vascular growth and development, extracellular matrix remodeling and lung development (Additional file 1: Table S1). Validation RT-PCR was performed on lungs of saline-treated animals held in normoxia and hyperoxia, and animals treated with high dose of omeprazole held in hyperoxia (Fig. 7). Three genes were examined, and all three were significantly downregulated after hyperoxic exposure (p = 0.023, p < 0.001 and p = 0.038 for CA4, SCGB1A1 and VEGF resp.), but only the increase of SCGB1A1 after treatment with omeprazole reached statistical significance.Fig. 7 validation RT-PCR. Fold changes of CA4, SCGB1A1 and VEGF using the ΔΔCt method, measured by RT-PCR in freshly harvested lung tissue at day 5 of life. Animals held in normoxia injected. with saline, or housed in hyperoxia (injected with saline or high dose of omeprazole). Bars are mean ± SEM. *p < 0.05 compared to normoxia, saline treated. #p < 0.05 compared to hyperoxia, saline treated Discussion The principal aim of this study was to assess the effect of omeprazole treatment on hyperoxia induced lung toxicity in preterm rabbit lungs. Our central hypothesis was that omeprazole attenuates the inflammatory and functional changes in the lung tissue, potentially through the induction of the cytochrome CYP1A1. This role of CYP1A1 was demonstrated by Moorthy et al. [5] in a study where rats were pretreated with an inhibitor of CYP1A1 (1-aminobenzotriazole) with significant worsening of hyperoxic lung injury compared to controls [5]. Furthermore, recent studies have demonstrated an attenuating effect of CYP1A1 inducers on hyperoxic lung injury both in vitro [10, 24] as well as in vivo [11–13]. In the first part of our study, we analyzed the dose dependency of omeprazole induced CYP1A1 levels. Prenatal administration with the given doses, either directly (ultrasound guided fetal administration) or indirectly (maternal administration), failed to induce CYP1A1 expression in the lung tissue at birth. This might be explained by the single injection used in all prenatal administrations. Postnatal administered omeprazole efficiently and persistently (d3 and d5) increases the expression of CYP1A1 in pups exposed to normoxia. Hyperoxia has a pronounced effect on the efficiency of induction. In hyperoxia the increases were more modest and decreased over time (levels were around 45 and 5 % of what was observed in pups exposed to normoxia on day 3 and 5 resp.). This less efficient and decreasing expression could be expected, as prolonged exposure to hyperoxia has been shown to decrease the expression of CYP1A1 in adult rats [4]. Our administration dosing and regimens were based on clinical practice and literature data. In infants omeprazole is usually given at 1–2 mg/kg/day, so the low dose reflects this current clinical regimen. Rodent data about the induction of CYP1A1 by omeprazole mention higher dosages (50 mg/kg/day, [12]). Shih [31] described an obvious species specific range of induction of CYP1A1 by omeprazole. After we established that omeprazole is an efficient CYP1A1 inducer, we set out to study the effect of omeprazole treatment on hyperoxia induced changes in preterm rabbit lungs. The exposure of rabbit lungs in the saccular stage of development to persistent high levels of oxygen (>95 %) leads to inflammatory changes and developmental arrest [26]. This becomes obvious in lung function testing by an increase in tissue damping and elasticity, a worsening in static compliance and elastance and a reduction in the total lung capacity. At the microscopic scale this corresponds to thickened septation and less developed alveoli. Omeprazole markedly attenuated all these functional changes already at the lowest dose without obvious dose dependency. Hyperoxia did not lead to changes in resistance in the conducting airways as the major pathological changes are taking place in the lung parenchyma. This improved function would suggest a normalization of the parenchymal architecture. Indeed the mean wall transection length (Lmw), an index of the thickness of alveolar septae decreased to values that were comparable to lungs of pups kept in normoxic conditions, yet only for the highest dose of omeprazole. The distal airway complexity increased as well, evidenced by a smaller alveolar diameter (represented by Lm) but only in the cohort with the highest OM dose. In order to analyze some possible effectors of this changes, we quantified the presence of surfactant protein B (a crude measure for lung maturation) and VEGF and one of its receptors Flk-1 (an important regulator during lung development and angiogenesis). Treatment with omeprazole had a profound dose-dependent effect on all three factors. As the rabbit is not an easy model to study up- and downstream effects (e.g. cytokine levels, immunological changes, etc.) of oxygen toxicity due to the paucity of commercially available reagents, we also performed a transcriptome analysis. Herein we demonstrated that omeprazole affects molecules involved in inflammation, reactive oxygen species (ROS) metabolism, vascular growth and development, extracellular matrix remodeling and lung development. These processes are thought to be key features of hyperoxia induced lung injury and BPD [32]. However, apart from 1 publication on CA4 in gastric mucosa [33], CYP1A1 is the only dysregulated molecule in our dataset which has previously been related directly to omeprazole. Of course, based on our experiment, it is impossible to state whether the other observed gene changes are direct effects of omeprazole, or whether they are secondary changes due to an attenuation of the oxidative stress by e.g. CYP1A1. The results of the validation RT-PCR did demonstrate a significant increase of SCGB1A1 after omeprazole treatment, no significant changes were found for CA4 nor VEGF. Further mechanistic research needs to be performed to validate our transcriptome data, and to further elucidate the pathways involved in the beneficiary effect of omeprazole. Other experiments already demonstrated that omeprazole modulates hyperoxia-induced lung pathology through induction of CYP1A1 [12, 24]. The importance of the CYP1A1 gene has been highlighted again in a recent study using CYP1A1 −/− mice which suggests a mechanistic role for CYP1A1 [16]. Although there was a clear dose-dependent induction of CYP1A1 after daily administrations of omeprazole, the functional outcomes had no linear correlation with CYP1A1 level, as lung function improved even at the lowest dose of omeprazole. Several explanations are possible. Firstly, several in vitro studies have demonstrated both the anti-inflammatory properties by inhibition of neutrophil function as well as anti-oxidant properties by direct scavenging activity against oxygen free radicals [34, 35]. Secondly, the induction of CYP1A1 occurs through activation of the aryl hydrocarbon receptor, as already demonstrated in multiple studies [10, 12, 13, 24]. Activation of the AhR does not only act on CYP1A1 but on a variety of molecules. The aryl hydrocarbon nuclear translocator (ARNT) is an important factor in the pathway of the hypoxia-inducible factors (HIF-1α and HIF-2α) as ARNT serves as a dimerization partner. Stabilization of HIF might lead to an increased expression of growth factors for pulmonary alveolarization and angiogenesis [36]. Furthermore, AhR is suggested to be a suppressor of lung inflammation through its interaction with nuclear factor-κB as has been shown in studies evaluating cigarette smoke [37, 38] or influenza [39]. Recently Stockinger [40] reviewed the anti-inflammatory properties of AhR activation. Lung-resident dentritic cells seem to play an important role by modulating the immune-suppressive enzyme indolamine 2, 3 dioxygenase. Furthermore, AhR activation has an important effect on the differentiation of IL-17 producing T-helper cells, and bronchial fibroblasts who are sensitive to IL-17 can produce inflammatory mediators and chemoattractants such as IL-6 and IL-8 in response to IL-17 stimulation. A recent study performed on neonatal mice has demonstrated a potentiation of hyperoxia induced lung injury after administration of omeprazole [25]. These results are in conflict with our data, but several explanations are possible. First there is the difference in experimental animal, as Shih has demonstrated the important species difference in CYP1A1 induction [31]. Secondly, the study of Shivanna showed an increase in AhR after a short period (4 days), but this effect disappeared after 14 days of hyperoxic exposure. So there might be a time period in which omeprazole has a benefit that disappears when administered for too long with even a toxic effect depending on dose and/or duration of treatment. A more recent study from the same research group has demonstrated that there is no potentiation of hyperoxia induced cell toxicity after administration of omeprazole [41]. This might again be explained by the shorter duration of hyperoxia and administration period of omeprazole, as well as the use of human pulmonary microvascular endothelial cells instead of rodents. More research is needed to address the most effective but also safe dose and duration of treatment, before this can be translated into human care. We acknowledge a number of limitations to our study. First, though lung development of the rabbit is closer to that of man than rodents, it still differs compared to larger animals like baboons or sheep [42]. Another limitation is the lack of data of normoxic animals treated with omeprazole, which in clinical practice would be a group not qualifying for treatment, but could have offered further insides in the effects seen after administration of omeprazole. Furthermore, the high concentration of oxygen (>95 %) and the relative short time interval of 5 days does not mimic entirely the human situation of what is referred to as the “new” BPD. In our previous model [26] we used 7 days of hyperoxic exposure, but decided for a shorter duration in an attempt to decrease mortality. We are however working to expand this model with chronic exposure to oxygen. Studies with larger animals will be needed to assess efficacy of the lowest dose, and safety of higher doses of omeprazole in the prevention of hyperoxia induced lung injury as long term exposure to omeprazole might have negative effects on outcome [25]. Another limitation might be the combined used of pre- and postnatal administration of omeprazole without assessing both independently. Possible side effects of gastric acid inhibitors used in preterm infants are the increased risk of bloodstream and respiratory infections and necrotizing enterocolitis [43] but these complications seem to occur more with the use of H2 antagonists rather than proton pump inhibitors like omeprazole [44]. The strengths of our study are the use of different doses of omeprazole administered through different administration routes, as well as the assessment of both lung function as well as histological parameters. Conclusions Postnatal administration of omeprazole induces CYP1A1 expression in a dose-dependent manner and combined maternal and neonatal administration improves neonatal pulmonary lung function. This is paralleled by a more mature lung architecture in preterm rabbit pups exposed to hyperoxia. As the lowest dose of omeprazole had a positive effect on lung damage, without a clear increase of CYPA1A, further research will be necessary to elucidate the exact mechanism at which omeprazole attenuates the hyperoxic lung injury. Additional file 10.1186/s12967-016-1009-3 Molecules of special interest. Abbreviations Atotal lung capacity ActBactin beta AhRaryl hydrocarbon receptor AhREaryl hydrocarbon response elements BNFbeta-naphthoflavone BPDbronchopulmonary dysplasia Cststatic compliance CYP1A1cytochrome P4501A1 DREdioxin response elements Eststatic elastance FCfold change FDRfalse discovery rate Flk-1fetal liver kinase 1 Gtissue damping Htissue elasticity HSPheat shock protein IPAingenuity pathway analysis Lmlinear intercept Lmwmean wall transection length MTBDmean terminal bronchiolar density OMomeprazole PAHplanar aromatic hydrocarbons Rnairway resistance ROSreactive oxygen species RPKMreads per kilobase per million reads RT-PCRreal-time polymerase chain reaction SP-Bsurfactant protein B URAupstream regulator analysis VEGFvascular endothelial growth factor XRExenobiotic response elements Authors’ contributions All authors have contributed to the conception of the study and (partly) writing of the manuscript. The final version of this manuscript has been approved by all authors listed, furthermore the authors agree to be accountable for all the aspects of the work: JR conception and design of the study, performing all animal experiments, gathering and interpreting all raw data, writing of the manuscript. JJ and TN help in performing all animal experiments, with special attention for lung function and histology/morphometry. Revising of the manuscript. JT conception of the study and all outcome parameters, interpretation of all data and revising the manuscript to its latest form. PB and TS responsible for the transcriptome analysis, both conception of this part of the manuscript as well as processing the tissue for sequencing and running of the entire pathway analysis, writing part of the manuscript and revising it. FL responsible for all PCR data, both creating and evaluating all primers as well as running the PCR and analysis all data. Revising the manuscript till its latest form. JV responsible for the lung function analysis, setup of all measurements and outcome parameters and interpreting those data. Revising the manuscript till its latest form. JD conception of the entire study with input in the different experimental groups and outcome parameters. Rewriting parts of the manuscript till its latest form. All authors read and approved the final manuscript. Acknowledgements We would like to thank Julio Finalet Ferreiro for his support using the array studio software; Rieta Van Bree for the PCR data and Godelieve Verbiest for lung stains. Competing interests The authors declare that they have no competing interests. Availability of date and materials All raw data are submitted as an excel file Ethics approval All experiments were approved by the Ethics committee for Animal Experimentation of the Faculty of Medicine of the KU Leuven (project number 058-2011). Animals were treated according to current guidelines of animal well-being. Funding This project has been funded by a grant from the KU Leuven (OT/13/115) and the Flemish Hercules foundation (large infrastructure investments AKUL/09/033). J. R. is recipient of a Marie Curie Industria-Academia Partnership Program grant (PIAP-GA-2009-251356). J. J is recipient of an Erasmus Mundi Doctoral grant from the European Commission (Framework Agreement number: 2013-0040). J. D. is beneficent of a fundamental clinical research grant of the Fonds Wetenschappelijk Onderzoek Vlaanderen (1801207), and JT of a research grant of the “Klinische Opleidings-en Onderzoeks-Raad” of the University Hospitals Leuven. ==== Refs References 1. Martin JA Hamilton BE Ventura SJ Osterman MJ Mathews TJ Births: final data for 2011 Nat Vital Stat Rep. 2013 62 1 1 69 2. Natarajan G Pappas A Shankaran S Outcomes of extremely low birth weight infants with bronchopulmonary dysplasia: impact of the physiologic definition Early Hum Dev 2012 88 7 509 515 10.1016/j.earlhumdev.2011.12.013 22236557 3. Short EJ Klein NK Lewis BA Cognitive and academic consequences of bronchopulmonary dysplasia and very low birth weight: 8-year-old outcomes Pediatrics 2003 112 5 e359 10.1542/peds.112.5.e359 14595077 4. Moorthy B Nguyen UT Gupta S Stewart KD Welty SE Smith CV Induction and decline of hepatic cytochromes P4501A1 and 1A2 in rats exposed to hyperoxia are not paralleled by changes in glutathione S -transferase-alpha Toxicol Lett 1997 90 1 67 75 10.1016/S0378-4274(96)03832-5 9020404 5. Moorthy B Parker KM Smith CV Bend JR Welty SE Potentiation of oxygen-induced lung injury in rats by the mechanism-based cytochrome P-450 inhibitor, 1-aminobenzotriazole J Pharmacol Exp Ther 2000 292 2 553 560 10640292 6. Yang F Coalson JJ Bobb HH Carter JD Banu J Ghio AJ Resistance of hypotransferrinemic mice to hyperoxia-induced lung injury Am J Physiol 1999 277 6 L1214 L1223 10600893 7. Jobe AH Hillman N Polglase G Kramer BW Kallapur S Pillow J Injury and inflammation from resuscitation of the preterm infant Neonatology. 2008 94 3 190 196 10.1159/000143721 18832854 8. Dalton TP Puga A Shertzer HG Induction of cellular oxidative stress by aryl hydrocarbon receptor activation Chem Biol Interact 2002 141 1–2 77 95 10.1016/S0009-2797(02)00067-4 12213386 9. Saugstad OD Bronchopulmonary dysplasia-oxidative stress and antioxidants Semin Neonatol 2003 8 1 39 49 10.1016/S1084-2756(02)00194-X 12667829 10. Bhakta KY Jiang W Couroucli XI Fazili IS Muthiah K Moorthy B Regulation of cytochrome P4501A1 expression by hyperoxia in human lung cell lines: implications for hyperoxic lung injury Toxicol Appl Pharmacol 2008 233 2 169 178 10.1016/j.taap.2008.08.016 18824009 11. Couroucli XI Liang YH Jiang W Prenatal administration of the cytochrome P4501A inducer, beta-naphthoflavone (BNF), attenuates hyperoxic lung injury in newborn mice: implications for bronchopulmonary dysplasia (BPD) in premature infants Toxicol Appl Pharmacol 2011 256 2 83 94 10.1016/j.taap.2011.06.018 21745492 12. Shivanna B Jiang W Wang L Couroucli XI Moorthy B Omeprazole attenuates hyperoxic lung injury in mice via aryl hydrocarbon receptor activation and is associated with increased expression of cytochrome P4501A enzymes J Pharmacol Exp Ther. 2011 339 1 106 114 10.1124/jpet.111.182980 21768223 13. Shivanna B Zhang W Jiang W Functional deficiency of aryl hydrocarbon receptor augments oxygen toxicity-induced alveolar simplification in newborn mice Toxicol Appl Pharmacol 2013 267 3 209 217 10.1016/j.taap.2013.01.003 23337360 14. Guengerich FP Cytochrome P450: what have we learned and what are the future issues? Drug Metab Rev 2004 36 2 159 197 10.1081/DMR-120033996 15237850 15. Sindhu RK Sakai H Kikkawa Y Effect of hyperoxia on rat pulmonary and hepatic cytochrome P450 monooxygenases Arch Toxicol 2000 73 10–11 540 546 10.1007/s002040050006 10663385 16. Lingappan K Jiang W Wang L Mice deficient in the gene for cytochrome P450 (CYP)1A1 are more susceptible than wild-type to hyperoxic lung injury: evidence for protective role of CYP1A1 against oxidative stress Toxicol Sci 2014 141 1 68 77 10.1093/toxsci/kfu106 24893714 17. Couroucli XI Welty SE Geske RS Moorthy B Regulation of pulmonary and hepatic cytochrome P4501A expression in the rat by hyperoxia: implications for hyperoxic lung injury Mol Pharmacol 2002 61 3 507 515 10.1124/mol.61.3.507 11854430 18. Quattrochi LC Tukey RH Nuclear uptake of the Ah (dioxin) receptor in response to omeprazole: transcriptional activation of the human CYP1A1 gene Mol Pharmacol 1993 43 4 504 508 8386305 19. Lesca P Peryt B Larrieu G Evidence for the ligand-independent activation of the AH receptor Biochem Biophys Res Commun. 1995 209 2 474 482 10.1006/bbrc.1995.1526 7733914 20. Daujat M Peryt B Lesca P Fourtanier G Domergue J Maurel P Omeprazole, an inducer of human CYP1A1 and 1A2, is not a ligand for the Ah receptor Biochem Biophys Res Commun. 1992 188 2 820 825 10.1016/0006-291X(92)91130-I 1280125 21. Yoshinari K Ueda R Kusano K Yoshimura T Nagata K Yamazoe Y Omeprazole transactivates human CYP1A1 and CYP1A2 expression through the common regulatory region containing multiple xenobiotic-responsive elements Biochem Pharmacol 2008 76 1 139 145 10.1016/j.bcp.2008.04.005 18502397 22. Lalkin A Loebstein R Addis A The safety of omeprazole during pregnancy: a multicenter prospective controlled study Am J Obstet Gynecol 1998 179 3 Pt 1 727 730 10.1016/S0002-9378(98)70072-9 9757979 23. Moore J Flynn RJ Sampaio M Wilson CM Gillon KR Effect of single-dose omeprazole on intragastric acidity and volume during obstetric anaesthesia Anaesthesia. 1989 44 7 559 562 10.1111/j.1365-2044.1989.tb11441.x 2774120 24. Shivanna B Chu C Welty SE Omeprazole attenuates hyperoxic injury in H441 cells via the aryl hydrocarbon receptor Free Radic Biol Med 2011 51 10 1910 1917 10.1016/j.freeradbiomed.2011.08.013 21906671 25. Shivanna B Zhang S Patel A Omeprazole attenuates pulmonary aryl hydrocarbon receptor activation and potentiates hyperoxia-induced developmental lung injury in newborn mice Toxicol Sci 2015 148 1 276 287 10.1093/toxsci/kfv183 26272953 26. Richter J Toelen J Vanoirbeek J Functional assessment of hyperoxia-induced lung injury after preterm birth in the rabbit Am J Physiol Lung Cell Mol Physiol 2014 306 3 277 283 10.1152/ajplung.00315.2013 27. DeKoninck P Endo M Sandaite I A pictorial essay on fetal rabbit anatomy using micro-ultrasound and magnetic resonance imaging Prenat Diagn 2014 34 1 84 89 10.1002/pd.4259 24151193 28. Vanoirbeek JA Rinaldi M De Vooght V Noninvasive and invasive pulmonary function in mouse models of obstructive and restrictive respiratory diseases Am J Respir Cell Mol Biol 2010 42 1 96 104 10.1165/rcmb.2008-0487OC 19346316 29. Roubliova XI Deprest JA Biard JM Morphologic changes and methodological issues in the rabbit experimental model for diaphragmatic hernia Histol Histopathol 2010 25 9 1105 1116 20607652 30. Mortazavi A Williams BA McCue K Schaeffer L Wold B Mapping and quantifying mammalian transcriptomes by RNA-Seq Nat Methods 2008 5 7 621 628 10.1038/nmeth.1226 18516045 31. Shih H Pickwell GV Guenette DK Bilir B Quattrochi LC Species differences in hepatocyte induction of CYP1A1 and CYP1A2 by omeprazole Hum Exp Toxicol 1999 18 2 95 105 10.1191/096032799678839699 10100022 32. Chess PR D’Angio CT Pryhuber GS Maniscalco WM Pathogenesis of bronchopulmonary dysplasia Semin Perinatol 2006 30 4 171 178 10.1053/j.semperi.2006.05.003 16860156 33. Puscas I Coltau M Baican M Domuta G Omeprazole has a dual mechanism of action: it inhibits both H(+)K(+)ATPase and gastric mucosa carbonic anhydrase enzyme in humans (in vitro and in vivo experiments) J Pharmacol Exp Ther. 1999 290 2 530 534 10411559 34. Wandall JH Effects of omeprazole on neutrophil chemotaxis, super oxide production, degranulation, and translocation of cytochrome b-245 Gut 1992 33 5 617 621 10.1136/gut.33.5.617 1319381 35. Yoshida N Yoshikawa T Tanaka Y A new mechanism for anti-inflammatory actions of proton pump inhibitors—inhibitory effects on neutrophil-endothelial cell interactions Aliment Pharmacol Ther 2000 14 Suppl 1 74 81 10.1046/j.1365-2036.2000.014s1074.x 10807407 36. Asikainen TM White CW Antioxidant defenses in the preterm lung: role for hypoxia-inducible factors in BPD? Toxicol Appl Pharmacol 2005 203 2 177 188 10.1016/j.taap.2004.07.008 15710178 37. Baglole CJ Maggirwar SB Gasiewicz TA Thatcher TH Phipps RP Sime PJ The aryl hydrocarbon receptor attenuates tobacco smoke-induced cyclooxygenase-2 and prostaglandin production in lung fibroblasts through regulation of the NF-kappaB family member RelB J Biol Chem 2008 283 43 28944 28957 10.1074/jbc.M800685200 18697742 38. Thatcher TH Maggirwar SB Baglole CJ Aryl hydrocarbon receptor-deficient mice develop heightened inflammatory responses to cigarette smoke and endotoxin associated with rapid loss of the nuclear factor-kappaB component RelB Am J Pathol 2007 170 3 855 864 10.2353/ajpath.2007.060391 17322371 39. Head JL Lawrence BP The aryl hydrocarbon receptor is a modulator of anti-viral immunity Biochem Pharmacol 2009 77 4 642 653 10.1016/j.bcp.2008.10.031 19027719 40. Stockinger B Di Meglio P Gialitakis M Duarte JH The aryl hydrocarbon receptor: multitasking in the immune system Annu Rev Immunol 2014 32 403 432 10.1146/annurev-immunol-032713-120245 24655296 41. Patel A Zhang S Moorthy B Shivanna B Omeprazole does not potentiate acute oxygen toxicity in fetal human pulmonary microvascular endothelial cells exposed to hyperoxia Pharmaceutica Analytica Acta. 2015 6 10 424 10.4172/2153-2435.1000424 26779382 42. Albertine KH Progress in understanding the pathogenesis of BPD using the baboon and sheep models Semin Perinatol 2013 37 2 60 68 10.1053/j.semperi.2013.01.001 23582959 43. Omari TI Haslam RR Lundborg P Davidson GP Effect of omeprazole on acid gastroesophageal reflux and gastric acidity in preterm infants with pathological acid reflux J Pediatr Gastroenterol Nutr 2007 44 1 41 44 10.1097/01.mpg.0000252190.97545.07 17204951 44. Kultursay N Gastroesophageal reflux (GER) in preterms: current dilemmas and unresolved problems in diagnosis and treatment Turk J Pediatr 2012 54 6 561 569 23692780

PMC005xxxxxx/PMC5002204.txt

==== Front TrialsTrialsTrials1745-6215BioMed Central London 155210.1186/s13063-016-1552-5ResearchRepresentativeness of the participants in the smoking Cessation in Pregnancy Incentives Trial (CPIT): a cross-sectional study Bessing Barnabas benabala@ymail.com 1Bauld Linda linda.bauld@stir.ac.uk 2Sinclair Lesley l.a.sinclair@stir.ac.uk 2Mackay Daniel F. daniel.mackay@glasgow.ac.uk 3Spence William william.spence@glasgow.ac.uk 4Tappin David M. david.tappin@glasgow.ac.uk 51 World Health Organization, Section of Expanded Programme on Immunization, Ground Floor, UNECA building, Box 3069, Addis Ababa, Ethiopia 2 Centre for Tobacco and Alcohol Studies, School of Health Sciences, University of Stirling, Stirling, FK9 4LA UK 3 Institute of Health and Wellbeing, Public Health, University of Glasgow, 1 Lilybank Gardens, Glasgow, G12 8RZ Scotland UK 4 School of Medicine, University of Glasgow, 1 Lilybank Gardens, Glasgow, G12 8RZ Scotland UK 5 Section of Child Health, School of Medicine, Glasgow University, Scottish Cot Death Trust, West Glasgow Ambulatory Hospital, Yorkhill, Glasgow, G3 8SJ UK 26 8 2016 26 8 2016 2016 17 1 42627 7 2015 13 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background The limited representativeness of trial samples may restrict external validity. The aim of this study was to ascertain the representativeness of the population enrolled in the Cessation in Pregnancy Incentives Trial (CPIT), a therapeutic exploratory study to examine the effectiveness of financial incentives for smoking cessation during pregnancy. Methods CPIT participants (n = 492) were compared with all self-reported smokers at maternity booking who did not participate in the trial (n = 1982). Both groups were drawn from the National Health Service (NHS) Greater Glasgow and Clyde area over a 1-year trial enrolment period. Variables used for comparison were age, area-based deprivation index, body mass index, gestation, and carbon monoxide (CO) breath test level. Chi-square and Mann-Whitney U tests were used to compare groups. Results From January to December 2012, 2474/13,945 (17.7 %) women, who booked for maternity care, self-reported as current smokers (at least one cigarette in the last week). Seven hundred and fifty-two were ineligible for trial participation because of a CO breath test level of less than 7 parts per million (ppm) used as a biochemical cut-off to corroborate self-report of current smoking. At telephone consent 301 could not be contacted, 11 had miscarried, 16 did not give consent and 3 opted out after randomisation, leaving 492 participants for analysis. There were no differences in demographic or clinical characteristics between trial participants, and self-reported smokers not enrolled in the trial in terms of CO breath test (as a measure of smoking level for those with a CO level of 7 ppm or higher), material deprivation (using an area-based measure), maternal age and maternal body mass index. Gestation at booking was statistically significantly lower for participants. Conclusions To ensure that all trial participants were smokers, biochemical validation excluded self-reported smokers with a CO level of less than 7 ppm from taking part in the trial, which excluded 30 % of self-reported smokers who were ‘lighter’ smokers. The efficacy of financial incentives would not have been likely to decrease if ‘lighter’ smokers had been included in the trial population. Trial participants were slightly earlier in their pregnancy at maternity booking, but this difference would not clinically affect the provision of financial incentives if provided routinely. Overall, the trial population was representative of all self-reported smokers with regard to available routinely collected data. Appropriate comparison of trial and target populations, with detailed reporting of exclusion criteria would contribute to the understanding of the wider applicability of trial results. Trial registration Current Controlled Trials ISRCTN87508788. Registered/Assigned on 1 September 2011. Keywords Clinical markersCross-sectional studiesData collectionDemographyTreatment effectivenesshttp://dx.doi.org/10.13039/501100000589Chief Scientist Office (GB)CZH/4/594issue-copyright-statement© The Author(s) 2016 ==== Body Background The fundamental evidence for decision-making in patient care has in the past been based on clinical experience [1]. More recently, clinicians and decision-makers have come to accept that rigorous research is required in addition to experience [2] in order to tackle growing population health challenges. Current intervention protocols and clinical guidelines combine high-quality research from individual patient treatment [3] with clinical experience. Randomised controlled trials (RCTs) with reliable internal validity are considered the ‘gold standard’ [4] to inform decision-making for both individual patients and population health [5–7]. Guidelines to help clinicians make quick but accountable decisions should be guided by accurate and reliable research findings from RCTs [3] to minimise bias. However, trial findings must benefit the target group with similar therapeutic needs (external validity) [4]. Clinicians [8–10] have questioned the effectiveness and reliability of applying trial evidence to target populations outside a trial setting. This may explain delayed use of trial evidence in clinical practice [10, 11]. In the USA, a large body of research has been reviewed showing that ethnic minority groups are under-represented in trials [12]. Much of this review focussed on reasons why ethnic minority groups are more difficult to access and are less likely to agree to take part and be retained in research studies. There are many plausible reasons, but in general trials have insufficient representation from ethnic minority groups for clinicians to feel confident that the trial results should lead to change in treatment for those groups. Trial selection criteria [6, 13, 14] and overall methodology to ensure internal validity may threaten representativeness. For instance, patients with comorbid conditions are often excluded from trials. In subsequent clinical practice, beneficial effects may not be realised perhaps due to side effects associated with a common comorbid condition that is highly represented in the target population. It is important, therefore, to compare trial participants with the target population with regard to demographic, clinical and other variables. Such comparison will help to uncover possible bias associated with nonrepresentative trial populations and allow a judgement to be made about the likely generalisability of trial results [4, 15]. Models of reasons for taking part and not taking part are important [16] to try to understand how better to run research projects that are representative and, therefore, produce results applicable to all groups in the target population. However, this present study did not attempt to understand the reasons behind lack of representativeness, but was limited to establishing representativeness or not, using limited criteria available from both the trial population and the target population from which the trial population was drawn. The Cessation in Pregnancy Incentives Trial (CPIT) [17, 18] was a therapeutic exploratory phase II trial to examine the efficacy and acceptability of using financial incentives to help pregnant smokers to make use of routine Stop Smoking Services (SSS) and or quit smoking during pregnancy. Women were eligible if they were self-reported smokers with an exhaled carbon monoxide level of at least 7 parts per million (ppm) (to biochemically demonstrate current smoking), aged 16 years or older (for consent), less than 24 weeks pregnant (to allow intervention lasting 12–16 weeks by SSS), resident in NHS Greater Glasgow and Clyde (to allow access for research nurses to collect biochemical verification measures of smoking cessation), and able to understand and speak English (for telephone consent). This study used 492 trial participants of 2494 (20 %) women who self-reported as current smokers (at least one cigarette in the last week) at their first maternity booking contact appointment – the target population. The trial showed the intervention to be effective [18] and cost-effective [19]; however, there remains a valid question: ‘Is the intervention likely to be effective if applied to the whole target population?’ Comparisons were, therefore, made between self-reported pregnant smokers not included in the trial or without analysable data (group B; Fig. 1) and trial participants enrolled in CPIT with analysable data (group A; Fig. 1) [17, 18]. Together these groups make up the target population for the intervention. The findings will help policy-makers to determine the representativeness of the trial population and the likely generalisability of financial incentives for smoking cessation during pregnancy.Fig. 1 Study sampling frame and sample. Two thousand four hundred and seventy-four women between January and December 2012 replied ‘Yes’ when asked by their midwife at their maternity booking visit if they had smoked at least one cigarette in the last week. Seven hundred and fifty-two had a carbon monoxide level (CO) less than 7 ppm, or unavailable in 172 (7 %); 13 (<1 %) were under 16 years of age; 243 (10 %) had a gestation ≥24 weeks; 593 (24 %) were not contactable by the Stop Smoking Service (SSS); 23 (<1 %) were non-English speaking; the trial was not discussed by the SSS with 31 (1 %); and permission was not obtained to pass contact details for 113 (5 %). These categories are not mutually exclusive. The remaining 823 (33 % of all self-reported smokers) were eligible for the trial and were passed to the trial team. A records check showed that five had miscarried and contact was not attempted. Three hundred and one could not be contacted by the trial team. Of the 517 contacted, 6 had miscarried and 16 did not consent to the trial. Of the 495 who consented and were randomised, 3 control participants opted out and did not want any data collected in the trial to be used leaving 492 trial participants with data for analysis making up group A. All self-reported smokers who were not included in the trial analysis (2474–492) made up group B (n = 1982) Methods The design of this representativeness study was cross-sectional, comparing the trial population with self-reported pregnant smokers not included in the trial in terms of available routinely collected demographic and clinical data. The setting was maternity booking in early pregnancy in NHS Greater Glasgow and Clyde in the West of Scotland where all pregnant women are routinely asked about their smoking status and a CO breath test is administered as an objective proxy measure of smoking status – about 97 % of pregnant women at first maternity booking in Greater Glasgow and Clyde undergo this test [20]. Client details, including self-reported smoking status and CO level are forwarded using an opt-out approach from the maternity booking appointment to the NHS SSS as recommended in national guidelines [21]. The SSS then telephones all pregnant smokers to discuss smoking and attempting to quit. During this contact, verbal permission was obtained from eligible potential trial participants to pass their contact details to the trial team. Ineligible pregnant smokers were those: with a CO breath test level less than 7 ppm or where no CO breath test level was available; where no contact was possible by the SSS; who were under 16 years of age; who were at 24 full weeks gestation or more at maternity booking (to allow SSS support to take place prior to expected delivery); who were not approached by SSS about the trial; who did not speak English; and who refused permission to pass contact details to the research team. The trial team attempted to contact pregnant smokers to discuss trial participation and a few who were contacted were no longer pregnant. The remaining patients were asked for verbal consent to take part in the trial. Those who agreed to take part were enrolled into the trial. Group A are trial participants (n = 492). Group B are self-reported smokers at maternity booking who were not included in the trial when analysed (n = 1982). Variables used to compare group A with group B Variables from the two population groups were gathered from routinely collected maternity service data: CO breath test level, height, weight, postcode, age and gestation. Gestation was recorded as estimated gestation (in weeks calculated from recall of last menstrual period) at the date of referral to the SSS and ranged from 4 to 40 weeks. Body mass index (BMI) was derived from maternal height and weight using the formula; Wtinkilos/Ht2inmetres. Finally, material deprivation was measured using the Scottish Index of Multiple Deprivation (SIMD) [22] with postcode as proxy. ‘SIMD ranks small areas (called datazones) from most deprived (ranked 1) to least deprived (ranked 6505) using 38 indicators from seven domains; income, employment, health, education, skills and training, housing, geographic access and crime’ [22]. In order to compare the CO levels (and, therefore, heaviness of smoking [23]) between those enrolled in the trial and those who chose or could not be enrolled in the trial, comparison between groups A and B was made after excluding those with a CO level less than 7 ppm. The aim was to show that apart from the exclusion criterion of a CO level less than 7 ppm the groups were similar in heaviness of smoking. Statistical analyses Statistical analyses were performed with Stata 12.10 [24]. Descriptive statistics (percentages, medians and interquartile ranges) were derived for all variables. Univariate statistical comparison used Pearson X2 test for linear trend and Mann-Whitney U tests to ascertain possible group differences between group B – self-reported smokers not included in the trial and group A – the trial participants. We used multivariable logistic regression to control for the possible effects of confounding among our variables. Nonlinearities for continuous variables were tested using polynomial terms. Results Study setting and sample characteristics Of the women who booked for maternity care, 2474/13,945 (17.7 %) were self-reported smokers, 172 (7.0 %) were excluded from all analyses because their CO level was missing or unknown, 752 (30.4 %) were excluded because their CO level was below 7 ppm, and 727 (29.4 %) did not meet the other inclusion criteria (Fig. 1). After checking routine records, 5 (<1 %) had miscarried and, therefore, contact was not attempted, and for 301 (12 %) contact was unsuccessful. Out of the 517 women contacted, consent for enrolment in the trial was not given by 16 women, 6 had miscarried, and the remaining 495 were enrolled. Three participants withdrew, refusing permission for their data to be used, leaving 492 trial participants’ data for analysis (Fig. 1) as group A. Those not included in the trial, 1982 (2474 − 492), made up the comparison group B. Comparison of groups The median CO breath test result for both self-reported smokers at maternity booking not included in the trial (excluding those with a CO level less than 7 ppm) and the trial participants was 12 ppm (Table 1). A statistically significant difference (P < 0.001) was seen for gestation at booking with those in the trial booking at slightly earlier gestation (median 12.9 weeks compared with 13.3 weeks for those not included in the trial).Table 1 Summary of baseline characteristics of nontrial (group B) and trial (group A) groups taken from all self-reported smokers at maternity booking from January to December 2012 Characteristics Nontrial group B (n = 1982) Trial group A (n = 492) P value CO level (ppm) for those with CO ≥7 ppm, median (interquartile range) 12 (10–16) 12 (10–17) 0.98 Gestation (weeks), median (interquartile range) 13.3 (11.7–15.3) 12.9 (11.3–14.3) <0.001a Deprivation quintiles, n (%)  Most deprived   1st 1331 (67.4) 322 (65.5) 0.58b   2nd 319 (16.2) 83 (16.9)   3rd 168 (8.5) 51 (10.4)   4th 100 (5.1) 21 (4.3)   5th 56 (2.8) 15 (3.1)  Least deprived   Missing 8 0  Age under 20 272 (13.7) 59 (12.0) 0.09c   20–24 599 (30.3) 126 (25.6)   25–29 491 (24.8) 136 (27.6)   30–34 359 (18.1) 113 (23.0)   35+ 258 (13.0) 58 (11.8)   Missing 3 0 BMI categories, n (%)   Underweight 90 (4.6) 19 (4.0) 0.15d   Normal weight 921 (47.5) 215 (44.7)   Overweight 532 (27.5) 139 (28.9)   Obese 395 (20.4) 108 (22.5)   Missing 44 11 aMann-Whitney U test bChi-square test of trend cChi-square test of trend dChi-square test of trend BMI body mass index, CO carbon monoxide Table 2 shows the odds ratios for a multivariable logistic regression of trial status on gestation, age category, deprivation category and BMI category. All variables with the exception of estimated gestation are not statistically significant. The effect of estimated gestation is that of increasing the odds into the trial. However, the square term shows that at higher gestational ages the odds of being in the trial declines significantly, P = 0.007.Table 2 Odds ratios (ORs) for a multivariable logistic regression of trial status on gestation, age category, deprivation category and body mass index (BMI) category Characteristics OR (95 % CI) P value Gestation (weeks) 1.12 0.122 Gestation (squared) 0.99 0.007 Deprivation quintiles, n (%) Most deprived  1st 1.0  2nd 1.13 (0.86, 1.49) 0.383  3rd 1.25 (0.89, 1.77) 0.201  4th 0.81 (0.50, 1.33) 0.416  5th 1.08 (0.59, 1.96) 0.806 Least deprived  Age  under 20 1.0  20–24 0.91 (0.64, 1.29) 0.590  25–29 1.20 (0.85, 1.71) 0.301  30–34 1.34 (0.93, 1.93) 0.117  35+ 1.10 (0.72, 1.67) 0.651 BMI categories, n (%)  Underweight 1.0  Normal weight 1.19 (0.70, 2.03) 0.526  Overweight 1.38 (0.80, 2.39) 0.251  Obese 1.35 (0.77, 2.37) 0.288 Discussion Comparing trial participants and self-reported smokers not included in the trial Overall, there was no evidence from available routinely collected demographic or clinical characteristics that trial participants differed in a way that would make a difference to implementation or effectiveness of the incentives intervention. This suggests that the trial sample was representative of the whole target population of self-reported smokers identified at maternity booking in NHS Greater Glasgow and Clyde (Table 1). This is similar to randomised controlled trial studies in cannabis dependence [25] and also in smoking and chewing tobacco cessation trials [26, 27] none of which were among pregnant women. However, other studies [28, 29] found their trial populations to be nonrepresentative in terms of demographic and clinical characteristics. Evidence from systematic reviews [8, 30] indicates that pharmaceutical trials often exclude the majority of patients based on age, comorbidity, multiple drug use and other unexplained reasons, resulting in nonrepresentative samples. Our results indicate that the CPIT population was a representative sample of all self-reported smokers identified at first maternity booking contact visit (the target population) in NHS Greater Glasgow and Clyde Health Board area. An exhaled CO level of less than 7 ppm excluded 30 % of self-reported smokers. The aim was to make sure that all participants were smokers rather than nonsmokers masquerading as smokers to try to receive incentive payments. A previous study in the same geographical area [31], before the incentives trial, showed that 36 % of self-reported smokers at maternity booking (with no advantage gained by calling themselves smokers) had a CO level of less than 7 ppm. It therefore seems likely that self-reported smokers who were excluded because their CO breath test gave a reading less than 7 ppm were true smokers but light smokers [23]. Light smokers are more likely to successfully quit with effective support [32]. Therefore, excluding these women is not likely to have overestimated the true effectiveness of financial incentives if offered to all self-reported smokers identified at maternity booking. Gestation was significantly different with self-reported smokers who were not included in the trial, having a higher gestation at maternity booking of 13.3 weeks compared to trial participants’ 12.9 weeks. However, this difference is small and likely to be skewed by the few women who book late during pregnancy as women with greater than 24 weeks gestation at maternity booking were also excluded. This difference in gestation between groups A and B would not be large enough to affect implementation of financial incentives if they were rolled out across the health board area. Achieving representativeness and external validity alongside internal validity is a difficult task especially in clinical trials. This study, using data from a novel trial for smoking cessation in pregnancy, showed that routinely collected data from maternity booking can provide large general target and trial population groups for adequate comparison, essential when assessing representativeness. Data items available for this study were by no means exhaustive, but those available show that participants were of similar age, material deprivation status, heaviness of smoking addiction and gestation to those women who did not take part from the target population. Therefore, it seems likely that an effective smoking cessation intervention demonstrated among trial participants would be generalisable to the whole of the target population of pregnant smokers identified at the first maternity booking visit. Limitations of the study Consideration of other important covariates with regard to representativeness of the trial population which were not available – such as marital status, education, partner smoking status, employment, symptoms of psychological distress, and others known to influence smoking during pregnancy – may have produced different results. Reasons why the study population might not be representative such as ‘program benefits’ and ‘barriers to participation’ [16] were not considered but are of great importance when beginning to design a public health intervention strategy and through the stages of testing the intervention – pilot, definitive trial and implementation. Implications for practice To help assess applicability and to enable clinicians to replicate intervention strategies to improve health, researchers can make use of available routinely collected data in intervention studies in a relatively inexpensive way. This would allow assessment of the representativeness of the study population compared with the target population from which it was drawn and for whom the intervention is planned. Detailed reporting of reasons for exclusions and presentation of results by age, sex, socioeconomic status and other important characteristics, such as ethnic group, could help clinicians and policy-makers to take more informed decisions on whether to implement interventions. Enrolment and retention or not in public health intervention trials should be modelled carefully at every stage of the research process from design through pilot and definitive trial stages to the end of implementation [16]. In this way, barriers to participation and retention in trials which often mirror barriers to implementation can be understood and removed if possible at an early stage. If barriers cannot be removed or the program benefits cannot be effectively enhanced to overcome barriers, then public health interventions can be abandoned or redesigned before large amounts of public money are spent on interventions that will not be effective for the target population. Conclusion This study shows that the trial population who took part in the Cessation in Pregnancy Incentives Trial (CPIT) were representative of the target population of pregnant smokers identified at maternity booking in the trial catchment area in relation to important demographic and clinical variables. This finding supports the view that the effectiveness of the intervention examined in the trial would be generalisable if offered to all pregnant smokers identified in the trial area at maternity booking and perhaps more widely. To underpin guidelines relevant to the majority of the population that call for quality, efficient, effective and holistic health care, trial participant representativeness to the target population should be examined alongside internal validity in the ranking of evidence. Abbreviations BMIBody mass index COCarbon monoxide CPITCessation in Pregnancy Incentives Trial NHSNational Health Service ppmParts per million RCTRandomised controlled trial SSSNHS Stop Smoking Service Acknowledgements Mrs Lydia A. Yamoah: Principal of Midwifery and Health Care Assistants Training College, Mampong Ashanti. Ghana. She contributed critically to the drafting, fine tuning and proofreading the draft. Commonwealth Commission UK for providing financial support to undertake the Master’s degree program and the thesis project. We would like to acknowledge the help and support of NHS Greater Glasgow and Clyde R&D department, in particular Dr Roma Armstrong and Brenda Colvin, without whose help this trial would not have been possible. Funding This study was funded by a grant from the Chief Scientist Office Scottish Government CZH/4/594, Glasgow Centre for Population Health, NHS Greater Glasgow and Clyde Endowments, the Royal Samaritan Endowment Fund and Glasgow Children’s Hospital Charity. Availability of data and materials Raw data from this trial may be made available on request by Professor Tappin. Authors’ contributions BB conceived the study, was involved in designing the study, data analyses and drafting of the article. LB was involved in designing the study and drafting the article. LS was involved in data acquisition and cleaning, designing the study, and drafting and proofreading the article. DM was involved in data cleaning, analyses and proofreading the article. WS was involved in designing the study, and drafting and proofreading the article. DT conceived the study, was involved in designing the study, and drafting and proofreading the article. All authors read and approved the final manuscript. Authors’ information 1. Barnabas Bessing (field epidemiologist): WHO International STOP Consultant. Has extensive midwifery/nursing teaching and public health working experience. Undertook this work as a Master’s thesis, and was involved in design, data analyses, and writing the paper. 2. Linda Bauld (Co-principal Investigator): Professor of Health Policy, University of Stirling and UK Centre for Tobacco and Alcohol Studies, has extensive experience of quantitative work related to smoking cessation. She supported the trial on a day-to-day basis and was involved in writing the paper. 3. Lesley Sinclair (Trial Manager): has experience of data management and the management of clinical trials. She ran the trial on a day-to-day basis, assisted with study design, and data acquisition and cleaning for analyses. 4. Daniel F Mackay: Reader in Public Health, has extensive experience in quantitative research work and data analyses. He assisted in the organisation, analyses of the data and was involved in the supervision of the project work. 5. William Spence: University Teacher in Public Health, has extensive experience in qualitative work and teaching. He was fully involved in the project design, supervision, and in writing the paper. 6. David Tappin (Co-Principal Investigator): Professor of Clinical Trials for Children is based at Glasgow University. He coordinated and managed the overall running of the trial, fully supervised this project and was closely involved in data analyses and in writing the paper. Competing interests The authors declare that they have no competing interests. Consent for publication Not applicable. Ethics approval and consent to participate The Cessation in Pregnancy Incentives Trial (CPIT) was approved by the NHS West of Scotland Research Ethics Committee 2 (11/AL/0204) and a substantial amendment granted by the NHS West of Scotland Research Ethics Subcommittee 4 allowed anonymous use of routine data on nontrial participants without direct patient consent. The Caldicott guardian also granted approval for BB to use patient-level data without direct patient identifiers. Trial participants in CPIT gave individual consent. ==== Refs References 1. Travers J Marsh S Williams M Weatherall M Caldwell B Shirtcliffe P External validity of randomised controlled trials in asthma: to whom do the results of the trials apply? Thorax. 2007 62 219 23 10.1136/thx.2006.066837 17105779 2. Herland K Akselsen JP Skjønsberg OH Bjermer L How representative are clinical study patients with asthma or COPD for a larger ‘real life’ population of patients with obstructive lung disease? Respir Med. 2005 99 11 9 10.1016/j.rmed.2004.03.026 15672843 3. Fortin M Dionne J Pinho G Gignac J Almirall J Lapointe L Randomized controlled trials: do they have external validity for patients with multiple comorbidities? Ann Fam Med. 2006 4 104 8 10.1370/afm.516 16569712 4. Petersen MK Andersen KV Andersen NT Soballe K ‘To whom do the results of this trial apply?’: external validity of a randomized controlled trial involving 130 patients scheduled for primary total hip replacement Acta Orthop. 2007 78 12 8 10.1080/17453670610013367 17453387 5. Antes G The evidence base of clinical practice guidelines, health technology assessments and patient information as a basis for clinical decision-making Z Arztl Fortbild Qualitatssich. 2004 98 180 4 15250384 6. Le Strat Y Rehm J Le Foll B How generalisable to community samples are clinical trial results for treatment of nicotine dependence: a comparison of common eligibility criteria with respondents of a large representative general population survey Tob Control. 2011 20 338 43 10.1136/tc.2010.038703 21212379 7. Seale JP Gebski VJ Keech AC Generalising the results of trials to clinical practice Med J Aust. 2004 181 558 60 15540970 8. Jones R Jones RO McCowan C Montgomery AA Fahey T The external validity of published randomized controlled trials in primary care BMC Fam Pract. 2009 10 5 10.1186/1471-2296-10-5 19152681 9. Okuda M Hasin DS Olfson M Khan SS Nunes EV Montoya I Generalizability of clinical trials for cannabis dependence to community samples Drug Alcohol Depend. 2010 111 177 81 10.1016/j.drugalcdep.2010.04.009 20537813 10. Dawes M Co-morbidity: we need a guideline for each patient not a guideline for each disease Fam Pract. 2010 27 1 2 10.1093/fampra/cmp106 20081211 11. Flather M Delahunty N Collinson J Generalizing results of randomized trials to clinical practice: reliability and cautions Clin Trials. 2006 3 508 12 10.1177/1740774506073464 17170034 12. Yancey AK Ortega AN Kumanyika SK Effective recruitment and retention of minority research participants Annu Rev Public Health. 2006 27 1 28 10.1146/annurev.publhealth.27.021405.102113 16533107 13. Zimmerman M Chelminski I Posternak MA Generalizability of antidepressant efficacy trials: differences between depressed psychiatric outpatients who would or would not qualify for an efficacy trial Am J Psychiatry. 2005 162 1370 2 10.1176/appi.ajp.162.7.1370 15994721 14. Hoertel N Le Strat Y De Maricourt P Limosin F Dubertret C Are subjects in treatment trials of panic disorder representative of patients in routine clinical practice? Results from a national sample J Affect Disord. 2013 146 383 9 10.1016/j.jad.2012.09.023 23084184 15. Cole SR Stuart EA Generalizing evidence from randomized clinical trials to target populations The ACTG 320 Trial Am J Epidemiol. 2010 172 107 15 10.1093/aje/kwq084 20547574 16. Spoth R Redmond C Parent motivation to enroll in parenting skills programs: A model of family context and health belief predictors J Fam Psych. 1995 9 294 310 10.1037/0893-3200.9.3.294 17. Tappin DM Bauld L Tannahill C de Caestecker L Radley A McConnachie A The Cessation in Pregnancy Incentives Trial (CPIT): study protocol for a randomised controlled trial Trials. 2012 13 113 10.1186/1745-6215-13-113 22818493 18. Tappin D Bauld L Purves D Boyd K Sinclair L MacAskill S Financial incentives for smoking cessation in pregnancy: randomised controlled trial BMJ. 2015 350 h134 10.1136/bmj.h134 25627664 19. Boyd KA Briggs AH Bauld L Sinclair L Tappin D Are financial incentives cost-effective to support smoking cessation during pregnancy? Addiction 2016 111 2 360 70 10.1111/add.13160 26370095 20. Director of Public Health, Greater Glasgow and Clyde NHS Board. Update on progress with recommendations from the Director of Public Health report: keeping health in mind. Board paper 12/52. 2014. http://library.nhsggc.org.uk/mediaAssets/Board%20Papers/12-52.pdf page 6. Accessed 18 Feb 2015 21. National Institute of Health and Care Excellence: quitting smoking in pregnancy and follow childbirth (PH26).2010. https://www.nice.org.uk/guidance/PH26. Accessed 18 Feb 2015 22. Scottish Index of Multiple Deprivation 2012. http://simd.scotland.gov.uk/publication-2012 Accessed 18 Feb 2015 23. Deveci SE Deveci F Acik Y Ozan T The Measurement of exhaled carbon monoxide in healthy smokers and non-smokers Respir Med. 2004 98 551 6 10.1016/j.rmed.2003.11.018 15191041 24. Stata C. Stata Statistical Software Release 7.0: Programming: Stata Corporation, 2001. 25. Frewen AR Baillie AJ Montebello ME Are cannabis users who participate in a randomized clinical trial different from other treatment seekers? J Subst Abuse Treat. 2009 36 339 44 10.1016/j.jsat.2008.07.004 18834692 26. Graham AL Papandonatos GD DePue JD DePue JD Pinto BM Borrelli B Lifetime characteristics of participants and non-participants in a smoking cessation trial: implications for external validity and public health impact Ann Behav Med. 2008 35 295 307 10.1007/s12160-008-9031-1 18414962 27. Howard-Pitney B Fortmann SP Killen JD Generalizability of findings from a chewing tobacco cessation clinical trial Nicotine Tob Res. 2001 3 347 52 10.1080/14622200110050439 11694202 28. Moore DA Goodall RL Ives NJ Hooker M Gazzard BG Easterbrook PJ How generalizable are the results of large randomized controlled trials of antiretroviral therapy? HIV Med. 2001 1 149 54 10.1046/j.1468-1293.2000.00019.x 11737343 29. Licht RW Gouliaev G Vestergaard P Frydenberg M Generalisability of results from randomised drug trials. A trial on antimanic treatment Br J Psychiatry 1997 170 264 7 10.1192/bjp.170.3.264 9229034 30. Van Spall HGC Toren A Kiss A Fowler RA Eligibility criteria of randomized controlled trials published in high-impact general medical journals: a systematic sampling review JAMA. 2007 297 1233 40 10.1001/jama.297.11.1233 17374817 31. Usmani ZC Craig P Shipton D Tappin D Comparison of CO breath testing and women’s self-reporting of smoking behaviour for identifying smoking during pregnancy Subst Abuse Treat Pr. 2008 3 4 10.1186/1747-597X-3-4 32. Kotz D Fidler J West R Very low rate and light smokers: smoking patterns and cessation-related behaviour in England, 2006–11 Addiction 2012 107 5 995 1002 10.1111/j.1360-0443.2011.03739.x 22126678

PMC005xxxxxx/PMC5002205.txt

==== Front BMC Vet ResBMC Vet. ResBMC Veterinary Research1746-6148BioMed Central London 79710.1186/s12917-016-0797-2Research ArticleComparison of bulk milk antibody and youngstock serology screens for determining herd status for Bovine Viral Diarrhoea Virus Booth Richard E. +441707666211rbooth@rvc.ac.uk Brownlie Joe jbrownlie@rvc.ac.uk Royal Veterinary College, Hawkshead Lane, North Mymms, Hertfordshire, AL9 7TA UK 26 8 2016 26 8 2016 2016 12 1 17715 12 2015 10 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background This paper examines the use of Bulk Milk antibody (BM Ab), Youngstock (YS) serology (Check Tests) and Bulk Milk PCR (BM PCR) for determining the presence or absence of animals persistently infected (PI) with Bovine Viral Diarrhoea Virus (BVDV) within a herd. Data is presented from 26 herds where average herd sizes were 343 and 98 animals for dairy and beef units respectively. Seventeen herds had sufficient data to analyse using Receiver Operating Characteristic (ROC) and probability curves enabling calculation of the sensitivity and specificity of BM Ab and YS Check tests for determining the presence of PI animals within herds in this dataset. Results Using BM Ab to screen a herd for the presence of PI animals, achieved a herd level sensitivity and specificity of 80.00 % (44.39–97.48 %) and 85.71 % (42.13–99.64 %) respectively (95 % confidence intervals quoted). Sensitivity and specificity of YS Check Tests at a cut off of 3/10 Ab positive YS were 81.82 % (48.22–97.72 %) and 66.67 % (22.28–95.67 %) respectively (95 % confidence interval). These results were achieved by comparing the screening tests to whole herd PI searches that took place 1–19 months after the initial screen with a mean interval of 8 months. Removal of this delay by taking BM samples on the day of a whole herd test and simulating a YS Check Test from the herd test data produced improvements in the reliability of the Check Tests. BM Ab sensitivity and specificity remained unchanged. However, the Check Test sensitivity and specificity improved to 90.9 % (58.72–99.77 %) and 100 % (54.07–100 %) respectively (95 % confidence interval) at a cut of off 2.5/10 Ab positive animals. Our limited BM PCR results identified 5/23 dairy farms with a positive BM PCR result; two contained milking PIs, two had non-milking PIs and another had no PIs identified. Conclusions Delaying a PI search following an initial herd screen decreased the diagnostic accuracy and relevance of our results. With careful interpretation, longitudinal surveillance using a combination of the techniques discussed can successfully determine farm status and therefore allow changes in BVDV status to be detected early, thus enabling prompt action in the event of a BVDV incursion. Keywords Bovine Viral Diarrhoea VirusHerd screenHerd statusBulk milk antibodyYoungstock check testhttp://dx.doi.org/10.13039/501100000277Department for Environment, Food and Rural AffairsSE0777Brownlie Joe issue-copyright-statement© The Author(s) 2016 ==== Body Background Bovine Viral Diarrhoea Virus (BVDV) is an economically important pestivirus recognised for causing infertility, immunosuppression and, as a consequence, high levels of secondary disease in cattle herds worldwide [1–7]. The losses associated with BVDV infection are well documented and the disease is commonly quoted to cost the UK farming industry £40 million per year primarily as a consequence of sub-optimal fertility and immunosuppression [8]. At the individual animal level, the most recent published estimates of losses due to BVDV infection are €32 and €63 per cow per year in beef and dairy systems respectively within the Irish cattle sector [9]. Similar figures of £37 per cow per year exist for beef suckler herds in the UK [10], but less detail is reported at the individual cow level within the UK dairy sector. Significantly, a number of European countries have recognised the losses caused by BVDV and undertaken national eradication; whilst Norway, Sweden, Finland and Denmark have eradicated it [11, 12], other countries e.g. Austria, Switzerland, Germany, Belgium, Ireland and Scotland are in the process of eradication [11, 13–15]. The Scottish and Irish programmes include control measures with regulations to prevent the sale of persistently infected (PI) carrier animals [13, 14]. This will impact on further eradication efforts throughout England and Wales since clearly it would be beneficial for these programmes to be compatible with those underway in Scotland and Ireland in order to facilitate trade. The persistently infected (PI) animal is infected as a foetus in the first trimester of pregnancy and born immunotolerant to the infecting strain of BVDV; thereafter becoming a lifelong shedder of the virus [16, 17] excreting large quantities of BVDV in all body secretions [18, 19]. Control of PI animals is critical to the successful control of BVDV transmission both within and between herds. Lindberg et al. 1999 [20] drew conclusions from epidemiological studies of BVDV and stated that “in practice, a herd is not infected until one or more PIs have been established” and, in addition to this, “BVDV cannot persist in a herd where contacts between PI animals and susceptible animals in early pregnancy do not occur”. This highlights the pivotal role of PI animals in the epidemiology of BVDV and the need to both identify and cull them as part of any successful BVDV eradication programme. The more rapid the culling of PIs from infected herds, the greater the health and productivity advantage. Two approaches to BVDV eradication have been utilised by those European countries that have programmes in place. In some, often where seroprevalence was deemed to be high, the decision was made at the outset to test directly for PI animals at the national level without establishing status at the herd level first i.e. Switzerland and Ireland [14, 21]. In both programmes, specialised ear tags were used to collect ear notch tissue samples for testing for BVDV antigen or RNA. Within the Swiss programme, the aim was to test the whole cattle population [21] whereas, in Ireland it has been compulsory to test newborn animals as part of the official tagging process from January 1st 2013 onwards [14]. In contrast, the Scandinavian and Scottish programmes have elected firstly to establish individual herd status and then proceed with eradication in infected herds [13, 20]. At the outset of these eradication programmes, effective herd level screening was essential in order to distinguish between BVDV-infected (contains PI animals) and BVDV-free herds (no PI animals present). This allows resources to be focussed on either identification and culling of PI animals or on surveillance and protection. It is always essential that effective biosecurity is implemented to prevent the re-introduction of infection [20, 22, 23]. Depending upon local circumstances and, in particular cattle density, enhanced biosecurity may be combined with vaccination to protect against herd re-infection [24]; this is especially important where biosecurity does not meet the stringent standards outlined in the Cattle Health Certification Standards (CHeCS) technical document [25]. Within England and Wales, many farmers will not know their current BVDV status and it is hoped that this manuscript will provide practitioners with further advice on the use of appropriate herd level diagnostics to determine whether PI animals are likely to be present. Current recommendations for UK herd BVDV accreditation are documented in the CHeCS technical document [25]. Two options are permissible for herd BVD accreditation. The first involves screening successive calf crops for BVDV antigen and accreditation is achieved following negative test results over a 2 year period. The second method involves antibody (Ab) screening of bulk milk samples and youngstock (YS) cohorts (commonly referred to as ‘Check Tests’ or ‘Spot Tests’) in order to accurately assess herd BVDV status. BM Ab has played a major role in the control of cattle diseases since the mid-1980s [26] and was widely used within the Scandinavian BVDV control programmes to establish dairy herd status. The CHeCS technical document recommends the use of quarterly BM Ab screening tests at the point of establishing and monitoring herd status although it recognises a positive result may occur in a herd that has been historically infected; at this point the screening of first lactation animals may provide a more up to date status of the dairy herd [25]. The recommendations within CHeCS for YS screening in the UK are largely based on the work conducted in Denmark and the USA by Houe 1992, Houe 1994 and Houe et al. 1995 [27–29]. The first study determined that herds ranging in size from 96 to 324 animals (with a mean of 135) could be grouped into BVDV infected and BVDV-free herds by testing only three or five YS 6–18 months of age [27]. In a further study, forty-two herds were investigated ranging in size from 40 to 157 animals and in a detailed survey of seven of these herds, low levels of YS seroprevalence were found to correlate with the absence of PI animals (0–1/10 YS Ab positive at the spot test) and high levels of YS seroprevalence correlated with the presence of PIs (8–10/10 YS Ab positive at the spot test) [28]. An additional test available for milk samples is bulk milk PCR (BM PCR) which enables the screening of bulk tank milk for the presence of PI animals. This allows the user to establish the status of the milking animals that contribute to the sample, but not necessarily the entire herd. Initially BM PCR was reported to be sensitive enough to detect one PI animal in 160 milking animals [30], but more recently this has increased to an upper limit of 300 recommended by many diagnostic laboratories [31] with further laboratories reporting a higher diagnostic sensitivity allowing this upper limit to increase further to 1000 animals (Karen Bond, National Milk Laboratories, Personal Communication). This paper examines and discusses the use and practical implications of BM Ab, YS Check Tests and BM PCR for determining the presence or absence of PI animals using data collected from 26 working UK farms involved in a pilot BVDV eradication programme [32] where average herd sizes were 343 (interquartile range: 224 to 431) and 98 (interquartile range: 58 to 121) animals for the dairy and beef units respectively. Methods Farm recruitment began in April 2006 and was largely complete by April 2007. The farms involved and further details of their recruitment are described by Booth & Brownlie 2012 [32]. Farm ID numbers used throughout this manuscript are consistent with those used by Booth & Brownlie 2012 [32] and Booth et al. 2013 [33] enabling cross referencing between publications. All samples were collected by veterinary surgeons as part of the routine infectious disease surveillance on the farms involved and the herd owners gave signed consent for the data collected to be reported anonymously. Upon recruitment to the pilot BVDV eradication programme, each farm was screened to determine herd status (BVDV-infected/BVDV-free) using a combination of BM Ab, BM PCR and YS Check Tests. Throughout this study, Check Tests consisted of 10 animals of approximately 9 months of age (range 6–18 months, with a preference of 9–12 months) from each separate management group on the farm. If fewer than ten animals were available in the age range required for the Check Test then all available animals were sampled. All samples were submitted to and tested by the Animal Health and Plant Agency (AHPA) via AHPA Starcross. BM samples were submitted in universal containers with Bronopol preservative (AHPA, Weybridge) and blood samples were submitted in heparinised vacutainers. AHPA test codes for the BVDV BM Ab ELISA and PCR were TC0123 and TC0709 respectively. All BM Ab results were reported as Optical Density (OD) ratios and interpreted such that OD ratios <0.1 = negative, 0.1–0.35 = low positive, 0.35–0.7 = mid positive and >0.7 = high positive. BM PCR results were reported as either positive or negative. All blood samples tested for Ab (AHPA test code TC0390) were reported as OD ratios and interpreted such that <0.2 = negative and > 0.2 = positive. Further details of the laboratory tests used in this paper including test sensitivity, specificity and further details regarding cut off values have been described previously by Booth and Brownlie 2012 [32]. Since the data presented in this paper was collected the TC0390 and TC0123 ELISA tests have been superseded and it should be noted that the stated cut-off values are not applicable to the ELISA tests currently offered by the AHPA. Following each initial screen, the results were assessed in order determine herd exposure and whether BVDV was likely to be active on the farm and thus whether PI animals were likely to be present. For the purposes of this project, active infection based upon the initial screens was defined cautiously as any herd with >1/10 Ab positive YS and/or a positive BM PCR. Where evidence of exposure to BVDV indicated that the infection was likely to be active, a whole herd test (WHT) was performed where the entire herd was blood sampled in order to identify any PI animals present. Due to the longevity of the immune response following field infection with BVDV [7, 34] high BM Ab levels alone were not deemed sufficient to advise WHTs in the absence of a positive BM PCR or Ab positive YS and farms with this combination of initial results were subjected to further regular surveillance (described below). In beef herds, initial herd screens consisted solely of Check Tests. The laboratory tests and testing regimes used for WHTs have been described previously by Booth and Brownlie 2012 [32]. If identified, PIs were either culled or retained on the farm of origin; this was the farmer’s choice. Where PI animals were identified, the recommendation was to cull them - sale of PI animals (unless direct to slaughter) was not permitted as a condition of membership of this study. In herds where initial screening did not indicate an active BVDV infection, the farmer was given the choice of either blood sampling the whole herd to confirm the accuracy of the screen, or implementing regular herd surveillance. If a WHT was selected, once it was confirmed that there were no PI animals within the herd, the farm implemented regular herd surveillance from that point onwards. Within this study, regular herd surveillance consisted of BM Ab testing (at least quarterly), BM PCR testing (at least yearly) and Check Tests performed at least once a year. In herds undergoing regular surveillance, if active infection (as defined above) was suspected at any point, surveillance frequency was initially increased meaning that a further Check and BM Ab test was performed within 4–8 weeks of the preceding one. In the event that active infection was still suspected, a WHT was performed to search for PIs. Again, in beef herds, surveillance consisted solely of Check Tests. All farms with data reported in this manuscript were monitored for at least 3 years (the majority joined the study in 2006/2007 and remained members until it ended in 2014), thus the data collected accurately reflects the number of PI animals present on each of the study farms over the period presented. On farms that did not undergo WHTs because ongoing herd surveillance indicated that there was no source of BVDV exposure, the assumption was made that no PI animals were present at any time. For herds that underwent WHTs and had full BM Ab and Check Test results from their initial screen, Receiver Operating Characteristic (ROC) curves were generated in SPSS (SPSS Inc., Chicago) to compare the ability of BM Ab and Check Tests to determine the presence of PI animals in a herd. The optimal cut-off value, sensitivity and specificity of each test was then determined. Using logistic regression, predicted probability curves were produced in ‘R’ (The R Foundation for Statistical Computing, 2008) using “xyplot <lattice>” for BM Ab and YS Check Tests to illustrate how the probability of identifying a herd containing a PI(s) alters according to the results of the initial screens. For some herds there were considerable delays between the initial herd screen and the WHT. This consisted of time needed to persuade the farmer to undertake a WHT and agreement on a suitable date (at times, harvesting or silaging meant a considerable delay was incurred). In order to investigate the effect of this delay on the accuracy and relevance of the initial herd screens, BM Ab and BM PCR results from the day of the WHT were collected. In order to simulate a YS Check Test as if performed on the day of the WHT, ten Ab results from YS aged as close to 9–12 months as possible were randomly selected from those sampled during the WHT. The YS chosen were selected using a random number generator in Microsoft Excel (2007) which was used to select ten animals from a list of those present on the day of the WHT arranged in age order starting with the animal closest to 9 months of age on the day of the WHT. Where insufficient animals were present within the 9–12 month range, the range was expanded to 7–13 months. The ROC and probability curve calculations described above were then repeated using data from the day of the WHT in effect removing any delay between conducting a herd screen and acting on the results. Results Twenty-six farms (Table 1) from the original 41 study members described by Booth and Brownlie 2012 [32] had results for both an initial herd screen to determine BVDV status and either whole herd tests (WHT) to identify PI animals or sufficient surveillance (described above) to determine that PI animals were unlikely to be present during the study period. Individual details of the 26 farms analysed in this manuscript are shown in Table 1. The group consisted of three beef herds and 23 dairy herds. Average herd sizes were 343 (interquartile range: 224 to 431) and 98 (interquartile range: 58 to 121) animals for the dairy and beef units respectively. Of the farms involved, the majority, 20/26 (77 %), were vaccinating against BVDV using either Pregsure BVD (Pfizer Animal Health, UK) or Bovilis BVD (MSD Animal Health, Milton Keynes, Buckinghamshire) (Table 1). All vaccines were administered according to the datasheet recommendations from the manufacturer with the primary course completed at an appropriate time prior to first service; a minimum of 14 days and 4 weeks for Pregsure and Bovilis BVD respectively.Table 1 Study Farm Details Farm ID Dairy/Beef Vaccinating over study period? Suspicion of active infection following initial screen?^ Was whole herd testing performed? PIs present at WHT Total number of PIs identified Time between initial herd screen and WHT (Months) Age range of random sample (Months) Herd size at recruitment 1 Dairy Y* Y Y 2 3 11 9–12 349 2 Dairy N Y Y 0 12 7 9–12 542 3 Dairy Y* N Y 0 0 10 9–12 450 5 Dairy Y* Y Y 2 2 7 9–12 527 6 Beef Y* N N - 0 - - 61 7 Dairy Y* Y Y 0 0 9 9–12 381 8 Dairy Y* Y Y 0 0 3 11.5–13 381 11 Dairy Y* N N - 0 - - 217 13 Dairy N N N - 0 - - 201 15 Dairy Y* N Y 1 1 9 7.7–12 633 17 Beef Y* N N - 0 - - 180 18 Dairy Y* Y Y 4 6 9 10–12.6 307 19 Dairy N Y Y 1 1 5 9–12 214 20 Dairy Y* Y Y 3 3 4 9–12 309 21 Beef N N N - 0 - - 54 24 Dairy N N N - 0 - - 286 25 Dairy Y* N Y 0 0 9 9.3–13.5 300 26 Dairy Y* Y Y 2 3 5 8.5–12.3 170 27 Dairy Y* N Y 1 1 11 9–12 192 28 Dairy Y+ Y Y 0 0 12 9–12 335 29 Dairy Y+ Y Y 0 0 12 8–11 412 30 Dairy Y+ N N - 0 - - 360 37 Dairy Y+ Y Y 3 5 2 10.1–12.3 402 38 Dairy Y+ Y Y 2 2 3 8.3–10.9 542 39 Dairy N N Y 0 0 19 9–12 143 40 Dairy Y* Y Y 5 9 1 9–12 230 Mean 8 315 Max 19 633 Min 1 54 Details of the farms analysed in this manuscript: Farm type, vaccination status (* vaccinating using Pregsure BVD and + vaccinating using Bovilis BVD), Suspicion of active infection upon initial screening (^ according to the cautious initial interpretation), Total number of PIs identified during the study, Whether a whole herd test was performed, Total number of PIs identified at the WHT (where performed), The delay between initial screening and WHT if conducted, Age range of animals selected for the random screening at the WHT and herd size at recruitment. Farm numbers can be cross referenced with Booth and Brownlie 2012 [32] and Booth et al. 2013 [33] With the exception of Farms 2, 5 and 15, all farms reported in this manuscript reared homebred youngstock on the farm where they were born.Farms 2, 5 and 15 removed animals at, or shortly after, weaning to separate units where they were reared to return to the home farm for first calving. Farms 5 and 15 only reared their own animals at these separate sites. Farm 2 reared bull calves and non-replacement heifers on the main farm unit as beef animals, however they also utilised a heifer rearer for their replacement dairy heifers. These animals were moved to the heifer rearer after weaning where they were kept with animals from another farm with no attempt to separate them. Once at service age, a bull was introduced and the heifers returned to the main farm to calve. The results of the initial herd screens for all 26 farms are presented in Fig. 1 alongside the number of PIs found in each herd at the WHT. The total number of PIs identified on each farm and the initial number found at the WHT are shown in Table 1. With the exception of Farm 2, in all herds where PIs were identified, at least one of those PIs was present from the outset of the study. For Farm 2, the initial herd screens identified 40 % seroconversion at the YS Check Test (conducted at the heifer rearing unit) and a mid-positive BM Ab titre. The seropositive youngstock initiated a WHT on this farm at both the main farm where the milking animals were kept and the heifer rearing unit. Permission was not given to test the animals at the heifer rearing unit that were mixed with those from Farm 2. No PIs were identified at the Farm 2 WHT in year 1, however, due to the seroconversion noted in the youngstock, newborn animals were tested as they were born onto the farm and the first PI was identified in year 2 of the study as a ‘trojan’ PI born to an non-PI heifer returning from a heifer rearing unit where mixing with animals from multiple sources had occurred [32]. Indeed the subsequent two PIs detected on Farm 2 were also born to non-PI heifers returning from the heifer rearer. For these reasons, within the ROC and probability curve analysis described below, Farm 2 was treated as positive when analysing Check Test results, but negative in the BM analyses. Following the birth of the third PI the adult herd had been infected long enough to produce a fourth PI from a 3rd lactation animal. In total, twelve PIs were confirmed on this farm (with two antigen tests four weeks apart) with a further 9 that either died or were culled before they could be confirmed with a second antigen test [32].Fig. 1 The results of the initial herd screens of the 26 study farms. The figure illustrates the results of the initial herd screens for the 26 farms presented in this manuscript. ^ indicates the 19 herds where WHTs were performed. The left y-axis displays the number of PIs identified at the WHT (where performed) depicted for each farm by the solid bars - the Farm 2 bar is hashed to represent in-utero PI animals. For the 7 farms that did not undertake WHTs because their regular surveillance did not indicate exposure, it was assumed that no PIs were present. The right y-axis displays the BM Ab OD ratio. Along the x -axis, each column represents an individual farm stating both the percentage of antibody positive youngstock (YS) at the Check Test (* all farms except Farms 24, 27 & 28 submitted ten animals who presented 9, 8 & 7 youngstock respectively) and the Farm number. (Graph Key attached to the bottom of Fig. 1) Seven of the 26 study herds (the three beef units; Farms 6, 17 & 21 and four of the dairy units; Farms 11, 13, 24 &30) returned regular surveillance results which indicated that they were unlikely to have been exposed to a source of BVDV infection. These herds were excluded from the statistical analysis because they did not undertake WHTs. The remaining 19 herds undertook WHTs however two of these, Farms 27 and 28, were excluded from the statistical analysis since only 7 & 8 YS respectively were available in the appropriate age range for their initial screens. Therefore, of the 26 study herds, a total of 17 were suitable for ROC and probability curve analysis since they had both 10 YS and a BM Ab sampled at initial recruitment and also conducted a WHT. Furthermore, it is only these 17 farms that contribute to the sensitivity and specificity results stated below. Bulk milk antibody analysis BM Ab titres on the 23 dairy farms observed in Fig. 1 range from mid to high positive. Interestingly, high positive BM Ab results occur on six farms where no PIs were identified (Farms 7, 8, 25, 28, 29 & 30) and mid positive results occur on two farms where PIs were identified (Farms 1 and 2). On Farm 2 however, it should be noted that the dairy herd would not have had contact with PI animals by this point. Of the dairy farms observed in Fig. 1, none began the study with low positive or negative BM Ab titres, yet 11/23 (48 %) had no PI animals identified. The BM Ab ROC curve in Fig. 2 and its coordinates (Table 2) indicate that the optimal cut-off point for distinguishing whether or not a herd contained a PI animal using the BM Ab titre is achieved at an OD ratio of 0.7950 units with a sensitivity and specificity of 80.00 % (44.39–97.48 %) and 85.71 % (42.13–99.64 %) respectively (95 % confidence intervals quoted). Of the 17 farms analysed in Fig. 2, using a BM OD ratio cut off of 0.7950 there were two false negative farms (i.e. would be declared free of PIs when they were present) (Farms 1 & 20) and one false positive (i.e. would be thought to contain PI animals when there were none present) (Farm 7). For Farm 2, the BM Ab OD ratio of 0.46 units and interpretation that there were not PI animals present can be considered correct for the adult dairy herd given that it had not been exposed to PI animals at this point. Fig. 3a displays the predicted probability curve for BM Ab OD ratio calculated for the 17 farms that were used for the ROC curve analysis. The curve shows a gradual but limited increase in probability of detecting a PI as BM Ab OD ratio increases.Fig. 2 ROC curves demonstrating the performance of BM Ab and YS Check Tests as predictors of the presence of a PI animal(s) where delays occurred between the initial herd screen and the WHT. The ROC curves were generated from the seventeen dairy farms in Fig. 1 with full initial herd screens and WHT results. YS = Youngstock check testing and BM Ab = Bulk milk antibody testing. Farm 2 is considered negative for BM Ab analysis and positive for YS analysis. AUC = Area under curve. Coordinates of the curves are given in Table 2 Table 2 Coordinates of the ROC curves in Fig. 2 Test result variables Positive if greater than or equal to a or b# Sensitivity 1 - Specificity aNumber of Youngstock antibody positive out of ten tested with delays between the YS screen and whole herd test −1.00 1.000 1.000 1.00 .909 .500 3.00 .818 .333 4.50 .636 .333 5.50 .636 .167 6.50 .545 0.000 7.50 .455 0.000 8.50 .273 0.000 9.50 .091 0.000 11.00 0.000 0.000 bBulk milk antibody OD ratio with delays between the BM Ab screen and whole herd test −.5630 1.000 1.000 .4440 1.000 .857 .4575 1.000 .714 .5225 1.000 .571 .6520 .900 .571 .7315 .900 .429 .7590 .800 .429 .7840 .800 .286 .7950 .800 .143 .8025 .700 .143 .8075 .600 .143 .8285 .500 .143 .8535 .400 .143 .8650 .300 .143 .8850 .200 .143 .9600 .100 .143 1.1885 0.000 .143 2.3570 0.000 0.000 # The smallest cutoff value is the minimum observed test value minus 1, and the largest cutoff value is the maximum observed test value plus 1. All the other cutoff values are the averages of two consecutive ordered observed test values Fig. 3 Predicted Probability Curves indicating the probability of a study herd containing a PI animal with the stated BM Ab and YS Check Test results. Data points demonstrating PI presence are ‘jittered’ around 0 or 1 on the y axis to show the number of farms that contribute to the curve at each point, a BM Ab OD Ratio and the probability of identifying a herd containing a PI animal with delays between performing the BM test and the WHT; b Number of YS positive out of the ten tested and the probability of identifying a herd containing a PI animal with delays between performing the Check Test and the WHT; c BM Ab OD Ratio and the probability of identifying a herd containing a PI animal with no delay between performing the BM test and the WHT; d Number of YS positive out of the ten tested and the probability of identifying a herd containing a PI animal with no delays between performing the Check Test and the WHT. Farm 2 is considered negative for BM Ab analysis and positive for YS Check Test analysis Check test analysis In Fig. 1, antibody positive YS identified at the initial Check Test are presented as a percentage of the total tested since Farms 24, 27 and 28 only presented 9, 8 and 7 YS respectively in the required age range. All other farms had ten YS available for the Check Test. When observing the number of antibody positive YS identified on initial herd screening in Fig. 1, the number of antibody positive YS increases with the likelihood of identifying a herd containing a PI animal. Using the Check Test results alone and the original cut-off (defined above) of >1/10 Ab positive animals signifying a BVDV infected farm results in two false negative farms (Farms 15 & 27) and four false positives (Farms 7, 8, 28 & 29). Analysing the Check Test results from the seventeen dairy herds that undertook WHTs, with Farm 2 included as infected since three heifers amongst the group sampled were carrying in-utero PIs at this point, the coordinates (Table 2) of the ROC curve in Fig. 2, indicate two potential cut off points. The first uses a cut off of 5.5/10 YS BVDV Ab positive (thus interpreting results with either ≤5 YS or ≥6 YS as herds that are negative and positive respectively) and returns a sensitivity and specificity of 63.64 % (30.79–89.07 %) and 83.33 % (35.88–99.58 %) respectively (95 % confidence intervals quoted). The second cut off is at 3/10 YS BVDV Ab positive and returns a sensitivity and specificity of 81.82 % (48.22–97.72 %) and 66.67 % (22.28–95.67 %) respectively (95 % confidence intervals quoted). The second cut off of 3/10 YS BVDV Ab positive may be preferred marginally due to the slightly higher sensitivity and thus smaller likelihood of mis-diagnosing a farm that contains a PI as negative. However, neither cut-off can be considered ideal. The predicted probability curve in Fig. 3b is generated for the same seventeen herds that underwent ROC curve analysis and shows how the probability of detecting a herd containing a PI increases as the number of Ab positive YS out of ten tested increases. Bulk milk PCR analysis Not all dairy farms had BM PCR results at this stage of the study - this was simply due to the fact that the test was not offered by the AHPA at the start of farm recruitment. With the exception of Farm 29, all farms returning a positive BM PCR result in Fig. 1 had PI animals identified; Farms 38 & 40, where Farm 38 had a milking PI animal yet Farm 40 did not. There were also six farms which were BM PCR negative and of these, three (Farms 18, 37 & 20) contained PIs none of which were of milking age. The ages of the PI animals identified on these farms (and throughout this study) are discussed in Booth and Brownlie 2012 [32]. At this point, Farm 29 is worth highlighting since it returned a positive BM PCR result, a high positive BM Ab result and 5/10 animals Ab positive at the Check Test yet no PIs were identified on the premises either at the WHT or in the testing following this. The delay between the initial herd screens and WHTs to identify PI animals is shown in Table 1 and ranges from 1–19 months a mean interval between initial screens and WHTs of 8 months. The delay for Farm 29 was 12 months between the initial screens presented in Fig. 1 and the WHT to identify PI animals. The impact of these delays between initial screens and what was actually found at the WHT could be considerable and therefore the initial screens presented in Fig. 1 may not accurately reflect herd status at the time the WHT was conducted. For this reason, further data was collected from the farms presented in Fig. 1 so that for those farms that underwent a WHT, the equivalent of an initial screen was generated for that day meaning that the delay between the initial screen and the WHT was effectively 0 days. The age groups of the randomly selected YS are detailed in Table 1 and for most farms, it was possible to select animals from within the 9–12 month range. The YS selected for Farms 18 and 37 were just outside of the 9–12 month range whilst for Farms 8, 15, 25, 26 and 38 it was necessary to extend the range to 7–13 months. These data are presented in Fig. 4. Farms 6, 11, 13, 17, 21, 24, & 30 did not undergo WHT since the results of their regular surveillance (BM Ab, BM PCR and Check Tests) did not justify testing all stock. These farms are included in Fig. 4 and the data presented for each are the year 2 Check Test results which coincided with a quarterly BM Ab and yearly PCR test (where available).Fig. 4 The results of simulated initial herd screens of the 19 herds that undertook WHTs and the Year 2 screens of the remaining 7 herds. The figure illustrates the results of the simulated initial herd screens for the 19 herds that undertook WHTs and the Year 2 screens for the 7 herds that did not. ^indicates that the farm underwent whole herd testing. The left y-axis displays the number of PIs identified at the WHT (where performed) depicted for each farm by the solid bars – the Farm 2 bar is hashed to represent in-utero PI animals. For the 7 farms that did not undertake WHTs because their regular surveillance did not indicate exposure, it was assumed that no PIs were present. The right y-axis displays the BM Ab OD ratio. Along the x -axis, each column represents an individual farm stating both the percentage of antibody positive youngstock (YS) at the Check Test (* all farms submitted ten animals except Farm 30 where only 9 animals of appropriate age were available) and the Farm number. (Graph Key attached to the bottom of Fig. 4) Bulk milk antibody analysis without delay In the 23 dairy farms represented in Fig. 4 high positive BM Ab results occur on nine infected farms and two that are BVDV-free (Farms 29 & 30), mid positive BM Ab results occur on six BVDV-free farms, Farm 2 and two infected farms (Farms 15 & 19). Finally, three BVDV-free dairy farms in Fig. 4 now return low positive BM Ab results. On Farm 2, the dairy herd was still un-exposed at this point hence should be considered a negative herd when assessing BM Ab. Analysis of BM Ab results from the 17 dairy herds that undertook WHTs produced the ROC curve in Fig. 5, the coordinates of which (Table 3) demonstrate that the optimal BM Ab OD ratio cut-off for distinguishing whether a herd contained a PI animal or not occurs at 0.7 OD ratio units. This also coincides with the AHPA cut off for a high positive result using this test. With this dataset, a sensitivity and specificity of 80.00 % (44.39–97.48 %) and 85.71 % (42.13–99.64 %) respectively (95 % confidence intervals quoted) was achieved. The predicted probability curve in Fig. 3c shows better distinction using BM Ab OD ratio at this cut off compared to Fig. 3a.Fig. 5 ROC curves demonstrating the performance of BM Ab and YS Check Tests as predictors of the presence of a PI animal(s) where there is no delay between the initial herd screen and the WHT. The ROC curves were generated from the seventeen dairy farms in Fig. 4 with full initial herd screens and WHT results. YS = Youngstock check testing and BM Ab = Bulk milk antibody testing. Farm 2 is considered negative for BM Ab analysis and positive for YS analysis. AUC = Area under curve. Coordinates of the curves are given in Table 3 Table 3 Coordinates of the ROC curves in Fig. 5 Test result variables Positive if greater than or equal to a or b# Sensitivity 1 - Specificity aNumber of youngstock antibody positive out of ten tested without delay −1.00 1.000 1.000 .50 .909 .500 1.50 .909 .167 2.50 .909 0.000 3.50 .818 0.000 5.00 .727 0.000 6.50 .636 0.000 7.50 .545 0.000 9.00 .273 0.000 11.00 0.000 0.000 bBulk milk antibody OD ratio without delay −.8500 1.000 1.000 .2400 1.000 .857 .3750 1.000 .714 .4300 1.000 .571 .4550 1.000 .429 .4950 1.000 .286 .5300 .900 .286 .6050 .800 .286 .7000 .800 .143 .7650 .700 .143 .8250 .600 .143 .8550 .500 .143 .8750 .400 .143 .8950 .300 .143 .9100 .200 .143 .9250 .200 0.000 .9800 .100 0.000 2.0300 0.000 0.000 # The smallest cutoff value is the minimum observed test value minus 1, and the largest cutoff value is the maximum observed test value plus 1. All the other cutoff values are the averages of two consecutive ordered observed test values Check test analysis without delay Analysing the farms in Fig. 4 using the cautious cut off (defined earlier) of >1/10 Ab positive youngstock results in the false positive misclassification of Farms 13 & 29 and the false negative misclassification of Farm 15. Analysing the Check Test results from the 17 farms produces the ROC curve in Fig. 5, the coordinates of which (Table 3) indicate that an optimal cut-off point occurs at 2.5/10 animals with a sensitivity and specificity of 90.9 % (58.72–99.77 %) and 100 % (54.07–100 %) respectively (95 % confidence intervals quoted). Utilising this cut-off would result in the false negative misclassification of one farm (Farm 15) from the 17 analysed in Fig. 5. The predicted probability curve in Fig. 3d illustrates an improved distinction compared to the tests explored in Fig. 3a-c. Bulk milk PCR analysis without delay At this point all farms that returned a positive BM PCR in Fig. 4 had PIs identified (Farms 1, 19 & 38) where Farms 1 & 38 contained milking PIs, but Farm 19 did not. Five out of the twelve farms that returned BM PCR negative results also contained PI animals at the time the BM was sampled (Farms 20, 26, 27, 37 & 40) however, none of the PI animals on these farms were of milking age. These results indicate that when herd screening is performed with no delay between that and the WHT to identify PI animals, the sensitivity and specificity of both the BM antibody and the YS cohort screens for determining PI presence in a herd are improved. Furthermore, the predictive probabilities of both tests improve when the delay between screening and WHTs is removed. Discussion This manuscript presents BM Ab, BM PCR and YS Ab (Check Test) data from twenty six working farms that were recruited to a pilot BVDV eradication programme. Although there is general agreement in the current literature that BM Ab levels and youngstock Check Tests are appropriate ways to assess BVDV status at the herd level [20, 25–27, 35, 36], there remains some confusion about their reliability and the impact of historic infection and vaccination. We discuss these issues in the context of our study herds below. The recruited herds were predominantly located in Somerset, UK and the large proportion of dairy herds in the study population reflects the region from which herds were sampled. The average study herd size appears large when compared to the average herd size of 120 animals quoted by Defra for the Taunton area at the time of recruitment [37]. However, the Defra figure may be falsely low as this will include small holdings and in the authors’ experience, the herd sizes represented here are typical of the range of farms in this region. Within the analyses conducted, the wide 95 % confidence intervals quoted throughout highlight that there is a degree of uncertainty surrounding the results and thus the figures should be used with caution. This is largely due to the small sample size of 17 herds which underwent statistical analysis. Despite these potential limitations, the improvements demonstrated in sensitivity and specificity and also in the ROC and probability curves by removing the delay between screening a herd and acting on those results are substantial and worth discussion. Of the 26 herds that were recruited, only 17 were suitable for statistical analysis since BM and YS samples were collected in addition to individual animal samples to determine the BVDV status of all animals on this subset of farms. Two farms that had conducted a WHT were deemed unsuitable for inclusion in the statistical analysis as ten animals were not available for the initial YS screens and are included in Figs. 1 and 4. The seven remaining farms did not undergo WHTs and were therefore excluded from the statistical analysis, but are presented for completeness in Figs. 1 and 4. Whilst we believe that the regular surveillance would have detected infection on these farms over, and beyond, the course of the study period analysed here, it remained an assumption that they were BVDV free and so it would have been inappropriate to influence the ROC, probability, sensitivity and specificity analyses on this basis. This point is especially pertinent when we consider that Farm 15 returned a mid positive BM Ab titre at the WHT and also negative YS results in both the initial Check Test and the simulated Check Test from the WHT yet a PI was present throughout the period. This was primarily due to herd dynamics and is discussed in relation to this herd’s results below. The Use of BM Ab Tests for determining Herd BVDV Status BM Ab tests provide an initial, rapid and cost effective assessment of the BVDV status in dairy cattle at the herd level. It is largely accepted that high BM Ab levels correlate with a high probability of the presence of a PI. However, we found several instances where herd BVDV status would have been incorrectly classified if based on an interpretation of the BM Ab result alone. The sensitivity and specificity of the test throughout this study was 80.00 and 85.71 % respectively however, the probability curves indicate that it is difficult to accurately define a cut-off point where we can say with reasonable certainty that a herd does or does not contain a PI animal. The removal of the delay between the BM Ab screening and further investigations does improve the confidence that one can have in the interpretation of the test but as a sole screening method, the usefulness of BM Ab remains limited in many circumstances. We know that both historic BVDV infections and herd vaccination can have a significant effect on correct BM Ab interpretation [33, 38] and both of these factors are likely to have played a significant role in the test results and interpretation here. BVDV antibodies following a natural infection may persist for 3 or more years in an individual animal [7, 34]. When one considers that BM Ab assesses all milking animals including the oldest in the herd, it explains why it can take in excess of 1000 days for BM Ab to show any real decrease following the removal of PI animals [38]. Booth et al. 2013 [33] demonstrated that Farms 13 & 24, which were BVDV-free farms that neither vaccinated nor experienced active BVDV infection during the study period experienced a slow yet consistent decrease in the BM Ab titre; in this type of herd BM Ab can be a sensitive surveillance test for monitoring for disease incursion into the milking group since an increase in Ab titre can be considered of significance. Vaccination can also confound interpretation of BM Ab results further. With the exception of Farms 13, 24 & 39, all BVDV-free dairy herds in this study vaccinated during the period data was collected (Table 1) and all began the study with mid to high BM Ab titres (Fig. 1). In the longitudinal analysis of BM Ab trends by Booth et al. 2013 [33], BVDV-free vaccinating herds did not demonstrate a clear decrease in BM Ab. The fact that the majority of the study herds were utilising BVDV vaccines, in combination with the likelihood that many of them may have experienced historic infection and thus contain older seropositive animals may explain the higher level of misclassification of false positive farms observed with the BM Ab tests compared to YS Check Tests. In practice, careful thought needs to be given to the application and timing of BM Ab testing in herds utilising BVDV vaccines and the use of this test may be deemed inappropriate in some systems. This is discussed in further detail by Booth and et al. 2013 [33]. Whilst vaccination and historic infection would explain those farms that returned high positive BM Ab results in the absence of PIs, it is more difficult to explain those herds (Farms 15 and 19) that returned only mid-positive BM Ab results in Fig. 4 when PI animals were present within the herd. For these farms, the reasons are likely due to the location of the PI animal(s) within their herds:For Farm 15, the lack of a high positive BM Ab result is harder to explain since the PI was of milking age. She was however a poor animal and often separated from the milking herd in a hospital pen due to illness. Farm 19 contained a PI animal of approximately 1 year of age and yet returned a mid positive BM Ab result. Upon initial interpretation, it may simply be possible that the PI did not have contact with the milking herd, however, analysis of the BM PCR results bring this interpretation into question since the herd was positive on BM PCR testing. It is difficult to provide an explanation for this combination of results especially given that at the initial screen for Farm 19 (conducted 5 months prior to the WHT), the herd returned a high positive BM Ab result (Fig. 1). Farm 2 highlights the need for the practitioner to consider the epidemiological units that make up the herd structure in detail both before testing and before coming to a decision on the status of the herd based on the results of the tests. At both the initial herd screen and at the point of the WHT, the milking herd on Farm 2 had not been recently exposed to BVDV, hence the mid positive result. If a single assessment had been performed on a bulk milk sample, and decision reached that the herd was uninfected, then infection in the heifers would have gone unnoticed until subsequent BM Ab titres had increased. Our Check Test results indicated that there was likely exposure in the heifer group at the calf rearer and that this was almost certainly due to the biosecurity breakdown involving mixing with animals from another farm. The return of three heifers carrying PI foetuses to the main farm in combination with the farmers reluctance to cull PI animals once identified and the failure to engage the owner of the other animals at the heifer rearer in the scheme eventually resulted in infection of the main herd and an increase in the BM Ab titre to a high positive (data not shown). This farm is discussed in more detail by Booth and Brownlie 2012 [32]. The use of young stock (YS) check tests for determining herd BVDV status The real value of YS antibody tests is that young animals will be seronegative if unexposed and maternally derived Ab (MDA) have waned; if they do have antibodies demonstrating an active immune response (in contrast to MDA), they will have been infected only within their short lifetime (e.g. 9 months). That would indicate a recent, or present, infection on the farm. Assessing animals at, or after, 9 months of age is a valuable time to initiate a YS screen as MDA will have has waned and it is typically pre-vaccination. YS below six months of age should not be selected for Check Testing since MDA are highly likely to be present. Chamorro et al. 2014 [39] demonstrated that in extreme cases, BVDV MDA can take 11 months to wane and that MDA are commonly present up to 6 months of age; hence have the potential to confound interpretation of the test. In the authors’ experience, testing animals at least 9 months of age, especially in beef suckler herds where colostral transfer of MDA is more reliable and higher quality than in dairy systems, provides a safer margin for ensuring that MDA is not present. Whilst the majority of the herds in this study did vaccinate, they only initiated the primary course of vaccine prior to first service. The result was that we were able to assess the YS cohorts in these herds without the influence of vaccine. There is the possibility on Farms 8 and 25, that some of the older YS selected from the WHT in order to simulate a Check Test on that day could have been vaccinated as they were at the older end of the 7–13 month range, but in all the other herds, this was unlikely to have influenced the results. In this study, we achieved both a high sensitivity and specificity when assessing the ability of the Check Test to determine whether PI animals were present in the seventeen study herds analysed when the Check Test was simulated from the WHT results. Furthermore, the probability curves displayed an increasing ability to predict the presence of a PI animal within the study herds at the suggested cut off of 2.5/10 animals Ab positive once the delay between screening and the WHT had been reduced. Whilst we have sampled randomly within the animals tested at the WHT in order to generate the ‘without delay’ Check Test, it is important to mention that in the field, careful thought should be given to the contact of the cohort(s) tested with the rest of the herd and this is discussed in the context of Farm 15 below. Previous epidemiological studies have shown that when two or fewer YS of the ten tested are Ab positive, there are unlikely to be PIs present on a farm [27–29] and the optimal cut off from this manuscript of 2.5/10 YS Ab positive supports this finding, even in larger UK herds. The authors would urge readers to use some caution when applying this cut-off value in those herds that return 1/10 or 2/10 positive animals since this may be the start of seroconversion in a group, detection of a declining Ab response to vaccine or MDA. The most effective way to distinguish this is to re-test the same animals at least 28 days later to determine whether seroconversion is underway. Farm 15 was consistently classified as negative when interpreting YS Check Tests when, in fact, it contained one adult PI animal. The explanation is that the PI animal was bought in as a heifer and, after taking considerable time to get in calf, finally produced a dead calf that was quickly removed. All calves born on Farm 15 were routinely raised on a different unit to adult stock and thus the adult PI animal never had contact with young animals and did not produce a live PI calf that was mixed with the YS. This again highlights the importance of considering the mixing of groups within a farm unit. If the cohorts selected for Check Tests have had reasonable exposure to the rest of the herd, the results may be extrapolated to provide an indication of herd status. However, cohorts of animals reared as isolated groups (or even in some cases as separate herds until first calving) are not suitable for selection for Check Tests to determine whole herd status but only group status. Furthermore, Farm 15 highlights the importance of testing purchased stock however, this animal was purchased before Farm 15 was recruited and was therefore only detected once the study commenced. In a herd with a high positive BM Ab titre due to historic infection and/or vaccination, sampling cohorts of YS may be the only way to accurately obtain an up to date assessment of herd status and the predicted probability and ROC curves in Figs. 3d and 5 respectively support that this should provide a relatively accurate analysis. In beef herds, monitoring antibody levels in YS cohorts is the only method of assessing the herd hence it is re-assuring that this test performs well. In traditional cow:calf suckler units, Check Tests are potentially much more reliable than in dairy units since there is a greater degree of mixing of animals thus separate epidemiological groups are less likely to occur. The use of bulk milk PCR for determining herd BVDV status BM PCR provided a highly sensitive and cost-effective test for adult PI animals contributing to the milk tank. Unfortunately, our PCR data is incomplete as the test was not widely available at the time of recruitment of many farms. Over the course of the study, our limited BM PCR results did identify 5 farms with a positive test; on two of these farms (Farms 1 & 38), PIs were found within the milking herd. On Farm 40, no PIs of milking age were identified and only two older animals had left the milking herd in the time between screening and the WHT. Whilst it is possible that either/both of these culled animals had been PI, another potential hypothesis for the positive BM PCR could be that a number of pregnant animals were carrying PIs and thus the level of virus circulating in the milking herd was sufficient to create detectable levels of viral RNA in the bulk milk. On Farm 19 a positive BM PCR was also returned at the whole herd test, but no milking PI animals were identified, nor were further PIs beyond the initial one identified in the youngstock so it would seem unlikely that this could be due to pregnant animals carrying PIs in this instance and may simply be due to contact between the milking animals and YS group containing the PI on this unit. This concept would be supported if the matching milk sample had returned a high positive Ab titre, but as described above, it is difficult to reconcile why this was not the case and a mid positive titre was returned. In contrast, Farm 29 returned a positive BM PCR result yet no PIs were identified at the WHT. On this farm it is highly likely that the delay in conducting a WHT (12 months) resulted in the death or removal of any PIs prior to identification. The fact that 5/10 YS on Farm 29 tested Ab positive at the initial screen and that the herd had a high positive BM Ab titre support the conclusion that PIs had been present on the farm. It is important to note that BM PCR can only test those animals contributing to the bulk tank on the day the sample was taken. Our results show that some of the herds that did contain PIs did not return a BM PCR positive test result. PIs later identified on these farms had not contributed to the bulk tank. For this reason, provided the upper limit of milking contributors to a sample, as recommended by the testing laboratory, is not exceeded, a negative BM PCR is a reliable indication that any animals contributing to the BM tank, are not PI. A negative result is not, however, a reliable indication that there is not a PI(s) animal elsewhere in the herd. The effect of delays between herd screens and whole herd testing Apart from the 19 month delay between screening and WHTs on Farm 39, most herds were investigated within 8 months of their initial screen. Farm 39 consistently appeared BVDV-free in the regular surveillance conducted and so there was no urgency to perform a WHT in this herd so perhaps it skews the recorded delays unnecessarily. The mean interval is not ideal but in part reflects the reality in practice where results are awaited, reported to the farmer and then a period of discussion and organisation, often avoiding harvesting or silaging, has to occur in order to arrange whole herd testing. If the delay between screening the herd and taking action based on those results is prolonged, the results of the initial screens may not accurately reflect the BVDV status of the herd in question. The data presented here reflects this and in all cases shows an increase in the diagnostic accuracy and relevance of the screens once these delays are removed. Farms 7, 8, and 29 had delays between initial screens and WHT of 9, 3 and 12 months respectively. Based upon the original YS criteria of >1/10 YS Ab positive indicating an infected herd all were classified as infected. Despite this, no PIs were identified on these three farms during the course of the study (Fig. 1). Farm 8 had a relatively low number of YS seropositive at the initial screen (2/10) which declined further to 1/10 YS Ab positive at the WHT three months later and so is unlikely to have been an infected herd. Whilst no PIs were found, it is highly likely that both farms 29 & 7 were infected due to the number of seropositive YS detected at the initial Check Test; 6/10 and 5/10 YS Ab positive for each farm respectively (Fig. 1). An infected status is further supported for Farm 29 since it returned a positive BM PCR result at the initial screen (Fig. 1). Due to the delays in conducting WHTs on these two farms, it is possible that any PIs had unknowingly been removed from these units prior to undertaking WHTs since both farms show less evidence of BVDV exposure in Fig. 4. Figure 4 demonstrates that, based on Check Tests, herd screens performed closer to the WHT would have indicated that herds 7 & 29 did not contain PI animals at the point of the WHT since their results were 1/10 and 2/10 YS Ab positive respectively (Fig. 4) – this is below the cut off of 2.5/10 animals indicated in the ROC curve in Fig. 5. The result for Farm 29 is further supported by the fact that it was also BM PCR negative at this stage. These results indicate that significant alterations in herd infection status may have occurred on Farms 7 and 29 in the time between the initial screen and WHT. On Farms 15 & 27 the delay between the initial screens and whole herd tests was 8 & 9 months respectively. Both farms 15 & 27 bought PI animals shortly before the screens presented in Fig. 1 were performed (data not shown). It is evident that for Farm 27, the initial Check Test was likely performed before enough time had elapsed for seroconversion to have occurred within the group screened since continued surveillance on Farm 27 would have detected the presence of the PI animal, since the YS seroprevalence increased from 0/10 YS Ab positive in Fig. 1 to 10/10 YS Ab positive (Fig. 4). This further highlights the need to develop an ongoing surveillance plan. Farm 15 would however remain undetected through YS surveillance and the reasons for this have been discussed above. Conclusions The data presented indicate that the approach to establishing herd status endorsed by CHeCS [25] is robust. Check Tests, BM Ab and BM PCR, if used and interpreted appropriately are useful tools for establishing a herds BVD status. Minimising the delay between screening a herd and acting quickly on the results obtained will increase their diagnostic accuracy and relevance. Whilst a good starting point, a high positive BM Ab result alone is not enough to classify a herd as infected but it should trigger further investigation. In the same way a mid positive BM Ab result does not definitively mark a herd as BVDV-free. BM Ab can prove a useful surveillance tool in established BVDV free herds with low or negative BM Ab titres however; such results were rarely encountered in this study. For this reason, if the study population is representative of herds in cattle dense areas where there is no systematic control of BVDV and biosecurity is difficult to implement, then low or negative BM Ab samples could be considered rare. In this paper, Check Tests performed with the greatest accuracy when diagnosing herd BVD status; only one farm was misclassified when using YS screens (Fig. 4). BM PCR may be used at the same point of the initial screen of YS and BM antibody although, on cost grounds, it would be reasonable to postpone this test until Check Test and BM Ab results indicate the likely presence of a PI. Finally, it is paramount that longitudinal surveillance using a combination of the techniques discussed in this paper is performed on farms wishing to monitor their BVDV status since this allows for changes in status to be detected early thus enabling prompt action in the event of a BVDV incursion. Abbreviations AbAntibody AHPAAnimal Health and Plant Agency BMBulk milk BVDVBovine Viral Diarrhoea Virus CHeCSCattle Health Certification Standards PCRPolymerase chain reaction PIPersistently infected ROCReceiver operating characteristic WHTWhole herd test YSYoungstock Acknowledgements This study was funded by Defra under project grant SE0777. The authors would like to thank Mick Cranwell (AHPA, Starcross) for his help and support with the laboratory testing, Paddy Gordon for his help during the study design and setup, and the practitioners and staff at Shepton Veterinary Group and Synergy Farm Health for their extensive involvement in this work. Funding This study was funded by Defra under project grant SE0777. Availability of data and materials All datasets supporting the analysis of this study are included in the published article. Should further information be required, this will be available upon request from the corresponding author. Authors’ contributions RB and JB conducted the sample collection from all herds involved. RB performed the analysis of the datasets. RB and JB contributed jointly to the drafting of the manuscript. All authors read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. Consent to publish Not applicable. Ethics approval and consent to participate All samples were collected by veterinary surgeons as part of the routine infectious disease surveillance on the farms involved and the herd owners gave signed consent for the data collected to be reported anonymously. ==== Refs References 1. Potgieter LN Immunology of bovine viral diarrhea virus Vet Clin North Am Food Anim Pract 1995 11 501 20 10.1016/S0749-0720(15)30464-3 8581860 2. Virakul P Fahning ML Joo HS Zemjanis R Fertility of cows challenged with a cytopathic strain of Bovine Viral Diarrhea virus during an outbreak of spontaneous infection with a noncytopathic strain Theriogenology 1988 29 441 9 10.1016/0093-691X(88)90246-4 16726367 3. Edwards S Wood L Hewitt-Taylor C Drew TW Evidence for an immunocompromising effect of bovine pestivirus on bovid herpesvirus 1 vaccination Vet Res Commun 1986 10 297 302 10.1007/BF02213992 3016976 4. Wray C Roeder PL Effect of bovine virus diarrhoea-mucosal disease virus infection on salmonella infection in calves Res Vet Sci 1987 42 213 8 3035663 5. Ellis JA Davis WC Belden EL Pratt DL Flow cytofluorimetric analysis of lymphocyte subset alterations in cattle infected with bovine viral diarrhea virus Vet Pathol 1988 25 231 6 10.1177/030098588802500308 3293305 6. McGowan MR Kirkland PD Richards SG Littlejohns IR Increased reproductive losses in cattle infected with bovine pestivirus around the time of insemination Vet Rec 1993 133 39 43 10.1136/vr.133.2.39 8212473 7. Moerman A Straver PJ de Jong MC Quak J Baanvinger T van Oirschot JT A long term epidemiological study of bovine viral diarrhoea infections in a large herd of dairy cattle Vet Rec 1993 132 622 6 10.1136/vr.132.25.622 8394609 8. Bennett R Ijpelaar J Economic Assessment of Livestock Diseases in Great Britain. Final Report to Defra Book Economic Assessment of Livestock Diseases in Great Britain. Final Report to Defra 2003 9. Stott AW Humphry RW Gunn GJ Higgins I Hennessy T O'Flaherty J Graham DA Predicted costs and benefits of eradicating BVDV from Ireland Irish Vet J 2012 65 12 10.1186/2046-0481-65-12 10. Gunn GJ Stott AW Humphry RW Modelling and costing BVD outbreaks in beef herds Vet J 2004 167 143 9 10.1016/S1090-0233(03)00112-6 14975388 11. Lindberg A Brownlie J Gunn GJ Houe H Moennig V Saatkamp HW Sandvik T Valle PS The control of bovine viral diarrhoea virus in Europe: today and in the future Rev Sci Tech 2006 25 961 79 17361763 12. Uttenthal A Stadejek T Nylin B Genetic diversity of bovine viral diarrhoea viruses (BVDV) in Denmark during a 10-year eradication period APMIS 2005 113 536 41 10.1111/j.1600-0463.2005.apm_227.x 16086824 13. The Scottish Government. The Scottish BVD Eradication Scheme [http://www.gov.scot/Topics/farmingrural/Agriculture/animal-welfare/Diseases/disease/bvd/eradication]. Accessed 8 Dec 2016. 14. Animal Health Ireland. BVD - National Eradication Programme [http://animalhealthireland.ie/?page_id=220]. Accessed 8 Dec 2016. 15. Sarrazin S Epidemiological reflections on bovine viral diarrhoea virus control in Belgian cattle herds based on experimental infections and observational studies 2015 Ghent Ghent University, Department of Reroduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine 16. Brownlie J Clarke MC Howard CJ Experimental infection of cattle in early pregnancy with a cytopathic strain of bovine virus diarrhoea virus Res Vet Sci 1989 46 307 11 2544969 17. Peterhans E Jungi TW Schweizer M BVDV and innate immunity Biologicals 2003 31 107 12 10.1016/S1045-1056(03)00024-1 12770540 18. Mars MH Bruschke CJ van Oirschot JT Airborne transmission of BHV1, BRSV, and BVDV among cattle is possible under experimental conditions Vet Microbiol 1999 66 197 207 10.1016/S0378-1135(99)00009-7 10227122 19. Barlow RM Nettleton PF Gardiner AC Greig A Campbell JR Bonn JM Persistent bovine virus diarrhoea virus infection in a bull Vet Rec 1986 118 321 4 10.1136/vr.118.12.321 3010538 20. Lindberg AL Alenius S Principles for eradication of bovine viral diarrhoea virus (BVDV) infections in cattle populations Vet Microbiol 1999 64 197 222 10.1016/S0378-1135(98)00270-3 10028173 21. Presi P Struchen R Knight-Jones T Scholl S Heim D Bovine viral diarrhea (BVD) eradication in Switzerland--experiences of the first two years Prev Vet Med 2011 99 112 21 10.1016/j.prevetmed.2011.01.012 21371766 22. Moennig V Houe H Lindberg A BVD control in Europe: current status and perspectives Anim Health Res Rev 2005 6 63 74 10.1079/AHR2005102 16164009 23. Niskanen R Lindberg A Transmission of bovine viral diarrhoea virus by unhygienic vaccination procedures, ambient air, and from contaminated pens Vet J 2003 165 125 30 10.1016/S1090-0233(02)00161-2 12573600 24. Houe H Lindberg A Moennig V Test strategies in bovine viral diarrhea virus control and eradication campaigns in Europe J Vet Diagn Invest 2006 18 427 36 10.1177/104063870601800501 17037609 25. CHeCS. Cattle Health Certification Standards Technical Document [http://www.checs.co.uk/wp-content/uploads/2015/08/CHeCS-Technical-Document-final-2015.pdf]. Accessed 8 Dec 2016. 26. Pritchard G Milk antibody testing in cattle Practice 2001 27. Houe H Serological analysis of a small herd sample to predict presence or absence of animals persistently infected with bovine viral diarrhoea virus (BVDV) in dairy herds Res Vet Sci 1992 53 320 3 10.1016/0034-5288(92)90133-M 1334566 28. Houe H Bovine virus diarrhoea virus: detection of Danish dairy herds with persistently infected animals by means of a screening test of ten young stock Prev Vet Med 1994 19 241 8 10.1016/0167-5877(94)90092-2 29. Houe H Baker JC Maes RK Ruegg PL Lloyd JW Application of antibody titers against bovine viral diarrhea virus (BVDV) as a measure to detect herds with cattle persistently infected with BVDV J Vet Diagn Invest 1995 7 327 32 10.1177/104063879500700305 7578446 30. Drew TW Yapp F Paton DJ The detection of bovine viral diarrhoea virus in bulk milk samples by the use of a single-tube RT-PCR Vet Microbiol 1999 64 145 54 10.1016/S0378-1135(98)00266-1 10028169 31. Animal Health and Plant Agency: Disease Surveillance Tests - Bovine viral diarrhoea virus in bulk milk samples (up to 300 contributors) [http://science.vla.gov.uk/Tests/SearchResultsDetailed.aspx?SiteName=DST&Search=&TestCode=TC0709&PriceListCategoryId=16]. Accessed 8 Dec 2016. 32. Booth RE Brownlie J Establishing a pilot bovine viral diarrhoea virus eradication scheme in Somerset Vet Rec 2012 170 73 10.1136/vr.100191 22027186 33. Booth RE Cranwell MP Brownlie J Monitoring the bulk milk antibody response to BVDV: the effects of vaccination and herd infection status Vet Rec 2013 172 449 10.1136/vr.101195 23474586 34. Fredriksen B Sandvik T Loken T Odegaard SA Level and duration of serum antibodies in cattle infected experimentally and naturally with bovine virus diarrhoea virus Vet Rec 1999 144 111 4 10.1136/vr.144.5.111 10070699 35. Niskanen R Relationship between the levels of antibodies to bovine viral diarrhoea virus in bulk tank milk and the prevalence of cows exposed to the virus Vet Rec 1993 133 341 4 10.1136/vr.133.14.341 8236677 36. Valle PS Wayne Martin S Skjerve E A Bayesian approach to estimating the performance of a bovine virus diarrhoea virus (BVDV) antibody ELISA bulk-tank milk test Prev Vet Med 2001 50 71 87 10.1016/S0167-5877(01)00221-5 11448496 37. The Cattle Book 2008. [https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/69220/pb13572-cattlebook-2008-090804.pdf]. Accessed 8 Dec 2016. 38. Houe H Epidemiological features and economical importance of bovine virus diarrhoea virus (BVDV) infections Vet Microbiol 1999 64 89 107 10.1016/S0378-1135(98)00262-4 10028165 39. Chamorro MF Walz PH Haines DM Passler T Earleywine T Palomares RA Riddell KP Galik P Zhang Y Givens MD Comparison of levels and duration of detection of antibodies to bovine viral diarrhea virus 1, bovine viral diarrhea virus 2, bovine respiratory syncytial virus, bovine herpesvirus 1, and bovine parainfluenza virus 3 in calves fed maternal colostrum or a colostrum-replacement product Can J Vet Res 2014 78 81 8 24688168

PMC005xxxxxx/PMC5002206.txt

==== Front BMC Med Res MethodolBMC Med Res MethodolBMC Medical Research Methodology1471-2288BioMed Central London 18310.1186/s12874-016-0183-6DebateResuming the discussion of AMSTAR: What can (should) be made better? Wegewitz Uta wegewitz.uta@baua.bund.de 1Weikert Beate weikert.beate@baua.bund.de 1Fishta Alba Fishta.alba@baua.bund.de 1Jacobs Anja Anja.jacobs@g-ba.de 2Pieper Dawid +49 221 98957-40dawid.pieper@uni-wh.de 31 Federal Institute for Occupational Safety and Health (BAuA), Nöldnerstr. 40-42, 10317 Berlin, Germany 2 The Federal Joint Committee (G-BA), Wegelystr. 8, 10623 Berlin, Germany 3 Institute for Research in Operative Medicine, Witten/Herdecke University, Ostmerheimer Str. 200 (Building 38), 51109 Cologne, Germany 26 8 2016 26 8 2016 2016 16 1 11113 2 2016 3 7 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Evidence syntheses, and in particular systematic reviews (SRs), have become one of the cornerstones of evidence-based health care. The Assessment of Multiple Systematic Reviews (AMSTAR) tool has become the most widely used tool for investigating the methodological quality of SRs and is currently undergoing revision. The objective of this paper is to present insights, challenges and potential solutions from the point of view of a group of assessors, while referring to earlier methodological discussions and debates with respect to AMSTAR. Discussion One major drawback of AMSTAR is that it relies heavily on reporting quality rather than on methodological quality. This can be found in several items. Furthermore, it should be acknowledged that there are now new methods and procedures that did not exist when AMSTAR was developed. For example, the note to item 1 should now refer to the International Prospective Register of Ongoing Systematic Reviews (PROSPERO). Furthermore, item 3 should consider the definition of hand-searching, as the process of reviewing conference proceedings using the search function (e.g. in Microsoft Word or in a PDF file) does not meet the definition set out by the Cochrane Collaboration. Moreover, methods for assessing the quality of the body of evidence have evolved since AMSTAR was developed and should be incorporated into a revised AMSTAR tool. Summary Potential solutions are presented for each AMSTAR item with the aim of allowing a more thorough assessment of SRs. As the AMSTAR tool is currently undergoing further development, our paper hopes to add to preceding discussions and papers regarding this tool and stimulate further discussion. Keywords Systematic reviewMethodsDecision makingEvidence-Based Medicineissue-copyright-statement© The Author(s) 2016 ==== Body Background Evidence syntheses, and in particular systematic reviews (SRs), have become one of the cornerstones of evidence-based health care. If SRs are methodologically sound, they are considered to provide the highest level of evidence for medical decision-making. Over recent years, the Assessment of Multiple Systematic Reviews (AMSTAR) tool has become the most widely used tool for investigating the methodological quality of SRs [1, 2]. It was developed based on the Overview Quality Assessment Questionnaire [3, 4] and the checklist by Sacks [5] and consists of 11 items, each of which is categorized into a standardized set of four possible responses: “yes”, “no”, “can’t answer” or “not applicable”. The items relate to a priori design, study selection and data extraction, the literature search, grey literature, the list of included and excluded studies, study characteristics, critical appraisal, the formulation of conclusions, the combination of study results, publication bias and conflicts of interest. The measurement properties were recently described in a SR and found to be satisfactory in terms of interrater reliability, validity and applicability (scoring takes between 10 and 20 minutes), although it has also been stated that some methodological issues are still open to discussion, such as the test–retest reliability and whether and how an overall score should be calculated [6]. It should also be noted that several authors have modified or augmented the original AMSTAR tool, for example by adding new items, splitting items, or altering the rationales for answering or scoring items [7–11]. Two recently published papers present AMSTAR from the perspective of an assessor [12, 13]. Both papers describe challenges that the assessors faced when applying AMSTAR to SRs and present potential solutions to these challenges. The main idea of our paper is to continue the discussion raised by these papers and to present another perspective from a different group of assessors. In doing so, we will focus only on the evaluation of SRs of randomized controlled trials (RCTs) using AMSTAR. This is an ideal time for discussion, as the AMSTAR tool is currently undergoing further development. According to the developers of AMSTAR, this development will be two-fold [14]. Firstly, a new version of AMSTAR will improve upon the original tool. Secondly, a version of AMSTAR will be developed for SRs of non-randomized studies; this will be called AMSTAR-NRS. According to the tool’s developers, AMSTAR can be applied to a wide variety of SRs, although they recognize that its original development only took account of SRs of RCTs that evaluated treatment interventions [15]. The further development of AMSTAR is a sensible step. Even in the first AMSTAR paper, the tool’s developers acknowledged that new evidence would modify current thinking and that an update would be inevitable [16]. Now that AMSTAR is to be updated and further developed, it would make sense to gather the experiences of AMSTAR users from around the world, as these might provide important ideas that could be taken into account in the new, updated version of the tool. The following discussion is structured according the order of the AMSTAR items, and each item is discussed separately. A comparison with other comparable AMSTAR discussion papers is included at the end of the discussion. Discussion Was an ‘a priori’ design provided? The research question and inclusion criteria should be established before the conduct of the review. Note: Need to refer to a protocol, ethics approval, or pre-determined/a priori published research objectives to score a “yes.” [17]. An a priori protocol is an important component of SRs. The intention is to reduce the risk of bias related to selective reporting by stating a priori hypotheses and methods explicitly [18]. However, a thorough evaluation should focus on discrepancies between the protocol and the final review, which research has shown to be fairly common [19, 20]. It is not sufficient only to refer to a protocol, therefore, as this would primarily reflect the quality of reporting. In addition, clarification should be given of which aspects are included in a protocol or publication that is published a priori. Times have changed since the development of AMSTAR. Some years ago, it was quite common for protocols typically only to be available for Cochrane reviews. In recent years, there have been many efforts to increase the quantity of systematic review protocols. It is worth mentioning the creation of the International Prospective Register of Ongoing Systematic Reviews (PROSPERO), which was launched in February 2011 [21]. This allows the documentation of 22 mandatory and 18 optional items with regard to the a priori design and conduct of a review. At the time of writing (February 2016), approximately 10,000 records (including Cochrane protocols) can be found in PROSPERO. It would probably make sense to mention PROSPERO as an example under this item. Furthermore, Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) were published last year [22], and SR protocols are published by journals such as Systematic Reviews, for example.2. Was there duplicate study selection and data extraction? There should be at least two independent data extractors and a consensus procedure for disagreements should be in place. Note: 2 people do study selection, 2 people do data extraction, consensus process or one person checks the other’s work. [17]. This item addresses the two critical aspects of study selection and data extraction. Study selection should be performed by two people independently according to AMSTAR. It is important to bear in mind that there are also new approaches such as “liberal accelerated study selection,” in which the second reviewer reviews only the items that were excluded by the first reviewer [23]. Erroneously included records can still be excluded at a later stage. So far, this approach has been more prevalent in rapid reviews than in SRs. However, there is no obvious reason why this method should introduce bias into SRs. AMSTAR provides for the possibility of one author checking the data extracted by another. This procedure is seen as error-prone because the quality of data extraction is not as good as if the reviewers had acted independently [24]. However, to the best of our knowledge, this is the only study that has investigated the issue of double data extraction; accordingly, the evidence is very limited. Furthermore, the reviewers’ levels of expertise might be presumed to influence data extraction. For instance, an inexperienced reviewer might check the work of an expert in a much more lenient manner than would be appropriate. This item will also need to be updated for future applications, as automated (computerized) data extraction is bound to emerge as an issue in systematic reviews [25]. In such situations, a human could check the correctness and completeness of the extracted data. As a concluding remark, we would like to point out that the inclusion of two different aspects (study selection and data extraction) within one item might unnecessarily impede the application of AMSTAR in the event that only one of the two aspects is fulfilled.3. Was a comprehensive literature search performed? At least two electronic sources should be searched. The report must include years and databases used (e.g., Central, EMBASE, and MEDLINE). Key words and/or MESH terms must be stated and where feasible the search strategy should be provided. All searches should be supplemented by consulting current contents, reviews, textbooks, specialized registers, or experts in the particular field of study, and by reviewing the references in the studies found. Note: If at least 2 sources + one supplementary strategy used, select “yes” (Cochrane register/Central counts as 2 sources; a grey literature search counts as supplementary). [17]. For the sake of reproducibility and to ensure comprehensive assessment of the literature search strategy, assessors should have access to details of the entire search strategy. However, AMSTAR simply requires the reporting of keywords and MESH terms for the awarding of a “yes” judgement. The full search strategy is to be set out “where feasible” but it is not obvious what this means. Today, it is possible in most cases to set out the full search strategy in an online supplement or at least to make it available upon request. As a matter of common practice, therefore, review authors provide the full search strategy. Another problem centres around whether supplementary search strategies are adequate. In our opinion, the type of supplementary strategy used in SRs forms a critical aspect of a comprehensive literature search. In the case of SRs of clinical studies, for example, authors should focus their search on clinical study registers rather than on screening text books or dissertations. Furthermore, it should be clarified that it is mandatory for review authors to review the references in the identified studies. We therefore propose that the note be expanded as follows: If at least 2 sources + one adequate supplementary strategy were used, and references of the included studies were screened, select “yes”. Furthermore, it should be borne in mind that different authors might have different definitions of hand-searching. Nowadays, conference proceedings can also be reviewed by using the search function, as most conference proceedings are available as PDF files; this falls under the definition of hand-searching given by the Cochrane Collaboration.4. Was the status of publication (i.e. grey literature) used as an inclusion criterion? The authors should state that they searched for reports regardless of their publication type. The authors should state whether or not they excluded any reports (from the systematic review), based on their publication status, language etc. Note: If review indicates that there was a search for “grey literature” or “unpublished literature,” indicate “yes.” SINGLE database, dissertations, conference proceedings, and trial registries are all considered grey for this purpose. If searching a source that contains both grey and non-grey, must specify that they were searching for grey/unpublished lit. [17]. This item addresses multiple aspects of study selection: the publication status, the publication type, and language restrictions. Some concerns are associated with it. In order to fulfil this item, review authors need to search for grey literature. However, all that is required is a statement of whether or not the publication type and language restriction were used as inclusion criteria. Ignoring studies written in languages other than English could introduce a risk of bias. However, evidence shows that omitting non-English articles may have only a small effect, if any, on the conclusions of a systematic review [26, 27]. Nevertheless, it appears to be important to bear the research objective in mind; in other words, the preceding statement might not apply to research areas with a high prevalence of non-English articles and research [27, 28]. Furthermore, if review authors searched databases that appear to contain grey literature but did not mention this explicitly, AMSTAR recommends a “no” rating. This seems to be a questionable choice, as it is obvious that this item might in fact have been fulfilled. In conclusion, this item tends to evaluate the reporting quality rather than the methodological quality.5. Was a list of studies (included and excluded) provided? A list of included and excluded studies should be provided. Note: Acceptable if the excluded studies are referenced. If there is an electronic link to the list but the link is dead, select “no.” [17]. This item covers the documentation of included and excluded studies. In our opinion, it would be better to split the item. Complete documentation of excluded studies is an important aspect that review authors seldom fulfil. This should therefore be addressed in a single, additional item, as most recently published SRs provide a full list of included studies. However, we are aware that this would yield a large number of “yes” responses for the documentation of included studies and therefore to little variation across ratings. Another strategy for the documentation of excluded studies is proposed by the Cochrane Collaboration. According to the Cochrane Handbook, the list should not include all studies retrieved by the literature search but should contain studies that the reader might expect to be included [29]. In more specific terms, this means that Cochrane review authors present studies that are eligible at first glance but do not meet the eligibility criteria. In addition, well-known reports will also be considered if they are likely to be relevant to some readers. One can assume that this strategy introduces some subjectivity to the documentation process, which in most cases will not be reproducible by review assessors. We propose obligatory documentation of all studies excluded after the full text versions are evaluated. A description of other relevant studies may be helpful for the reader, and this could be presented on a voluntary basis.6. Were the characteristics of the included studies provided? In an aggregated form such as a table, data from the original studies should be provided on the participants, interventions and outcomes. The ranges of characteristics in all the studies analyzed e.g., age, race, sex, relevant socioeconomic data, disease status, duration, severity, or other diseases should be reported. Note: Acceptable if not in table format as long as they are described as above. [17]. This item is one of the most subjective, as there is no defined threshold for assigning a “yes” or “no” rating. In our experience, the majority of reviews meet the minimum requirements for this item, resulting in a high proportion of “yes” answers.7. Was the scientific quality of the included studies assessed and documented? ‘A priori’ methods of assessment should be provided (e.g., for effectiveness studies if the author(s) chose to include only randomized, double-blind, placebo controlled studies, or allocation concealment as inclusion criteria); for other types of studies alternative items will be relevant. Note: Can include use of a quality scoring tool or checklist, e.g., Jadad scale, risk of bias, sensitivity analysis, etc., or a description of quality items, with some kind of result for EACH study (“low” or “high” is fine, as long as it is clear which studies scored “low” and which scored “high”; a summary score/range for all studies is not acceptable). [17]. We believe that this item would benefit from some additional guidance. In our understanding, the item will most likely reflect reporting quality if the assessors simply look at whether an assessment has been made of the scientific quality of the included studies. The item should therefore ask whether the methodological quality has also been adequately assessed. The response will to some extent be subjective, as adequacy can be judged differently by different people. This might not be as important for SRs of randomized controlled trials, where the Cochrane Risk of Bias Tool can be used as a standard. In our opinion, the choice of the most adequate tool is not clear in the case of non-randomized studies. For example, the Newcastle–Ottawa Scale has been recommended by some journals but has also been criticized by others [30, 31]. It is important to add that the term “scientific quality” could be misleading in this context. This is because reporting quality might also be regarded as a form of scientific quality despite not being the focus of this item.8. Was the scientific quality of the included studies used appropriately in formulating conclusions? The results of the methodological rigor and scientific quality should be considered in the analysis and the conclusions of the review, and explicitly stated in formulating recommendations. Note: Might say something such as “the results should be interpreted with caution due to poor quality of included studies.” Cannot score “yes” for this question if scored “no” for question 7. [17]. With regard to the integration of the methodological quality of individual studies, SR authors rarely provide a definition, or a good description, of the process of deriving conclusions from studies’ results. Different groups of authors might draw different conclusions. To improve the plausibility of this step, we propose not only assessing the quality of individual studies but also discussing the quality of the body of evidence. Quality of evidence describes how confident one can be that the estimate of an intervention’s effectiveness is true [32]. An evaluation of the body of evidence allows a link to be drawn between the quality of the overall evidence and the strength of the conclusions. Multiple tools exist for assessing and characterizing the quality of a body of evidence [33]. When the quality assessment of the body of evidence is integrated into the AMSTAR checklist, examples should be given of existing methods, e.g. the system proposed by the Agency for Healthcare Research and Quality (AHRQ) [34] or by the Grading of Recommendations Assessment, Development and Evaluation (GRADE) Working Group [35].9. Were the methods used to combine the findings of studies appropriate? For the pooled results, a test should be done to ensure the studies were combinable, to assess their homogeneity (i.e., Chi-squared test for homogeneity, I2). If heterogeneity exists a random effects model should be used and/or the clinical appropriateness of combining should be taken into consideration (i.e., is it sensible to combine?). Note: Indicate “yes” if they mention or describe heterogeneity, i.e., if they explain that they cannot pool because of heterogeneity/variability between interventions. [17]. This item appears to be useful and easy to understand but suffers from some ambiguities. For example, in the case of strong clinical or statistical heterogeneity, the problem of heterogeneity might not be solved by using the random effects model. Readers are often faced with SRs that highlight substantial heterogeneity between the included studies. But how should these SRs be evaluated when the authors have conducted a meta-analysis without discussing the appropriateness of combining the studies? AMSTAR does not provide guidance for such cases. The item could therefore be improved by requiring SR authors to state or discuss the criteria used for qualitative and quantitative synthesis of the body of evidence [36]. The choice of whether to conduct qualitative or quantitative synthesis is not always an obvious one. For example, two SRs examined the effectiveness of qigong for chronic conditions in the elderly [37, 38]. Given the similarity of the two SRs, one would expect both to have used the same synthesis method. In fact, one SR performed a qualitative synthesis, while the other performed a quantitative synthesis without giving any rationale for this choice. Furthermore, clinical expertise in the research topic is required in order to evaluate a SR and especially when judging the clinical heterogeneity of the included studies. There may also be some obstacles to reaching a consensus on the assessment of reviews without meta-analyses. In this context, an explanation should be provided of whether this item is relevant to all SRs or only to SRs with a meta-analysis. Some authors provide a judgment of this item in all cases, while others state that it is not applicable in the case of narrative synthesis. Is it sufficient to cite clinical heterogeneity between studies without further explanation? In order to receive a score of “yes”, review authors should assess both the clinical and the statistical heterogeneity. Furthermore, we would like to point out that SRs can contain more than one pooled effect measure, i.e. multiple meta-analyses. This is often the case if more than one outcome is investigated and a meta-analysis is performed for each outcome. In such cases, it may be difficult to judge this item at the review level. Instead, it might be more appropriate to reach a judgment at the outcome level, which would make it easier to reflect differences in this item. For example, a SR might have applied fixed effect model meta-analysis appropriately to a number of outcomes. It would not be appropriate, however, to apply fixed effect model meta-analysis to all outcomes, as the best term to reflect the resulting judgment would probably be “partly”, which is not one of the possible answers under AMSTAR. We therefore believe that item 9 should ideally be answered at the outcome level. This would also be in line with the Cochrane Risk of Bias Tool or with GRADE [39].10.   Was the likelihood of publication bias assessed? An assessment of publication bias should include a combination of graphical aids (e.g., funnel plot, other available tests) and/or statistical tests (e.g., Egger regression test, Hedges-Olken). Note: If no test values or funnel plot included, score “no”. Score “yes” if mentions that publication bias could not be assessed because there were fewer than 10 included studies. [17]. In accordance with the above note, reviews with fewer than 10 included studies should receive a “yes” rating, provided that the authors stated that there were too few studies for an assessment of publication bias. In such cases, there is a high risk that reporting quality will be assessed instead of methodological quality. SRs that have fulfilled this item may vary substantially with respect to the assessment of publication bias. Though the difference between 9 and 10 may in fact be small, we appreciate that the AMSTAR tool provides a clearly defined threshold. Nonetheless, the idea that publication bias cannot be assessed due to a small number of studies means it is worth considering “not applicable” as a possible judgment. Furthermore, this item should be given a score of “yes” in the case of SRs that assess the quality of evidence using GRADE methodology [40]. One possible justification of this judgment is that GRADE requires its users to assess the likelihood of publication bias. However, some review authors simply provide the GRADE Summary of Findings Table without addressing the issue of publication bias. In such cases, we would appreciate further guidance on how to judge this item. In our opinion, in order to fulfil this item, SR authors ought to discuss the efforts and methods they used to assess publication bias even if they applied the GRADE methodology. In addition, we propose that assessors not only evaluate funnel plots or statistical tests but also consider the search strategy and/or potential conflicts of interest [41].11.   Was the conflict of interest included? Potential sources of support should be clearly acknowledged in both the systematic review and the included studies. Note: To get a “yes,” must indicate source of funding or support for the systematic review AND for each of the included studies. [17] This item clearly addresses the potential conflicts of interest both in the review itself and in the included studies. However, it emphasizes the sources of funding or support, whereas readers need also to be aware of any personal conflicts of interest. One example of personal involvement is the inclusion of studies undertaken by a review author. To minimize the potential for influence in judgments, the assessment of eligibility and risk of bias in primary studies should be carried out by an author who was not involved [42]. Generally, this item is seldom fulfilled even in Cochrane reviews (which are known to be of high quality). This is not surprising, as neither PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) [43] nor the Cochrane Handbook requires the reporting of funding sources for trials. We warmly welcome more-detailed assessment of conflicts of interest. This item should reflect the risk of bias due to conflicts of interest instead of simply examining whether they are reported, although we openly acknowledge that this task is a highly challenging one. Comparison with other assessors’ opinions Recently, two other debate articles have been published regarding AMSTAR that provide a discussion at the item level [12, 13]. Both articles are based on the author’s experiences with AMSTAR. Faggion Jr. focused on two points in his analysis: the ability of AMSTAR to assess the methodological quality of SRs and challenges in the interpretation of items and their scoring guidance. He also provided suggestions for improving the tool [12]. Burda et al. followed the same idea, while also providing suggestions for rewording items and their scoring guidance [13]. As they did this for all 11 items, the results could also be referred to as a Burda version of AMSTAR. In addition, they also discuss the relevance of AMSTAR response categories. For item 1, we placed greater emphasis on a published protocol rather than on an explicit statement in the review that review questions and inclusion criteria were defined a priori. There is still some debate as to whether checking someone else’s data extraction is an acceptable method of double data extraction given the lack of evidence in this field. With respect to item 2, all authors agree that it should be made much clearer that study selection and data extraction should be performed by two people. All three publications take the same approach to item 3. Faggion Jr. also presents a special emphasis on hand-searching [12], while only our manuscript strictly argues for the compulsory reporting of search strategy, assuming that this is feasible in all cases nowadays. Item 4 definitely needs to be reworded, as stated by all three publications, but while Faggion Jr. focuses on the number of sources [12], Burda et al. join us in highlighting the issue that this item is heavily reliant on reporting [13]. With respect to item 5, Faggion Jr. criticizes that this item does not state clearly whether the list of excluded studies should include hits from the stages of title/abstract screening or from the stage of full text evaluation [12]. He proposes reporting the full list of excluded studies from both stages, which is in line with the recommendation given by the PRISMA Statement [44]. However, a sensitive search strategy can result in a huge number of hits. From our point of view, we would like to question whether the benefit of this information justifies the huge effort involved in documenting the reasons for exclusion at the title/abstract screening stage. This concurs with Burda et al., who suggest only reporting the results of the full text screening [13]. Again, it is our opinion and the opinion of Burda et al. that item 6 will remain highly subjective [13], while Faggion Jr. proposes a threshold for a minimum amount of information [12]. In our view, this might not be feasible, as the minimum amount of information will probably depend on the objective of the review and the included studies. With respect to item 7, all of the publications recognise a need for revision. As a whole, it should be clarified that this item is about the risk of bias and that a “yes” score can only be obtained if an appropriate tool is selected for critical appraisal. With regard to item 8, Burda et al. also mention the GRADE approach [13], which Faggion Jr. had already mentioned in relation to item 7 [12]. Overall, all three publications arrive at a recommendation for the use of GRADE in SRs. There is a great deal of discussion with respect to item 9. While Faggion Jr. highlights the problems of quantitative synthesis [12], Burda et al. also make some points with regard to qualitative synthesis [13]. However, our manuscript is alone in calling for this item to be assessed at the outcome level rather than at the review level because heterogeneity can vary depending on the item being studied. The item concerning publication bias (item 10) raised discussions about the mixing methodological quality and reporting quality, as mentioned by Faggion Jr. [12] and by our group. This also raises the question of how this item should be rated. In this context, reference is made to GRADE both by Burda et al. and in our publication [13]. In general, all three author groups agree that item 11 must be clarified (e.g. that COI should be presented for all review authors), although there are differences regarding the degree of rigour needed to obtain a “yes.” Summary This debate paper presents methodological reflections in relating to AMSTAR, a validated and widely used tool for evaluating the methodological quality of SRs. It provides insights and challenges from the point of view of a group of assessors, while referring to earlier methodological discussions and debates with respect to AMSTAR. Potential solutions are presented for each AMSTAR item with the aim of allowing a more thorough assessment of SRs. As the AMSTAR tool is undergoing further development, our paper hopes to add to preceding discussions and papers regarding the AMSTAR tool and to stimulate further discussion. Abbreviations AMSTAR, Assessment of Multiple Systematic Reviews; PROSPERO, International Prospective Register of Ongoing Systematic Reviews; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses; RCTs, Randomized controlled trials; SRs, systematic reviews Funding There was no funding. Availability of data and materials Not applicable. Authors’ contributions All authors made substantial contributions to conception and design, discussion and interpretation of data, and development of recommendations. UW and DP drafted the manuscript. All authors have been involved in revising the manuscript critically for important intellectual content. All authors gave final approval of the version to be published and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Competing interests The authors declare that they have no competing interests. Consent for publication Not applicable Ethics approval and consent to participate Not applicable. ==== Refs References 1. Hartling L A descriptive analysis of overviews of reviews published between 2000 and 2011 PLoS One 2012 7 11 e49667 10.1371/journal.pone.0049667 23166744 2. Pieper D Overviews of reviews often have limited rigor: a systematic review J Clin Epidemiol 2012 65 12 1267 73 10.1016/j.jclinepi.2012.06.015 22959594 3. Oxman AD Guyatt GH Validation of an index of the quality of review articles J Clin Epidemiol 1991 44 11 1271 8 10.1016/0895-4356(91)90160-B 1834807 4. Oxman AD Agreement among reviewers of review articles J Clin Epidemiol 1991 44 1 91 8 10.1016/0895-4356(91)90205-N 1824710 5. Sacks HS Meta-analyses of randomized controlled trials N Engl J Med 1987 316 8 450 5 10.1056/NEJM198702193160806 3807986 6. Pieper D Systematic review found AMSTAR, but not R(evised)-AMSTAR, to have good measurement properties J Clin Epidemiol 2015 68 5 574 583 10.1016/j.jclinepi.2014.12.009 25638457 7. Johnson BT Methodological quality of meta-analyses on the blood pressure response to exercise: a review J Hypertens 2014 32 4 706 23 10.1097/HJH.0000000000000097 24463936 8. Huedo-Medina TB Methodologic quality of meta-analyses and systematic reviews on the Mediterranean diet and cardiovascular disease outcomes: a review Am J Clin Nutr 2016 103 3 841 50 10.3945/ajcn.115.112771 26864357 9. Berkhof M Effective training strategies for teaching communication skills to physicians: an overview of systematic reviews Patient Educ Couns 2011 84 2 152 62 10.1016/j.pec.2010.06.010 20673620 10. Kelley GA Kelley KS Effects of exercise in the treatment of overweight and obese children and adolescents: a systematic review of meta-analyses J Obes 2013 2013 783103 10.1155/2013/783103 24455215 11. Aziz T Methodological quality and descriptive characteristics of prosthodontic-related systematic reviews J Oral Rehabil 2013 40 4 263 78 10.1111/joor.12028 23330989 12. Faggion CM Jr Critical appraisal of AMSTAR: challenges, limitations, and potential solutions from the perspective of an assessor BMC Med Res Methodol 2015 15 63 10.1186/s12874-015-0062-6 26268372 13. Burda BU Holmer HK Norris SL Limitations of A Measurement Tool to Assess Systematic Reviews (AMSTAR) and suggestions for improvement Syst Rev 2016 5 1 58 10.1186/s13643-016-0237-1 27072548 14. N.N. 2015 09.09.2015]; Available from: http://www.amstar.ca/. 15. Shea BJ External validation of a measurement tool to assess systematic reviews (AMSTAR) PLoS One 2007 2 12 e1350 10.1371/journal.pone.0001350 18159233 16. Shea BJ Development of AMSTAR: a measurement tool to assess the methodological quality of systematic reviews BMC Med Res Methodol 2007 7 10 10.1186/1471-2288-7-10 17302989 17. AMSTAR working group. AMSTAR Checklist. 2016 20.05.16]; Available from: www.amstar.ca. 18. Kirkham JJ, Altman DG, Williamson PR. Bias due to changes in specified outcomes during the systematic review process. PLoS One. 2010;5(3). 19. Page MJ Bias due to selective inclusion and reporting of outcomes and analyses in systematic reviews of randomised trials of healthcare interventions Cochrane Database Syst Rev 2014 10 MR000035 25271098 20. Silagy CA Middleton P Hopewell S Publishing protocols of systematic reviews: comparing what was done to what was planned JAMA 2002 287 21 2831 4 10.1001/jama.287.21.2831 12038926 21. Booth A An international registry of systematic-review protocols Lancet 2011 377 9760 108 9 10.1016/S0140-6736(10)60903-8 20630580 22. Moher D Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement Syst Rev 2015 4 1 10.1186/2046-4053-4-1 25554246 23. Khangura S Evidence summaries: the evolution of a rapid review approach Syst Rev 2012 1 1 10 10.1186/2046-4053-1-10 22587960 24. Buscemi N Single data extraction generated more errors than double data extraction in systematic reviews J Clin Epidemiol 2006 59 7 697 703 10.1016/j.jclinepi.2005.11.010 16765272 25. Jonnalagadda SR Goyal P Huffman MD Automating data extraction in systematic reviews: a systematic review Syst Rev 2015 4 78 10.1186/s13643-015-0066-7 26073888 26. Morrison A The effect of English-language restriction on systematic review-based meta-analyses: a systematic review of empirical studies Int J Technol Assess Health Care 2012 28 2 138 44 10.1017/S0266462312000086 22559755 27. Pham B Language of publication restrictions in systematic reviews gave different results depending on whether the intervention was conventional or complementary J Clin Epidemiol 2005 58 8 769 76 10.1016/j.jclinepi.2004.08.021 16086467 28. Xue J Significant discrepancies were found in pooled estimates of searching with Chinese indexes versus searching with English indexes J Clin Epidemiol 2015 70 246 53 10.1016/j.jclinepi.2015.09.014 26410113 29. Higgins, JPTD, JJ. Chapter 7: Selecting studies and collecting data, in Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0, J.P.T.G. Higgins, S., Editor. 2011, The Cochrane Collaboration. 30. Hartling L Testing the Newcastle Ottawa Scale showed low reliability between individual reviewers J Clin Epidemiol 2013 66 9 982 93 10.1016/j.jclinepi.2013.03.003 23683848 31. Stang A Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses Eur J Epidemiol 2010 25 9 603 5 10.1007/s10654-010-9491-z 20652370 32. Schünemann HJ, et al. Interpreting Results and Drawing Conclusions, in Cochrane Handbook for Systematic Reviews of Interventions. 2008, John Wiley & Sons, Ltd. p. 359–387 33. IOM: Finding What Works in Health Care: Standards for Systematic Reviews. 2011; Available from: http://www.nationalacademies.org/hmd/Reports/2011/Finding-What-Works-in-Health-Care-Standards-for-Systematic-Reviews/Standards.aspx. Accessed 12 July 2016. 34. Berkman ND, et al. Grading the Strength of a Body of Evidence When Assessing Health Care Interventions for the Effective Health Care Program of the Agency for Healthcare Research and Quality: An Update. 2008 35. Guyatt G GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables J Clin Epidemiol 2011 64 4 383 94 10.1016/j.jclinepi.2010.04.026 21195583 36. Fu R, et al. Conducting Quantitative Synthesis When Comparing Medical Interventions: AHRQ and the Effective Health Care Program. 2008 37. Kemp CA Qigong as a therapeutic intervention with older adults J Holist Nurs 2004 22 4 351 73 10.1177/0898010104269313 15486154 38. Ng BH Tsang HW Psychophysiological outcomes of health qigong for chronic conditions: a systematic review Psychophysiology 2009 46 2 257 69 10.1111/j.1469-8986.2008.00763.x 19170945 39. Guyatt G GRADE guidelines: 11. Making an overall rating of confidence in effect estimates for a single outcome and for all outcomes J Clin Epidemiol 2013 66 2 151 7 10.1016/j.jclinepi.2012.01.006 22542023 40. Guyatt GH GRADE guidelines: 5. Rating the quality of evidence--publication bias J Clin Epidemiol 2011 64 12 1277 82 10.1016/j.jclinepi.2011.01.011 21802904 41. Deeks JJ, Higgins JPT, Altman DG. Analysing Data and Undertaking Meta-Analyses, in Cochrane Handbook for Systematic Reviews of Interventions. 2008, John Wiley & Sons, Ltd. p. 243–296 42. Kliner M Garner P When trial authors write Cochrane Reviews: competing interests need to be better managed Cochrane Database Syst Rev 2014 9 ED000089 25219979 43. Moher D Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement BMJ 2009 339 b2535 10.1136/bmj.b2535 19622551 44. Moher D Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement Ann Intern Med 2009 151 4 264 9 10.7326/0003-4819-151-4-200908180-00135 19622511

PMC005xxxxxx/PMC5002207.txt

==== Front J NeuroinflammationJ NeuroinflammationJournal of Neuroinflammation1742-2094BioMed Central London 68710.1186/s12974-016-0687-3ResearchEtifoxine improves sensorimotor deficits and reduces glial activation, neuronal degeneration, and neuroinflammation in a rat model of traumatic brain injury Simon-O’Brien Emmanuelle e.simonobrien@biocodex.fr Gauthier Delphine d.gauthier@biocodex.fr Riban Véronique v.riban@biocodex.fr Verleye Marc m.verleye@biocodex.fr Pharmacology Department, Biocodex, Chemin d’Armancourt, 60200 Compiègne, France 26 8 2016 26 8 2016 2016 13 1 20311 4 2016 18 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Traumatic brain injury (TBI) results in important neurological impairments which occur through a cascade of deleterious physiological events over time. There are currently no effective treatments to prevent these consequences. TBI is followed not only by an inflammatory response but also by a profound reorganization of the GABAergic system and a dysregulation of translocator protein 18 kDa (TSPO). Etifoxine is an anxiolytic compound that belongs to the benzoxazine family. It potentiates GABAergic neurotransmission, either through a positive allosteric effect or indirectly, involving the activation of TSPO that leads to an increase in neurosteroids synthesis. In several models of peripheral nerve injury, etifoxine has been demonstrated to display potent regenerative and anti-inflammatory properties and to promote functional recovery. Prior study also showed etifoxine efficacy in reducing brain edema in rats. In light of these positive results, we used a rat model of TBI to explore etifoxine treatment effects in a central nervous system injury, from functional outcomes to the underlying mechanisms. Methods Male Sprague-Dawley rats received contusion (n = 18) or sham (n = 19) injuries centered laterally to bregma over the left sensorimotor cortex. They were treated with etifoxine (50 mg/kg, i.p.) or its vehicle 30 min following injury and every day during 7 days. Rats underwent behavioral testing to assess sensorimotor function. In another experiment, injured rats (n = 10) or sham rats (n = 10) received etifoxine (EFX) (50 mg/kg, i.p.) or its vehicle 30 min post-surgery. Brains were then dissected for analysis of neuroinflammation markers, glial activation, and neuronal degeneration. Results Brain-injured rats exhibited significant sensorimotor function deficits compared to sham-injured rats in the bilateral tactile adhesive removal test, the beam walking test, and the limb-use asymmetry test. After 2 days of etifoxine treatment, behavioral impairments were significantly reduced. Etifoxine treatment reduced pro-inflammatory cytokines levels without affecting anti-inflammatory cytokines levels in injured rats, reduced macrophages and glial activation, and reduced neuronal degeneration. Conclusions Our results showed that post-injury treatment with etifoxine improved functional recovery and reduced neuroinflammation in a rat model of TBI. These findings suggest that etifoxine may have a therapeutic potential in the treatment of TBI. Keywords Traumatic brain injuryEtifoxineNeuroinflammationTSPONeurosteroidsFunctional recoveryCytokinesAstrogliosisNeuronal degenerationMicrogliaissue-copyright-statement© The Author(s) 2016 ==== Body Background Traumatic brain injury (TBI) is a leading cause of mortality and morbidity and can result in long-term physical, behavioral, and cognitive deficits. These deficits are due to both primary and secondary injuries. The primary injury occurs at the moment of TBI impact, with disruption of blood brain barrier and blood vessels that contribute to edema formation [1]. This primary injury precedes several downstream events contributing to secondary injury cascade of cellular, molecular, and metabolic pathological events, such as activation of brain-resident microglia and astrocytes, production of cytokines and chemokines, or recruitment of peripheral immune cells into the brain [2–4]. TBI is also followed by a profound reorganization of the GABAergic system and a dysregulation of the mitochondrial 18kDa translocator protein (TSPO) levels [5]. These events may lead to acute and chronic cell death and contribute to functional impairments [6, 7]. The onset of secondary injuries can be delayed for minutes to hours and persist for weeks to months [8, 9]. Etifoxine (EFX) ((2-ethylamino-6-chloro-4-methyl-4-phenyl-4H-3,1-benzoxazine hydrochloride) is an anxiolytic compound that belongs to the benzoxazine family. Its anxiolytic and anticonvulsant properties have been shown in rodents [10, 11], and it is effective in the treatment of adjustment disorders with anxiety in humans [12–14]. Etifoxine binds directly to the GABAA receptor, acting as a positive allosteric modulator, thus potentiating GABAergic synaptic transmission [10, 15–17]. EFX also activates TSPO [18] which plays an important role for the synthesis of neurosteroids [19]. Etifoxine was indeed shown to inhibit the binding of the selective TSPO ligand [3H]PK11195. In addition, i.p. administration of EFX increased plasma and brain concentrations of progesterone, pregnenolone, 5α-dihydroprogesterone, and allopregnanolone [18, 20]. The formation of these neurosteroids might result in a brain region-specific enhancement of GABAergic neuronal inhibition [21]. In the central nervous system (CNS), TSPO is generally expressed in both microglia and in reactive astrocytes [22]. Although TSPO level in the brain is low, it increases after brain injury [5, 22] and is thus a marker of brain injury and repair [5, 23–25]. Emerging evidence suggests that administration of TSPO selective ligands may be a therapeutic tool in the treatment of inflammatory conditions [26] and might be neuroprotective [27–29]. Some studies showed that EFX displays potent regenerative and anti-inflammatory properties, promotes functional recovery in experimental models of traumatic peripheral nerve injury, and reduces brain edema in rats [30–32]. To investigate further the potential neuroprotective and anti-inflammatory effect of EFX, we tested this compound in a model of TBI induced by controlled cortical impact (CCI) [33, 34] in male adult rats. This model has numerous advantages such as the possibility to control deformation parameters, including duration, depth, and velocity of impact. Methods Animal subjects All experiments were performed on male Sprague-Dawley rats weighing 250–300 g at the time of injury (Janvier laboratories, Le Genest Saint Isle, France). Rats were group-housed with ad libitum access to food and water in a temperature-controlled (22 ± 2 °C) and humidity-controlled (55 ± 10 %) environment. Lights were on a 12-h light/dark cycle (lights on at 7:00 a.m). Experiments were carried out in strict accordance with the European Community regulations for animal use in research (2010/63/EU directive), and all protocols were approved by the Local Ethical Committee (C2EAn°72). Drugs Etifoxine (batches 562 and 653, Biocodex, Gentilly, France) was suspended in 0.9 % saline containing 1 % Tween 80 (vol/vol). This compound and its vehicle were administered intraperitoneally (i.p.) in a volume of 5 mL/kg of bodyweight. Traumatic brain injury procedure Rats were subjected to TBI using a controlled cortical impactor (Impact One™, Leica Microsystem, Illinois, USA). Following deep anesthesia using isoflurane (induction 4 %, maintenance 2 %), rats were stabilized in a stereotaxic frame (David Kopf Instruments, California, USA) and placed on a heated pad. After exposing the skull, a 4-mm diameter circular craniotomy was performed using an electronic hand drill. Rats received contusion injury centered 2 mm lateral to the bregma over the left sensorimotor cortex with a velocity of 3.0 m/s, a depth of 1.5 mm, and a duration of 0.5 s [34–38]. The impactor tip was angled 8° vertically to maintain a perpendicular position in reference to the tangential plane of the brain convexity at the impact surface. A small quantity of saline solution at room temperature was directed at the site of drilling to prevent thermal injury [39]. After the impact, the scalp was sutured closed. Sham animals received craniotomy but no impact from the CCI device. All animals were closely monitored post-operatively with weight and health surveillance. Experimental design This study was divided in two parts. In the first part, we wanted to assess the effects of EFX treatment on sensorimotor deficits induced by CCI in rats. To this aim, we administered EFX at the anxiolytic-like effective dose of 50 mg/kg, i.p. [10], or its vehicle, 30 min after TBI (n = 10 for TBI-vehicle group, n = 8 for TBI-EFX group) or sham surgery (n = 9 for sham-vehicle group, n = 10 for sham-EFX group), then daily after behavioral testing during 1 week. The second part consisted in studying the effects of EFX treatment on neuroinflammation markers induced by TBI in our rat model (n = 5/group). The critical time-points to study were chosen based on the behavioral results. Effects of EFX treatment on sensorimotor performances Bilateral tactile adhesive removal test This test was performed on days 2, 5, and 7 post-CCI in order to test the forepaw sensitivity as well as motor impairments [40–42]. Each rat was placed into a transparent box (426 × 266 × 185 mm) during a habituation period of 60 s. Thereafter, two adhesive labels (13 mm diameter, Tough-Spots®) were applied with equal pressure on each animal’s wrist as bilateral tactile stimuli. The time to contact each paw (“contact time”) and to remove (“removal time”) the adhesive labels were measured with a maximum of 120 s. The contact time is defined as the time it takes for the rat to react to the presence of the adhesive labels. The rat may either shake its paw and/or bring it to its mouth. There were 5 days of pre-testing in order to obtain an optimal level of performance, to limit inter-individual variations and asymmetries. A trial ended when the rat removed both labels or after 3 min. Tapered beam walking test Motor performance and coordination can be measured with the tapered beam walking test [43]. Prior to surgery, all animals were trained to traverse a 165-cm long elevated, tapering beam with a 2-cm ledge in order to go in their home cage (Campden Instruments, Lafayette, IN, USA) (two trials per day for 5 days). Criterion performance was assessed as the ability to traverse the beam five times in a row without stopping. The day of the test (on days 2, 5, and 7 post-CCI), animals are videotaped while traversing the beam (four trials), and performance was rated on the percentage of contralateral hind limb faults (calculated by dividing the number of foot faults by the total number of steps and multiplying by 100) [44, 45]. Limb-use asymmetry test This procedure was adapted from Schallert et al. [46]. Injured animals use less often their contralateral limbs (injured site) to balance themselves while rearing along the wall of the cylinder [47]. Each animal was placed in a transparent glass cylinder (20 cm diameter and 38 cm height), and its spontaneous activity to explore the vertical surface with its forelimbs was analyzed by videotaping during 3 min. A mirror was placed behind the cylinder at an angle to allow the observer to record forelimb movements when the animal was turned away from the camera. Rats used either both forelimbs or a single one for an exploration. The number of both, right only, or left only explorations was counted. The laterality score was computed as follows: (number of left only − number of right only)/(number of left only + number of right only + number of both). Rats with unilateral lesion prefer to use the limb ipsilateral to the lesion. Effects of EFX treatment on neuroinflammation GFAP and CD68 immunostaining Forty-height hours after surgery (24 h after the second and last EFX administration), animals were deeply anesthetized with sodium pentobarbital (80 mg/kg) and perfused intracardially with phosphate buffer saline (PBS) at room temperature followed by ice-cold 4 % paraformaldehyde (PFA). Brains were removed and post-fixed in the same fixative for 24 h and soaked in 30 % sucrose solutions. Brains were then frozen and sectioned (50 μm) using a cryostat (Leica CM 1850, Leica, Wetzlar, Germany). Free-floating sections were immunostained for glial fibrillary acidic protein (GFAP), a marker of astrocytic activation, and cluster of differentiation 68 (CD68), a marker of macrophages and activated microglia, as follow: slides were rinsed in tris-buffered saline (TBS, pH 7.4), then incubated for 30 min with hydrogen peroxide to quench endogenous peroxidases, rinsed in TBS-Triton (TBS-T), and pre-incubated 60 min with 5 % blocking serum in TBS-T. Tissue sections were then incubated overnight at 4 °C with a rabbit anti-GFAP antibody (Abcam, Cambridge, UK) diluted (1:5000) or a mouse anti-CD68 antibody (Bio-Rad Laboratories, Hercules, CA, USA) diluted (1:200) in TBS-T containing 2 % normal goat serum (NGS). Slides were rinsed in TBS-T and incubated 30 min with a goat anti-rabbit biotinylated antibody (Abcam) diluted (1:500) or a goat anti-mouse biotinylated antibody (Abcam) diluted (1:500) in TBS-T and 2 % NGS. Slides were then rinsed and processed by using the Vectastain ABC Elite kit (Vector Laboratories, Burlingame, CA, USA) and stained with diaminobenzidine and hydrogen peroxide. Fluoro-jade B staining Fluoro-jade B (FJB) is a fluorescent marker that sensitively and specifically binds to degenerating neurons [48]. Brain sections were first incubated in absolute ethanol during 3 min, then 70 % ethanol during a minute, and rinsed in distilled water. Sections were then incubated in a solution of 0.06 % potassium permanganate for 15 min, rinsed in distilled water for a minute, and incubated in 0.0004 % solution of FJB (Interchim, Montluçon, France) for 30 min. Image analysis Slices were mounted on coverslips with Neo-Mount® mounting medium (VWR, Fontenay-sous-Bois, France) before analysis with an optical microscope (Leica DM 4000, Leica) coupled to a camera (Sony XCD-U100CR, Sony, Tokyo, Japan) and to an acquisition system (Archimed, Microvision Instruments, Evry, France). All images were acquired using the same light level by using a 20× (for CD68 and GFAP immunostaining) or 10× (for FJB staining) objective lens. The number of CD68+ cells was counted manually: three sections from each rat were selected for analysis. One section was at the epicenter of the impact (bregma 0 mm) while the other two were either 300 μm rostral or caudal to the epicenter. For GFAP expression, 12 non-consecutive sections were immunostained, representing the area of the lesion (approximately bregma +1.8 mm to bregma −1.8 mm) and quantified. Sections were digitized as grayscale images after background subtraction for quantification of intensity (Image J, NIH, version 1.46). For FJB staining, all sections were observed and photographed with blue excitation light (480 nm). The number of FJB+ cells was counted manually in the contusion margin along the cortex. For all analysis, brain sections from stereotaxically corresponding regions were used in sham-treated rats. Cytokine immunoassay Cytokine multiplex assay was performed on the cerebral cortex. Brain tissues were harvested and frozen 6 or 12 h post-injury separately from each hemisphere. Tissues were then homogenized in a lysis buffer (RIPA buffer, Abcam) containing protease inhibitor cocktail (Pierce™ protease inhibitor, Thermo Fisher Scientific, Waltham, MA, USA). Tissue homogenates were centrifuged at 13,000 rpm for 20 min at 4 °C. Supernatants were transferred to new tubes and used for analysis. All assays were performed according to the manufacturer’s instructions, in duplicates, and without adjustments to the recommended standard curve dilution. The concentration of 12 cytokines/chemokines (IL-1α, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12, TNF-α, GM-CSF, MCP-1, and MIP-1α) was determinate by Bio-Plex Pro™ rat cytokine assays (catalog number: 171-K1001M, Bio-Rad Laboratories, Hercules, CA, USA) in 500 μg of total protein of brain cortices homogenates (Bradford assay, Bio-Rad Laboratories). Briefly, magnetic beads conjugated with cytokine antibodies were loaded into the wells of a 96-well plate. After washing, standards and samples were added into wells and incubated for 60 min at room temperature on a shaking platform. The beads were washed and incubated with biotinylated detection antibody for 30 min at room temperature. Following the removal of excessive detection antibodies, streptavidin-phycoerythrin conjugate compound was added and allowed to incubate for 10 min at room temperature. Cytokines concentrations were determined by Bio-Plex Manager™ MP software version 6.1 and calculated against the standard curve. Samples with out of range (low) levels were assigned the limit of detection (LOD) (Table 1) of the assay for analysis purposes. The intra-assay percentage coefficient of variation (CV) ranged from 1.28 to 9.48 %.Table 1 Limit of detection (LOD) of the Bio-Plex Pro™ rat cytokine assays (Bio-Rad Laboratories, Hercules, CA, USA) Targets Assay sensitivity, pg/mL, LOD (limit of detection) IL-1α 1 IL-1β 2 IL-2 3 IL-4 1 IL-5 6 IL-6 10 IL-12 0.7 TNF-α 3 GM-CSF 0.6 MIP-1α 12 MCP-1 4 Statistical analysis All data are presented as mean ± standard error of the mean (SEM) and were submitted to a square root transformation to meet the requirements of normality if necessary. Behavioral data were analyzed using three two-way analysis of variance (ANOVA) with repeated measures (RM) with post hoc Tukey’s multiple comparison test. The first ANOVA consists in studying the effects of TBI on sensorimotor performances (comparison between sham-vehicle and TBI-vehicle groups) with two factors: time in RM and injury. Then, we did the same type of analysis for the effects of EFX treatment on the TBI group and on the sham-operated one. Significant main or interaction effects were followed by Tukey post hoc analyses. Biochemical data were analyzed using a two-way ANOVA (TBI or Sham) x (EFX or Vehicle) with post hoc Tukey’s multiple comparison test. Inflammatory load score (ILS) was calculated for each rats as follow: animals with the lowest values for a given cytokine (≤25th percentile) were assigned a 1, between the 26th and the 50th percentile animals were assigned a 2, between the 51th percentile and the 75th percentile animals were assigned a 3, and for the largest values (≥76th percentile), animals were assigned a 4. An ILS for each rat was calculated by summing the quartile rank for each of the pro-inflammatory cytokines. All statistical analyses were performed using SigmaStat® software v 3.5 (Systat Software Inc., San Jose, California, USA). Significance was set at p < .05. Results Etifoxine improves adhesive removal task at 2 days post-CCI Results showed that before surgery, all groups of rats felt the presence of the adhesive stuck on their wrists very quickly (less than 1 s for each animal). Statistical analyses (two-way RM ANOVA) revealed no significant effect of the injury factor (F1, 18 = 3.718, p = 0.071), a significant effect of time (F3, 54 = 10.873, p < 0.001) and a significant interaction (F3, 54 = 3.441, p = 0.023). Post hoc analyses showed that 2 days post-surgery, TBI-vehicle-treated rats needed significantly more time to feel the presence of the adhesive compared to pre-surgery (53.3 ± 16.5 s) (p < 0,001). Sham-operated animals did not present any deficit. In the injured group, two-way RM ANOVA showed no effect of treatment (F1, 16 = 2.208, p = 0.157), a significant effect of time (F3, 48 = 12.826, p < 0.001) and no significant interaction (F3, 48 = 1.954, p = 0.133). Post hoc analyses showed that EFX treatment significantly reduced the contact time in injured animals 2 days post-surgery (23.7 ± 9.7 s vs. 53.3 ± 16.5 s, p = 0.006). No significant differences were observed on day + 5 and day + 7 post-surgery in any groups (Fig. 1).Fig. 1 Bilateral adhesive removal test. Mean (+SEM) time to contact the contralateral adhesive in sham-vehicle (n = 9), TBI-vehicle (n = 10), sham-EFX (n = 10), and TBI-EFX (n = 8) groups. EFX treatment significantly improved the time to contact on day 2 post-CCI compared with vehicle treatment in TBI animals (## p < 0.01). ***p < 0.001 compared to pre-TBI; ### p < 0.001 and ## p < 0.01 compared to TBI-vehicle Mean removal time was 16.72 ± 1.85 s in all groups before surgery. Two-way RM ANOVA revealed a main effect of injury (F1, 18 = 6.150, p = 0.024), a main effect of time (F3, 54 = 10.747, p < 0.001) and no significant interaction (F3, 54 = 2.602, p = 0.062). Post hoc Tukey’s analyses revealed that injured-vehicle-treated animals spent significantly more time to remove the adhesive than sham-vehicle-treated animals at day 2 post-surgery (83.4 ± 16.2 s vs. 35.9 ± 12.6 s, p < 0.001). Two-way RM ANOVA showed no significant treatment effect in the injured group (F1, 16 = 2.744, p = 0.117) but a significant effect of time (F3, 48 = 14.258, p < 0.001) and no significant interaction (F3, 48 = 1.035, p = 0.386). Post hoc analyses showed that treatment with EFX significantly reduced the time to remove the adhesive in injured animals (52.6 ± 12.9 s vs. 83.4 ± 13.2 s, p = 0.043). No significant differences were observed on day + 5 and day + 7 post-surgery in any groups (Fig. 2).Fig. 2 Bilateral adhesive removal test. Mean (+SEM) time to remove the contralateral adhesive in sham-vehicle (n = 9), TBI-vehicle (n = 10), sham-EFX (n = 10), and TBI-EFX (n = 8) groups. EFX treatment significantly improved the time to remove the adhesive on day 2 post-CCI compared with vehicle treatment in TBI animals (# p < 0.05). ***p < 0.001 and *p < 0.05 compared to pre-TBI; ### p < 0.001 and # p < 0.05 compared to TBI-vehicle Etifoxine improves beam walk task at 2 days post-CCI Two-way repeated measures ANOVA revealed that exposure to CCI injury resulted in a main significant effect (F1, 18 = 10.014, p = 0.005), a significant main effect of time (F3, 54 = 10.624, p < 0.001) and a significant interaction (F3, 54 = 9.084, p < 0.001). Post hoc analyses showed that 2 days post-injury, TBI-vehicle-treated animals made more foot faults in comparison with sham-vehicle animals (39.8 ± 8.9 vs. 4.7 ± 2.5 %, p < 0.001). At day + 5 and day + 7 post-surgery, this impairment is not observed anymore. In the injured groups, two-way RM ANOVA showed no effect of treatment (F1, 16 = 1.296, p = 0.272), a significant main effect of time (F3, 48 = 14.666, p < 0.001) and no significant interaction (F3, 48 = 1.217, p = 0.314). Post hoc analyses showed that two days of EFX treatment significantly reduced the percentage of contralateral hind limb faults in TBI-EFX animals in comparison with TBI-vehicle-treated animals (25.3 ± 9.2 vs. 39.8 ± 8.9 %, p = 0.039) (Fig. 3).Fig. 3 Tapered beam walking test. Mean (+SEM) percentage of faults with the contralateral hind limb on the beam in sham-vehicle (n = 9), TBI-vehicle (n = 10), sham-EFX (n = 10), and TBI-EFX (n = 8) groups. Treatment with EFX significantly improved motor performance and coordination on day 2 post-CCI compared to vehicle treatment in TBI animals (# p < 0.05). ***p < 0.001 compared to pre-TBI; ### p < 0.001 and # p < 0.05 compared to TBI-vehicle Etifoxine improves limb-use asymmetry test at 2 days post-CCI Two-way ANOVA with repeated measures showed a significant effect of injury on rats (F1, 18 = 37.255, p < 0.001), a significant effect of time (F3, 54 = 17.433, p < 0.001) and a significant interaction (F3, 54 = 15.953, p < 0.001). Post hoc analyses indicated that injured-vehicle-treated rats had higher asymmetry scores at day + 2, day + 5, and day + 7 post-injury compared with sham-vehicle-treated animals (p < 0.001). Two-way RM ANOVA showed a significant main effect of EFX treatment in injured rats (F1, 16 = 7.203, p = 0.016), a significant effect of time (F3, 48 = 25.754, p < 0.001) and a significant interaction (F3, 48 = 3.280, p = 0.030). Tukey’s test showed a significant improvement in limb-use asymmetries in EFX-treated animals compared to vehicle-treated animals 2 days post-injury (p = 0.003) and 5 days post-injury (p = 0.017) (Fig. 4).Fig. 4 Limb-use asymmetry test. Mean (+SEM) asymmetry score in sham-vehicle (n = 9), TBI-vehicle (n = 10), sham-EFX (n = 10), and TBI-EFX (n = 8) groups. Higher scores indicate a greater use of the unimpaired forelimb. Treatment with EFX significantly improved asymmetry score on day 2 (## p < 0.01) and day 5 (# p < 0.05) post-CCI compared to vehicle treatment in TBI animals. ***p < 0.001 and **p < 0.01 compared to pre-TBI; ### p < 0.001, ## p < 0.01, and # p < 0.05 compared to TBI-vehicle Etifoxine decreases astrogliosis in the injured cortex at 2 days post-CCI Activated astrocytes were visualized by immunohistochemistry for GFAP. Two-way ANOVA showed a significant main effect of CCI (F1, 223 = 119.61, p < 0.001), a significant main effect of EFX treatment (F1, 223 = 21.34, p < 0.001), and no interaction (F1, 223 = 0.035, p = 0.851). Post hoc Tukey’s analyses showed that immunostaining for GFAP revealed evidence of a reactive astrogliosis in regions of the cortex underlying the impact site at day + 2 post-surgery (21.3 ± 1.6 % in the TBI-vehicle group vs. 6.8 ± 1.2 % in the sham-vehicle group, p < 0.001). EFX treatment significantly reduced GFAP immunoreactivity in injured animals (15.0 ± 1.3 % in the TBI-EFX group vs. 21.3 ± 1.6 % in the TBI-vehicle group, p = 0.001) as well as in sham animals (1.1 ± 0.3 % in the sham-EFX group vs. 6.8 ± 1.2 % in the sham-vehicle group, p = 0.001) (Fig. 5). Figure 5 shows the typical star-like hypertrophied aspect of astrocytes in response to injury which is a characteristic marker of the inflammatory cascade induced by CNS trauma and which contributes to later pathology.Fig. 5 Effect of EFX treatment on CCI-induced astrogliosis in rats 2 days post-injury. EFX (50 mg/kg, i.p.) or its vehicle were administered 30 min after surgery and 24 h after (n = 5/group). Astrogliosis was evaluated with GFAP immunostaining. TBI induced an increase in GFAP immunoreactivity in the ipsilateral side of the injury (p < 0.001). EFX treatment significantly reduced astrogliosis (p < 0.01) in injured animals in comparison with vehicle treatment. Images show representative sections from each group (ipsilateral and contralateral side). Data are mean + SEM. @@@ p < 0.001: comparison between sham-vehicle and sham-EFX groups; ### p < 0.001: comparison between sham-vehicle and TBI-vehicle groups; **p < 0.01: comparison between TBI-vehicle and TBI-EFX groups Etifoxine reduces activated microglia/macrophage density in the injured cortex at 2 days post-CCI Blood-borne and resident macrophages/activated microglia can be visualized by anti-CD68 staining (Fig. 6). Two-way ANOVA revealed a significant main effect of surgery (F1, 65 = 24.87, p < 0.001), a significant main effect of EFX treatment (F1, 65 = 13.09, p < 0.001), and no interaction (F1, 65 = 1.099; p = 0.298). As shown in Fig. 6, post hoc analyses showed a significant increase in CD68+ cells in the cortical contusion margin of rats 2 days post-CCI (584 ± 100 for TBI-vehicle group vs. 194 ± 61 for sham-vehicle group, p < 0.001). A significant reduction in the density of activated microglia/macrophages was observed in EFX-treated rats compared with vehicle controls (282 ± 33 in the TBI-EFX group vs. 584 ± 100 in the TBI-vehicle group, p < 0.001). No positive staining was observed in the hemisphere contralateral to the impact site.Fig. 6 Effect of EFX treatment on CCI-induced microglia/macrophage activation in rats 2 days post-injury. EFX (50 mg/kg, i.p.) or its vehicle were administered 30 min after surgery and 24 h after (n = 5/group). Microglia/macrophage activation was evaluated with CD68 immunostaining. TBI induced an increase in CD68 immunoreactivity in the ipsilateral side of the injury (p < 0.001). EFX treatment significantly reduced microglia/macrophage activation (p < 0.001) in injured animals in comparison with vehicle treatment. Images show representative sections from each group (ipsilateral and contralateral side). Data are mean + SEM. ### p < 0.001: comparison between sham-vehicle and TBI-vehicle groups; ***p < 0.001: comparison between TBI-vehicle and TBI-EFX groups Etifoxine reduces neuronal degeneration in the injured cortex at 2 days post-CCI Statistical analyses revealed a significant main effect of surgery (F1, 297 = 51.439, p < 0.001), a significant main effect of EFX treatment (F1, 297 = 20.042, p < 0.001), and a significant interaction (F1, 297 = 10.521, p = 0.001). Post hoc analyses showed that FJB+ cells with neuronal morphology were evident 2 days after CCI in the cortical contusion margin but not in the contralateral hemisphere. Ipsilateral minus contralateral FJB+ cells counting showed that neuronal degeneration was significantly more important in TBI-vehicle group compared with sham-vehicle group (98 ± 11 for TBI-vehicle group vs. 8 ± 3 for sham-vehicle group, p < 0.001). Etifoxine treatment significantly reduced the number of FJB+ cells in CCI rats compared with vehicle treatment (26 ± 5 for TBI-EFX group vs. 98 ± 11 for TBI-vehicle group, p < 0.001) (Fig. 7).Fig. 7 Effect of EFX treatment on CCI-induced neuronal degeneration in rats 2 days post-injury. EFX (50 mg/kg, i.p.) or its vehicle were administered 30 min after surgery and 24 h after (n = 5/group). Neuronal degeneration was evaluated with fluoro-jade B (FJB) staining. TBI induced an increase in FJB staining in the ipsilateral side of the injury (p < 0.001). EFX treatment significantly reduced neuronal degeneration (p < 0.001) in injured animals in comparison with vehicle treatment. Images show representative sections from each group (ipsilateral and contralateral side). Data are mean + SEM. ### p < 0.001: comparison between sham-vehicle and TBI-vehicle groups; ***p < 0.001: comparison between TBI-vehicle and TBI-EFX groups Etifoxine reduces cortical level of pro-inflammatory cytokines Different studies showed that the temporal change of several inflammatory cytokines peaked at 6 h after CCI [3, 49]. We therefore selected this time-point for evaluating EFX treatment on cytokine/chemokine expression. There are no differences within contralateral sites of injury between groups. Two-way ANOVA revealed statistical differences for four pro-inflammatory cytokines, IL-1α, IL-1β, IL-6, and TNF-α, and two chemokines: MCP-1 and MIP-1α. Post hoc testing showed that there was a significant increase in IL-1α, IL-1β, IL-6, TNF-α, MCP-1, and MIP-1α levels in the ipsilateral cortices to the injury site of TBI-vehicle-treated group in comparison with sham-vehicle-treated animals (p < 0.001). Statistical analyses also showed that EFX treatment significantly reduced this increase in pro-inflammatory cytokines and chemokines levels in the injured group compared to vehicle treatment (p < 0.001 for IL-1α, IL-1β, IL-6, and TNF-α, p < 0.05 for MCP-1 and MIP-1α). Data are summarized in Table 2. Because multiple testing across the analytes may not be fully accounted for by multiple test correction within each analyte, resulting in type I errors, we have also performed an inflammatory load score (ILS) analysis (Table 3). Two-way ANOVA revealed an overall effect of TBI on inflammation (F1,15 = 15.46, p = 0.001), an overall effect of EFX treatment (F1, 15 = 11.50, p = 0.004) on ILS, and no interaction (F1, 15 = 3.584, p = 0.078). Post hoc analysis showed that injured animals have a higher ILS compared to sham animals (p < 0.001) and that EFX treatment significantly reduced ILS in injured animals (p = 0.003) (see Table 3). We also tested the levels of pro-inflammatory cytokines IL-1α, IL-1β, IL-6, and TNF-α 12 h after surgery in order to study their evolution of expression. At 6 h time-point, the profound increase in pro-inflammatory cytokines associated with trauma was significantly reduced by EFX treatment (p < 0.001) compared with vehicle-treated rats to such an extent that these values were not different from those for the 12 h time-point (Fig. 8). We did not find any modifications of other tested cytokines in injured animals compared to sham animals (data not shown).Table 2 Concentrations of cytokines/chemokines in the injured cortex at 6 h post-CCI or sham-operation in rats treated with EFX or Vehicle control Sham-vehicle (pg/mL) Sham-EFX (pg/mL) TBI-vehicle (pg/mL) TBI-EFX (pg/mL) TBI-vehicle vs. TBI-EFX Ipsilateral Contralateral Ipsilateral Contralateral Ipsilateral Contralateral Ipsilateral Contralateral % decrease (ipsilateral side) Pro-inflammatory cytokines  IL-1α 5.75 ± 0.87 5.17 ± 0.22 5.53 ± 0.81 3.83 ± 0.88 40.41 ± 5.14*** 3.27 ± 0.63 13.55 ± 2.65** 1.32 ± 0.77 33  IL-1β 45.14 ± 2.89 37.68 ± 2.43 56.58 ± 9.43 34.77 ± 3.25 382.88 ± 64.66*** 33.98 ± 1.57 148.97 ± 25.50** 29.82 ± 1.72 39  IL-2 26.47 ± 0.75 24.84 ± 0.70 23.48 ± 0.50 22.89 ± 1.57 27.55 ± 4.42 21.55 ± 0.62 23.44 ± 1.64 20.46 ± 0.64 ns  IL-6 2.65 ± 0.33 1.76 ± 0.23 14.00 ± 8.89 3.21 ± 0.71 150.18 ± 21.25*** 4.34 ± 0.64 41.97 ± 10.25** 1.86 ± 0.51 28  IL-12 8.87 ± 0.29 8.54 ± 0.38 7.50 ± 0.24 7.50 ± 0.95 7.41 ± 0.26 6.95 ± 0.38 6.95 ± 0.20 7.10 ± 0.20 ns  TNF-α 5.34 ± 0.35 1.30 ± 0.14 5.93 ± 0.36 4.48 ± 0.84 16.34 ± 1.71*** 5.24 ± 0.62 9.91 ± 1.28** 5.91 ± 0.70 60  GM-CSF 2.49 ± 0.67 2.24 ± 0.49 2.06 ± 0.26 1.52 ± 0.29 1.20 ± 0.22 1.15 ± 0.17 1.07 ± 0.27 1.36 ± 0.14 ns Anti-inflammatory cytokines  IL-4 2.86 ± 0.13 2.50 ± 0.08 3.20 ± 0.53 2.85 ± 0.51 4.19 ± 0.78 2.54 ± 0.05 3.65 ± 0.72 2.32 ± 0.13 ns  IL-5 11.22 ± 0.24 10.74 ± 0.24 7.93 ± 0.31 6.72 ± 1.71 9.34 ± 0.63 9.24 ± 0.89 9.30 ± 1.18 9.00 ± 1.13 ns  IL-10 25.99 ± 1.60 24.13 ± 1.67 19.94 ± 2.04 19.66 ± 4.35 35.29 ± 4.98 17.52 ± 2.82 28.48 ± 2.45 23.12 ± 0.83 ns Macrophage-attracting chemokines  MIP-1α 1.71 ± 1.01 0.04 ± 0.04 2.64 ± 1.3 0.00 ± 0.00 90.42 ± 25.47*** 0.00 ± 0.00 36.99 ± 4.51* 0.24 ± 0.21 41  MCP-1 7.50 ± 1.97 4.30 ± 0.63 12.82 ± 5.16 4.25 ± 0.64 93.82 ± 21.66*** 5.53 ± 0.97 53.00 ± 7.45* 6.34 ± 1.55 56 Six hours post TBI, ipsilateral and contralateral cortices were analyzed with the Bio-Plex Pro™ rat cytokine assays for IL-1α, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12, TNF-α, GM-CSF, MIP-1α, and MCP-1 Data (n = 5/groups) represents mean ± SEM. p < 0.05 or less were considered statistically significant. Two-way analysis of variance was performed, followed by Tukey post hoc test NS not significant ***p < 0.001: comparison between sham-vehicle and TBI-vehicle groups in the ipsilateral side; **p < 0.001 and *p < 0.05: comparison between TBI-vehicle and TBI-EFX groups in the ipsilateral side Table 3 Inflammatory load score (ILS) of each group of rats (sham-vehicle, sham-EFX, TBI-vehicle, TBI-EFX, n = 5/groups) measured in the ipsilateral side of the lesion 6 hours post TBI Sham TBI Vehicle EFX p value Vehicle EFX p value Mean ILS (SEM) 2.371 (0.21) 2.114 (0.15) 0.292 3.371 (0.15) 2.464 (0.16) 0.003 Ipsilateral cortices were analyzed with the Bio-Plex Pro™ rat cytokine assays. The cytokines used for the score were IL-1α, IL-1β, IL-2, IL-6, IL-12, TNF-α, and GM-CSF. Data represents mean (SEM) Fig. 8 Cortical cytokine levels at 6 and 12 h post-CCI. The pro-inflammatory cytokines IL-1α, IL-1β, Il-6, and TNF-α were significantly elevated 6 h after CCI in the injury and margin regions compared to sham animals (***p < 0.001 for all groups). Statistical analyses showed no difference at 12-h time-point. Six hours after CCI, EFX treatment (50 mg/kg, i.p., 30 min after injury) significantly reduced IL-1α, IL-1β, IL-6, and TNF-α peaks (### p < 0.001 for all groups). ***p < 0.001: comparison between sham-vehicle and TBI-vehicle groups; ### p < 0.001: comparison between TBI-vehicle and TBI-EFX groups Discussion Our results showed that TBI model induced by CCI caused sensorimotor deficits in rats measured in three behavioral tasks: bilateral adhesive removal test, tapered beam walk test, and limb-use asymmetry test. An administration of EFX 30 min after the injury, followed by a second injection after 24 h, significantly improved functional impairments. This protective effect was associated with reduced neuroinflammation, astrogliosis, and neuronal degeneration. Following TBI, there is a wide range of cellular and biochemical pathological events resulting in neuron and glial cells death. Among these events, neuroinflammation is characterized by the activation of resident astrocytes and microglia and by the infiltration of leukocytes into the injury site. These events lead to the release of inflammatory cytokines and neurotoxic molecules which contribute to progressive cell death and functional impairments [50]. Our results suggest that EFX may reduce delayed neuronal death that contributes to neurological deficits [51] and promote survival in the injury border zone. This hypothesis is supported by the observations that EFX-treated animals showed fewer degenerative neurons than vehicle-treated rats as evaluated by FJB staining. Among the different mechanisms underlying delayed neuronal loss, post-traumatic neuroinflammation is a major factor. Activation of microglia is one of the central inflammatory responses after TBI [52]. Recent studies showed that brain inflammation has been associated with polarization of microglia and macrophages (M1 and M2 responses) [53–56]. M1 response is an activation of microglia/macrophages that produces pro-inflammatory substances such as IL-1β, IL-6, and TNF-α [56]. The M2 response is induced by anti-inflammatory cytokines such as IL-4 and IL-10 [53]. Inflammation is therefore an important tool for controlling TBI effects and enhancing recovery after injury. Thus, the suppression of M1 and/or the up-regulation of M2 response could be used to prevent secondary injury and promote repair processes in the brain. In a study using male Sprague-Dawley rats, Ansari showed an upregulation of pro-inflammatory cytokines (TNF-α, IL-1β, and IL-6) expression 6 h post-CCI and an upregulation of anti-inflammatory cytokines such as IL-4 and IL-10 but only 24 h post-injury [57]. These results support other studies showing that early after injury, M1 predominate over M2 response [54, 58, 59]. These observations are in accordance with our results showing an upregulation of pro-inflammatory cytokines such as TNF-α, IL-1α, IL-1β, and IL-6 early after injury but no upregulation of anti-inflammatory cytokines such as IL-4 and IL-10 6 h post-injury. Pro-inflammatory cytokines secreted by activated microglia also secrete chemokines that recruit additional inflammatory cells to the injury site and free radicals that are cytotoxic to neurons and can contribute to neurodegeneration after TBI [60, 61]. Elevated levels of pro-inflammatory cytokines have been detected in the serum and cerebrospinal fluid of TBI patients [62, 63] and in brain parenchyma of animals with experimental brain injuries [64–67]. The role of these pro-inflammatory cytokines has remained controversial, but several studies showed that suppressing their expression may reduce tissue damage and brain edema and also improve functional consequences [2, 68–78]. We demonstrated that acute administration of EFX significantly reduced TBI-induced up-regulation of pro-inflammatory cytokines. Since TNF-α, IL-1β, and IL-6 are major pro-inflammatory cytokines that mediate blood brain barrier disruption, edema, and programmed cell death resulting in loss of neurons [3, 71, 79–83], we can hypothesized that EFX may exerts its neuroprotective actions, at least in part, by suppressing the expression of these cytokines which are detrimental in the acute post-traumatic period. Experimental and clinical data show that GFAP and other intermediate filaments such as nestin and vimentin are considerably upregulated following TBI [84–86]. In our study, 2 days after TBI, there was a significant increase in GFAP immunoreactivity in the perimeter of the cortical lesion and in the subcortical white matter, reflecting increased astrocytic activation in the cortex ipsilateral to the injury site. This observation is in accordance with other studies [64, 85]. Two days of EFX treatment significantly reduced this astrocytic activation. This is important to note because astrocytic activation increases the production of pro-inflammatory cytokines, reactive oxygen species, excitatory amino acids, and nitrogen oxides and can also induce blood brain barrier disruptions that induce immune cell infiltration [87–89]. GFAP also promotes scarring and glial proliferation which is deleterious on neuronal regeneration and recuperation [90]. Our results are in accordance with other studies that show that TSPO ligands exert neuroprotective effects and reduce neural inflammation [26, 28–30]. Girard et al. indeed showed that EFX treatment decreased messenger RNA (mRNA) expression of pro-inflammatory cytokines TNF-α, Il-1β, and IL-6 and improved functional recovery in a model of peripheral nerve injury in rats and also had beneficial effects on axonal regrowth and on macrophage response [30, 32]. Ravikumar et al. showed that EFX treatment reduced lipopolysaccharide (LPS)-induced IL-1β and IL-6 release from rat primary astrocytes in a concentration-dependent manner [91]. Daugherty et al. also showed a decrease in pro-inflammatory cytokines after EFX treatment in a model of multiple sclerosis [92]. Etifoxine, which is a TSPO ligand, has been shown to be a potent enhancer of neurosteroidogenesis and to induce the synthesis of pregnenolone, progesterone, and allopregnanolone [18, 20, 93] which have demonstrated neuroprotective properties. Progesterone treatment following TBI in rats has indeed been demonstrated to reduce cerebral edema, cell death mediators expression, inflammatory response, reactive gliosis, and necrosis and to improve functional recovery [94–100]. In their studies, He at al. [8, 101] demonstrated that allopregnanolone significantly reduced neuronal loss and enhanced cognitive and behavioral recovery in a model of TBI in rats. These results suggest that EFX, a TSPO ligand, may protect neuronal function by inducing neurosteroids biosynthesis. In addition to its action on TSPO receptor, EFX exerts a direct effect on GABAA receptor as a positive allosteric modulator [10]. This is of interest because a significant decrease in inhibitory synaptic transmission has been shown in the basolateral amygdala of rats subjected to CCI, associated with a decrease of surface expression of GABAA receptors [102]. Furthermore, glutamatergic projections from the mediodorsal thalamic nucleus and cholinergic projections from the nucleus basalis magnocellularis to the prefrontal cortex are known to be modulated by GABA transmission [103, 104] and are highly sensitive to TBI [8, 94]. Excessive release of glutamate and acetylcholine following TBI causes excitotoxicity that could thus be prevented by direct action of EFX on inhibitory GABAA receptors and its indirect action on the synthesis of allopregnanolone, which also potentiate GABA transmission [105, 106]. Conclusions EFX treatment after TBI considerably improves functional outcomes. This effect is associated with reductions in neuronal degeneration, in glial scar formation and in microglial activation. Neuroinflammation is a major secondary injury mechanism that provides later neurodegeneration and neurologic impairments associated with TBI. We can hypothesize that functional and histological improvement of EFX-treated animals is due to its modulation of early inflammation mechanisms. These results are interesting because pre-clinical studies which combine both neurobehavioral and histological evaluations may improve the predictive value of animal models for clinical efficacy with novel neuroprotective agent [50]. In addition, the beneficial neuroprotective effects of EFX at an earliest stage of TBI might attenuate other deleterious effects such as anxiety disorders and memory disturbances which can appear several weeks later [84, 85]. Abbreviations TBITraumatic brain injury EFXEtifoxine GABAGamma-aminobutyric acid TSPOTranslocator protein 18 kDa kDaKilodalton CNSCentral nervous system CCIControlled cortical impact PBSPhosphate buffer saline PFAParaformaldehyde GFAPGlial fibrillary acidic protein CD68Cluster of differentiation 68 PHPotential of hydrogen TBSTris-buffered saline TBS-TTris-buffered saline-Triton NGSNormal goat serum FJBFluoro-jade B RPMRevolution per minute IL-1αInterleukin-1 alpha IL-1βInterleukin-1 beta IL-2Interleukin-2 IL-4Interleukin-4 IL-5Interleukin-5 IL-6Interleukin-6 IL-10Interleukin-10 IL-12Interleukin-12 TNF-αTumor necrosis factor alpha GM-CSFGranulocyte macrophage-colony stimulating factor MCP-1Monocyte chemotactic protein 1 MIP-1αMacrophage inflammatory protein 1 alpha SEMStandard error of the mean ANOVAAnalysis of variance RMRepeated measure LODLimit of detection ILSInflammatory load score LPSLipopolysaccharide Acknowledgements Not applicable. Funding Not applicable. Availability of data and materials Data supporting the conclusions are presented in the manuscript. Any further information are available to scientific communities upon direct contact to the authors. Authors’ contributions ESO designed the study, performed surgeries, performed behavioral studies and immunohistochemistry, performed statistical analyses, and wrote the manuscript. DG assisted with immunohistochemistry and performed all the image analyses. VR provided assistance with the experimental design and revised the manuscript. MV provided assistance with the experimental design and statistical analyses and revised the manuscript. All authors read and approved the final manuscript. Competing interests All authors are employees of Biocodex Pharmacology Department. Consent for publication Not applicable. Ethics approval and consent to participate Experiments were carried out in strict accordance with the European Community regulations for animal use in research (2010/63/EU directive) and all protocols were approved by the Local Ethical Committee (C2EA n°72). ==== Refs References 1. Pop V Badaut J A neurovascular perspective for long-term changes after brain trauma Transl Stroke Res 2011 2 533 545 10.1007/s12975-011-0126-9 22350620 2. Morganti-Kossmann MC Rancan M Stahel PF Kossmann T Inflammatory response in acute traumatic brain injury: a double-edged sword Curr Opin Crit Care 2002 8 101 105 10.1097/00075198-200204000-00002 12386508 3. Chen SF Hung TH Chen CC Lin KH Huang YN Tsai HC Lovastatin improves histological and functional outcomes and reduces inflammation after experimental traumatic brain injury Life Sci 2007 81 288 298 10.1016/j.lfs.2007.05.023 17612572 4. Kumar A Loane DJ Neuroinflammation after traumatic brain injury: opportunities for therapeutic intervention Brain Behav Immun 2012 26 1191 1201 10.1016/j.bbi.2012.06.008 22728326 5. Papadopoulos V Lecanu L Translocator protein (18 kDa) TSPO: an emerging therapeutic target in neurotrauma Exp Neurol 2009 219 53 57 10.1016/j.expneurol.2009.04.016 19409385 6. McIntosh TK Juhler M Wieloch T Novel pharmacologic strategies in the treatment of experimental traumatic brain injury: 1998 J Neurotrauma 1998 15 731 769 10.1089/neu.1998.15.731 9814632 7. Raghupathi R Cell death mechanisms following traumatic brain injury Brain Pathol 2004 14 215 222 10.1111/j.1750-3639.2004.tb00056.x 15193035 8. He J Hoffman SW Stein DG Allopregnanolone, a progesterone metabolite, enhances behavioral recovery and decreases neuronal loss after traumatic brain injury Restor Neurol Neurosci 2004 22 19 31 15096691 9. Sun DA Deshpande LS Sombati S Baranova A Wilson MS Hamm RJ Traumatic brain injury causes a long-lasting calcium (Ca2+)-plateau of elevated intracellular Ca levels and altered Ca2+ homeostatic mechanisms in hippocampal neurons surviving brain injury Eur J Neurosci 2008 27 1659 1672 10.1111/j.1460-9568.2008.06156.x 18371074 10. Schlichter R Rybalchenko V Poisbeau P Verleye M Gillardin J Modulation of GABAergic synaptic transmission by the non-benzodiazepine anxiolytic etifoxine Neuropharmacology 2000 39 1523 1535 10.1016/S0028-3908(99)00253-1 10854897 11. Verleye M Gillardin JM Effects of etifoxine on stress-induced hyperthermia, freezing behavior and colonic motor activation in rats Physiol Behav 2004 82 891 897 10.1016/S0031-9384(04)00313-0 15451655 12. Servant D Graziani PL Moyse D Parquet PJ Treatment of adjustment disorder with anxiety: efficacy and tolerance of etifoxine in a double-blind controlled study Encéphale 1998 24 569 574 9949940 13. Nguyen N Fakra E Pradel V Jouve E Alquier C Le Guern ME Efficacy of etifoxine compared to lorazepam monotherapy in the treatment of patients with adjustment disorders with anxiety: a double-blind controlled study in general practice Hum Psychopharmacol 2006 21 139 149 10.1002/hup.757 16625522 14. Stein DJ Etifoxine versus alprazolam for the treatment of adjustment disorder with anxiety: a randomized controlled trial Adv Ther 2015 32 57 68 10.1007/s12325-015-0176-6 25620535 15. Verleye M Schlichter R Gillardin JM Interactions of etifoxine with the chloride channel coupled to the GABA(A) receptor complex Neuroreport 1999 10 3207 3210 10.1097/00001756-199910190-00015 10574561 16. Verleye M Pansart Y Gillardin J Effects of etifoxine on ligand binding to GABA(A) receptors in rodents Neurosci Res 2002 44 167 172 10.1016/S0168-0102(02)00121-9 12354631 17. Hamon A Morel A Hue B Verleye M Gillardin JM The modulatory effects of the anxiolytic etifoxine on GABA(A) receptors are mediated by the beta subunit Neuropharmacology 2003 45 293 303 10.1016/S0028-3908(03)00187-4 12871647 18. Verleye M Akwa Y Liere P Ladurelle N Pianos A Eychenne B The anxiolytic etifoxine activates the peripheral benzodiazepine receptor and increases the neurosteroid levels in rat brain Pharmacol Biochem Behav 2005 82 712 720 10.1016/j.pbb.2005.11.013 16388839 19. Papadopoulos V Lecanu L Brown RC Han Z Yao ZX Peripheral-type benzodiazepine receptor in neurosteroid biosynthesis, neuropathology and neurological disorders Neuroscience 2006 138 749 756 10.1016/j.neuroscience.2005.05.063 16338086 20. Wolf L Bauer A Melchner D Hallof-Buestrich H Stoertebecker P Haen E Enhancing neurosteroid synthesis—relationship to the pharmacology of translocator protein (18 kDa) (TSPO) ligands and benzodiazepines Pharmacopsychiatry 2015 48 72 77 10.1055/s-0035-1558003 25654303 21. Rupprecht R Papadopoulos V Rammes G Baghai TC Fan J Akula N Translocator protein (18 kDa) (TSPO) as a therapeutic target for neurological and psychiatric disorders Nat Rev Drug Discov 2010 9 971 988 10.1038/nrd3295 21119734 22. Maeda J Higuchi M Inaji M Ji B Haneda E Okauchi T Phase-dependent roles of reactive microglia and astrocytes in nervous system injury as delineated by imaging of peripheral benzodiazepine receptor Brain Res 2007 1157 100 111 10.1016/j.brainres.2007.04.054 17540348 23. Chauveau F Boutin H Van CN Thominiaux C Hantraye P Rivron L In vivo imaging of neuroinflammation in the rodent brain with [11C]SSR180575, a novel indoleacetamide radioligand of the translocator protein (18 kDa) Eur J Nucl Med Mol Imaging 2011 38 509 514 10.1007/s00259-010-1628-5 20936410 24. Toth M Little P Arnberg F Haggkvist J Mulder J Halldin C Acute neuroinflammation in a clinically relevant focal cortical ischemic stroke model in rat: longitudinal positron emission tomography and immunofluorescent tracking Brain Struct Funct 2015 221 3 1279 1290 10.1007/s00429-014-0970-y 25601153 25. Wang Y Yue X Kiesewetter DO Niu G Teng G Chen X PET imaging of neuroinflammation in a rat traumatic brain injury model with radiolabeled TSPO ligand DPA-714 Eur J Nucl Med Mol Imaging 2014 41 1440 1449 10.1007/s00259-014-2727-5 24615467 26. Torres SR Frode TS Nardi GM Vita N Reeb R Ferrara P Anti-inflammatory effects of peripheral benzodiazepine receptor ligands in two mouse models of inflammation Eur J Pharmacol 2000 408 199 211 10.1016/S0014-2999(00)00760-3 11080527 27. Ferzaz B Brault E Bourliaud G Robert JP Poughon G Claustre Y SSR180575 (7-chloro-N, N,5-trimethyl-4-oxo-3-phenyl-3,5-dihydro-4H-pyridazino[4,5-b]indole-1 -acetamide), a peripheral benzodiazepine receptor ligand, promotes neuronal survival and repair J Pharmacol Exp Ther 2002 301 1067 1078 10.1124/jpet.301.3.1067 12023539 28. Ryu JK Choi HB McLarnon JG Peripheral benzodiazepine receptor ligand PK11195 reduces microglial activation and neuronal death in quinolinic acid-injected rat striatum Neurobiol Dis 2005 20 550 561 10.1016/j.nbd.2005.04.010 15916899 29. Veiga S Azcoitia I Garcia-Segura LM Ro5-4864, a peripheral benzodiazepine receptor ligand, reduces reactive gliosis and protects hippocampal hilar neurons from kainic acid excitotoxicity J Neurosci Res 2005 80 129 137 10.1002/jnr.20430 15696538 30. Girard C Liu S Cadepond F Adams D Lacroix C Verleye M Etifoxine improves peripheral nerve regeneration and functional recovery Proc Natl Acad Sci U S A 2008 105 20505 20510 10.1073/pnas.0811201106 19075249 31. Girard P Pansart Y Gillardin JM Preventive and curative effects of etifoxine in a rat model of brain oedema Clin Exp Pharmacol Physiol 2009 36 655 661 10.1111/j.1440-1681.2008.05127.x 19076164 32. Girard C Liu S Adams D Lacroix C Sineus M Boucher C Axonal regeneration and neuroinflammation: roles for the translocator protein 18 kDa J Neuroendocrinol 2012 24 71 81 10.1111/j.1365-2826.2011.02215.x 21951109 33. Lighthall JW Controlled cortical impact: a new experimental brain injury model J Neurotrauma 1988 5 1 15 10.1089/neu.1988.5.1 3193461 34. Dixon CE Clifton GL Lighthall JW Yaghmai AA Hayes RL A controlled cortical impact model of traumatic brain injury in the rat J Neurosci Methods 1991 39 253 262 10.1016/0165-0270(91)90104-8 1787745 35. Anderson GD Farin FM Bammler TK Beyer RP Swan AA Wilkerson HW The effect of progesterone dose on gene expression after traumatic brain injury J Neurotrauma 2011 28 1827 1843 10.1089/neu.2011.1911 21770760 36. Kim DH Ko IG Kim BK Kim TW Kim SE Shin MS Treadmill exercise inhibits traumatic brain injury-induced hippocampal apoptosis Physiol Behav 2010 101 660 665 10.1016/j.physbeh.2010.09.021 20888848 37. Haber M Abdel Baki SG Grin'kina NM Irizarry R Ershova A Orsi S Minocycline plus N-acetylcysteine synergize to modulate inflammation and prevent cognitive and memory deficits in a rat model of mild traumatic brain injury Exp Neurol 2013 249 169 177 10.1016/j.expneurol.2013.09.002 24036416 38. Briones TL Woods J Rogozinska M Decreased neuroinflammation and increased brain energy homeostasis following environmental enrichment after mild traumatic brain injury is associated with improvement in cognitive function Acta Neuropathol Commun 2013 1 57 10.1186/2051-5960-1-57 24252176 39. Robertson CS Garcia R Gaddam SS Grill RJ Cerami HC Tian TS Treatment of mild traumatic brain injury with an erythropoietin-mimetic peptide J Neurotrauma 2013 30 765 774 10.1089/neu.2012.2431 22827443 40. Schallert T Upchurch M Lobaugh N Farrar SB Spirduso WW Gilliam P Tactile extinction: distinguishing between sensorimotor and motor asymmetries in rats with unilateral nigrostriatal damage Pharmacol Biochem Behav 1982 16 455 462 10.1016/0091-3057(82)90452-X 7079281 41. Schallert T Upchurch M Wilcox RE Vaughn DM Posture-independent sensorimotor analysis of inter-hemispheric receptor asymmetries in neostriatum Pharmacol Biochem Behav 1983 18 753 759 10.1016/0091-3057(83)90019-9 6407036 42. Freret T Valable S Chazalviel L Saulnier R Mackenzie ET Petit E Delayed administration of deferoxamine reduces brain damage and promotes functional recovery after transient focal cerebral ischemia in the rat Eur J Neurosci 2006 23 1757 1765 10.1111/j.1460-9568.2006.04699.x 16623832 43. Schallert T Woodlee MT Fleming SM Krieglstein J Disentangling multiple types of recovery from brain injury Pharmacology of cerebral ischemia 2002 Stuttgart Medpharm Scientific Publishers 201 216 44. Fleming SM Salcedo J Fernagut PO Rockenstein E Masliah E Levine MS Early and progressive sensorimotor anomalies in mice overexpressing wild-type human alpha-synuclein J Neurosci 2004 24 9434 9440 10.1523/JNEUROSCI.3080-04.2004 15496679 45. Kaufman NA Beare JE Tan AA Vitek MP McKenna SE Hoane MR COG1410, an apolipoprotein E-based peptide, improves cognitive performance and reduces cortical loss following moderate fluid percussion injury in the rat Behav Brain Res 2010 214 395 401 10.1016/j.bbr.2010.06.017 20600347 46. Schallert T Fleming SM Leasure JL Tillerson JL Bland ST CNS plasticity and assessment of forelimb sensorimotor outcome in unilateral rat models of stroke, cortical ablation, parkinsonism and spinal cord injury Neuropharmacology 2000 39 777 787 10.1016/S0028-3908(00)00005-8 10699444 47. Woodlee MT Asseo-Garcia AM Zhao X Liu SJ Jones TA Schallert T Testing forelimb placing “across the midline” reveals distinct, lesion-dependent patterns of recovery in rats Exp Neurol 2005 191 310 317 10.1016/j.expneurol.2004.09.005 15649486 48. Schmued LC Hopkins KJ Fluoro-jade B: a high affinity fluorescent marker for the localization of neuronal degeneration Brain Res 2000 874 123 130 10.1016/S0006-8993(00)02513-0 10960596 49. Harting MT Jimenez F Adams SD Mercer DW Cox CS Jr Acute, regional inflammatory response after traumatic brain injury: implications for cellular therapy Surgery 2008 144 803 813 10.1016/j.surg.2008.05.017 19081024 50. Morganti-Kossmann MC Satgunaseelan L Bye N Kossmann T Modulation of immune response by head injury Injury 2007 38 1392 1400 10.1016/j.injury.2007.10.005 18048036 51. Fox GB Fan L Levasseur RA Faden AI Sustained sensory/motor and cognitive deficits with neuronal apoptosis following controlled cortical impact brain injury in the mouse J Neurotrauma 1998 15 599 614 10.1089/neu.1998.15.599 9726259 52. Streit WJ The role of microglia in brain injury Neurotoxicology 1996 17 671 678 9086488 53. Mantovani A Sica A Sozzani S Allavena P Vecchi A Locati M The chemokine system in diverse forms of macrophage activation and polarization Trends Immunol 2004 25 677 686 10.1016/j.it.2004.09.015 15530839 54. Hsieh CL Kim CC Ryba BE Niemi EC Bando JK Locksley RM Traumatic brain injury induces macrophage subsets in the brain Eur J Immunol 2013 43 2010 2022 10.1002/eji.201243084 23630120 55. Perego C Fumagalli S De Simoni MG Three-dimensional confocal analysis of microglia/macrophage markers of polarization in experimental brain injury J Vis Exp 2013 24056862 56. Hernandez-Ontiveros DG Tajiri N Acosta S Giunta B Tan J Borlongan CV Microglia activation as a biomarker for traumatic brain injury Front Neurol 2013 4 30 10.3389/fneur.2013.00030 23531681 57. Ansari MA Temporal profile of M1 and M2 responses in the hippocampus following early 24 h of neurotrauma J Neurol Sci 2015 357 41 49 10.1016/j.jns.2015.06.062 26148932 58. Turtzo LC Lescher J Janes L Dean DD Budde MD Frank JA Macrophagic and microglial responses after focal traumatic brain injury in the female rat J Neuroinflammation 2014 11 82 10.1186/1742-2094-11-82 24761998 59. Wang G Zhang J Hu X Zhang L Mao L Jiang X Microglia/macrophage polarization dynamics in white matter after traumatic brain injury J Cereb Blood Flow Metab 2013 33 1864 1874 10.1038/jcbfm.2013.146 23942366 60. Rodriguez-Rodriguez A Egea-Guerrero JJ Murillo-Cabezas F Carrillo-Vico A Oxidative stress in traumatic brain injury Curr Med Chem 2014 21 1201 1211 10.2174/0929867321666131217153310 24350853 61. Loane DJ Kumar A Microglia in the TBI brain: the good, the bad, and the dysregulated Exp Neurol 2016 275 Pt 3 316 327 10.1016/j.expneurol.2015.08.018 26342753 62. Goodman JC Robertson CS Grossman RG Narayan RK Elevation of tumor necrosis factor in head injury J Neuroimmunol 1990 30 213 217 10.1016/0165-5728(90)90105-V 2229409 63. Helmy A De Simoni MG Guilfoyle MR Carpenter KL Hutchinson PJ Cytokines and innate inflammation in the pathogenesis of human traumatic brain injury Prog Neurobiol 2011 95 352 372 10.1016/j.pneurobio.2011.09.003 21939729 64. Perez-Polo JR Rea HC Johnson KM Parsley MA Unabia GC Xu G Inflammatory consequences in a rodent model of mild traumatic brain injury J Neurotrauma 2013 30 727 740 10.1089/neu.2012.2650 23360201 65. Semple BD Bye N Rancan M Ziebell JM Morganti-Kossmann MC Role of CCL2 (MCP-1) in traumatic brain injury (TBI): evidence from severe TBI patients and CCL2-/- mice J Cereb Blood Flow Metab 2010 30 769 782 10.1038/jcbfm.2009.262 20029451 66. Mukherjee S Katki K Arisi GM Foresti ML Shapiro LA Early TBI-induced cytokine alterations are similarly detected by two distinct methods of multiplex assay Front Mol Neurosci 2011 4 21 10.3389/fnmol.2011.00021 21954376 67. Dalgard CL Cole JT Kean WS Lucky JJ Sukumar G McMullen DC The cytokine temporal profile in rat cortex after controlled cortical impact Front Mol Neurosci 2012 5 6 10.3389/fnmol.2012.00006 22291617 68. Lu KT Wang YW Wo YY Yang YL Extracellular signal-regulated kinase-mediated IL-1-induced cortical neuron damage during traumatic brain injury Neurosci Lett 2005 386 40 45 10.1016/j.neulet.2005.05.057 16024175 69. Pu B Xue Y Wang Q Hua C Li X Dextromethorphan provides neuroprotection via anti-inflammatory and anti-excitotoxicity effects in the cortex following traumatic brain injury Mol Med Rep 2015 12 3704 3710 26004757 70. Taupin V Toulmond S Serrano A Benavides J Zavala F Increase in IL-6, IL-1 and TNF levels in rat brain following traumatic lesion. Influence of pre- and post-traumatic treatment with Ro5 4864, a peripheral-type (p site) benzodiazepine ligand J Neuroimmunol 1993 42 177 185 10.1016/0165-5728(93)90008-M 8429103 71. Holmin S Mathiesen T Intracerebral administration of interleukin-1beta and induction of inflammation, apoptosis, and vasogenic edema J Neurosurg 2000 92 108 120 10.3171/jns.2000.92.1.0108 10616089 72. Venters HD Dantzer R Kelley KW A new concept in neurodegeneration: TNFalpha is a silencer of survival signals Trends Neurosci 2000 23 175 180 10.1016/S0166-2236(99)01533-7 10717677 73. Patel HC Boutin H Allan SM Interleukin-1 in the brain: mechanisms of action in acute neurodegeneration Ann N Y Acad Sci 2003 992 39 47 10.1111/j.1749-6632.2003.tb03136.x 12794045 74. Raivich G Liu ZQ Kloss CU Labow M Bluethmann H Bohatschek M Cytotoxic potential of proinflammatory cytokines: combined deletion of TNF receptors TNFR1 and TNFR2 prevents motoneuron cell death after facial axotomy in adult mouse Exp Neurol 2002 178 186 193 10.1006/exnr.2002.8024 12504878 75. Stover JF Schoning B Beyer TF Woiciechowsky C Unterberg AW Temporal profile of cerebrospinal fluid glutamate, interleukin-6, and tumor necrosis factor-alpha in relation to brain edema and contusion following controlled cortical impact injury in rats Neurosci Lett 2000 288 25 28 10.1016/S0304-3940(00)01187-3 10869807 76. Tehranian R Andell-Jonsson S Beni SM Yatsiv I Shohami E Bartfai T Improved recovery and delayed cytokine induction after closed head injury in mice with central overexpression of the secreted isoform of the interleukin-1 receptor antagonist J Neurotrauma 2002 19 939 951 10.1089/089771502320317096 12225654 77. Baratz R Tweedie D Wang JY Rubovitch V Luo W Hoffer BJ Transiently lowering tumor necrosis factor-α synthesis ameliorates neuronal cell loss and cognitive impairments induce by minimal traumatic brain injury in mice J Neuroinflammation. 2015 7 12 45 78. d'Avila JC Lam TI Bingham D Shi J Won SJ Kauppinen TM Microglial activation induced by brain trauma is suppressed by post-injury treatment with a PARP inhibitor J Neuroinflammation 2012 9 31 22335939 79. Kim KS Wass CA Cross AS Opal SM Modulation of blood-brain barrier permeability by tumor necrosis factor and antibody to tumor necrosis factor in the rat Lymphokine Cytokine Res 1992 11 293 298 1477182 80. Touzani O Boutin H Chuquet J Rothwell N Potential mechanisms of interleukin-1 involvement in cerebral ischaemia J Neuroimmunol 1999 100 203 215 10.1016/S0165-5728(99)00202-7 10695731 81. Zhang X Chen Y Jenkins LW Kochanek PM Clark RS Bench-to-bedside review: apoptosis/programmed cell death triggered by traumatic brain injury Crit Care 2005 9 66 75 10.1186/cc2950 15693986 82. Ziebell JM Morganti-Kossmann MC Involvement of pro- and anti-inflammatory cytokines and chemokines in the pathophysiology of traumatic brain injury Neurotherapeutics 2010 7 22 30 10.1016/j.nurt.2009.10.016 20129494 83. Woodcock T Morganti-Kossmann MC The role of markers of inflammation in traumatic brain injury Front Neurol 2013 4 18 10.3389/fneur.2013.00018 23459929 84. Nylen K Ost M Csajbok LZ Nilsson I Blennow K Nellgard B Increased serum-GFAP in patients with severe traumatic brain injury is related to outcome J Neurol Sci 2006 240 85 91 10.1016/j.jns.2005.09.007 16266720 85. Ekmark-Lewen S Lewen A Israelsson C Li GL Farooque M Olsson Y Vimentin and GFAP responses in astrocytes after contusion trauma to the murine brain Restor Neurol Neurosci 2010 28 311 321 20479526 86. Plog BA Dashnaw ML Hitomi E Peng W Liao Y Lou N Biomarkers of traumatic injury are transported from brain to blood via the glymphatic system J Neurosci 2015 35 518 526 10.1523/JNEUROSCI.3742-14.2015 25589747 87. Johnstone M Gearing AJ Miller KM A central role for astrocytes in the inflammatory response to beta-amyloid; chemokines, cytokines and reactive oxygen species are produced J Neuroimmunol 1999 93 182 193 10.1016/S0165-5728(98)00226-4 10378882 88. Shafer RA Murphy S Activated astrocytes induce nitric oxide synthase-2 in cerebral endothelium via tumor necrosis factor alpha Glia 1997 21 370 379 10.1002/(SICI)1098-1136(199712)21:4<370::AID-GLIA4>3.0.CO;2-7 9419012 89. Tanaka M Sotomatsu A Yoshida T Hirai S Nishida A Detection of superoxide production by activated microglia using a sensitive and specific chemiluminescence assay and microglia-mediated PC12h cell death J Neurochem 1994 63 266 270 10.1046/j.1471-4159.1994.63010266.x 8207432 90. Wei LC Shi M Chen LW Cao R Zhang P Chan YS Nestin-containing cells express glial fibrillary acidic protein in the proliferative regions of central nervous system of postnatal developing and adult mice Brain Res Dev Brain Res 2002 139 9 17 10.1016/S0165-3806(02)00509-6 12414089 91. Ravikumar B Crawford D Dellovade T Savinainen A Graham D Liere P Differential efficacy of the TSPO ligands etifoxine and XBD-173 in two rodent models of Multiple Sclerosis Neuropharmacology 2016 108 229 237 10.1016/j.neuropharm.2016.03.053 27039042 92. Daugherty DJ Selvaraj V Chechneva OV Liu XB Pleasure DE Deng W A TSPO ligand is protective in a mouse model of multiple sclerosis EMBO Mol Med 2013 5 891 903 10.1002/emmm.201202124 23681668 93. do Rego JL Vaudry D Vaudry H The non-benzodiazepine anxiolytic drug etifoxine causes a rapid, receptor-independent stimulation of neurosteroid biosynthesis PLoS One 2015 10 e0120473 10.1371/journal.pone.0120473 25785994 94. Roof RL Duvdevani R Braswell L Stein DG Progesterone facilitates cognitive recovery and reduces secondary neuronal loss caused by cortical contusion injury in male rats Exp Neurol 1994 129 64 69 10.1006/exnr.1994.1147 7925843 95. Galani R Hoffman SW Stein DG Effects of the duration of progesterone treatment on the resolution of cerebral edema induced by cortical contusions in rats Restor Neurol Neurosci 2001 18 161 166 11847439 96. Grossman KJ Goss CW Stein DG Effects of progesterone on the inflammatory response to brain injury in the rat Brain Res 2004 1008 29 39 10.1016/j.brainres.2004.02.022 15081379 97. Djebaili M Guo Q Pettus EH Hoffman SW Stein DG The neurosteroids progesterone and allopregnanolone reduce cell death, gliosis, and functional deficits after traumatic brain injury in rats J Neurotrauma 2005 22 106 118 10.1089/neu.2005.22.106 15665606 98. VanLandingham JW Cutler SM Virmani S Hoffman SW Covey DF Krishnan K The enantiomer of progesterone acts as a molecular neuroprotectant after traumatic brain injury Neuropharmacology 2006 51 1078 1085 10.1016/j.neuropharm.2006.07.015 16926035 99. Pettus EH Wright DW Stein DG Hoffman SW Progesterone treatment inhibits the inflammatory agents that accompany traumatic brain injury Brain Res 2005 1049 112 119 10.1016/j.brainres.2005.05.004 15932748 100. Cutler SM VanLandingham JW Murphy AZ Stein DG Slow-release and injected progesterone treatments enhance acute recovery after traumatic brain injury Pharmacol Biochem Behav 2006 84 420 428 10.1016/j.pbb.2006.05.029 16870241 101. He J Evans CO Hoffman SW Oyesiku NM Stein DG Progesterone and allopregnanolone reduce inflammatory cytokines after traumatic brain injury Exp Neurol 2004 189 404 412 10.1016/j.expneurol.2004.06.008 15380490 102. Almeida-Suhett CP Prager EM Pidoplichko V Figueiredo TH Marini AM Li Z Reduced GABAergic inhibition in the basolateral amygdala and the development of anxiety-like behaviors after mild traumatic brain injury PLoS One 2014 9 e102627 10.1371/journal.pone.0102627 25047645 103. Khateb A Fort P Williams S Serafin M Muhlethaler M Jones BE GABAergic input to cholinergic nucleus basalis neurons Neuroscience 1998 86 937 947 10.1016/S0306-4522(98)00094-3 9692729 104. Romanides AJ Duffy P Kalivas PW Glutamatergic and dopaminergic afferents to the prefrontal cortex regulate spatial working memory in rats Neuroscience 1999 92 97 106 10.1016/S0306-4522(98)00747-7 10392833 105. Akk G Covey DF Evers AS Steinbach JH Zorumski CF Mennerick S Mechanisms of neurosteroid interactions with GABA(A) receptors Pharmacol Ther 2007 116 35 57 10.1016/j.pharmthera.2007.03.004 17524487 106. Schumacher M Mattern C Ghoumari A Oudinet JP Liere P Labombarda F Revisiting the roles of progesterone and allopregnanolone in the nervous system: resurgence of the progesterone receptors Prog Neurobiol 2014 113 6 39 10.1016/j.pneurobio.2013.09.004 24172649

PMC005xxxxxx/PMC5002208.txt

==== Front BMC Complement Altern MedBMC Complement Altern MedBMC Complementary and Alternative Medicine1472-6882BioMed Central London 130810.1186/s12906-016-1308-5Research ArticleThe association between arterial stiffness and tongue manifestations of blood stasis in patients with type 2 diabetes Hsu Po-Chi 12Huang Yu-Chuen +886-4-22053366yuchuen@mail.cmu.edu.tw 34Chiang John Y. 56Chang Hen-Hong 37Liao Pei-Yung +886-4-723-859548795@cch.org.tw 8Lo Lun-Chien +886-4-723-8595126478@cch.org.tw 291 Graduate Institute of Chinese Medicine, China Medical University, Taichung, Taiwan 2 Department of Chinese Medicine, Changhua Christian Hospital, No.135 Nanxiao St., Changhua, Taiwan 3 School of Chinese Medicine, China Medical University, No. 91, Hsueh-Shih Rd., Taichung, Taiwan 4 Department of Medical Research, China Medical University Hospital, Taichung, Taiwan 5 Department of Computer Science and Engineering, National Sun Yat-sen University, Kaohsiung, Taiwan 6 Department of Healthcare Administration and Medical Informatics, Kaohsiung Medical University, Kaohsiung, Taiwan 7 Research Center for Chinese Medicine and Acupuncture, China Medical University, Taichung, Taiwan 8 Department of Endocrinology, Changhua Christian Hospital, No.135 Nanxiao St., Changhua, Taiwan 9 Graduate Institute of Statistical and informational Science, National Changhua University of Education, Changhua, Taiwan 27 8 2016 27 8 2016 2016 16 1 32430 9 2015 20 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Diabetes mellitus (DM) is a hypercoagulable state and is associated with highly increased risk of vascular complications. In the theory of traditional Chinese medicine (TCM), these vascular complications are classified as blood stasis. Diagnosis of the tongue plays an important role in TCM; a bluish tongue, petechiae, and engorged sublingual collateral vessels are manifestations of blood stasis. This study aimed to characterize the tongue manifestations of blood stasis and derive a relationship between blood stasis and vascular disorders in patients with type 2 DM. Method We conducted a cross-sectional study of 140 patients with type 2 DM, and compared demography, laboratory, physical examination, ankle brachial index(ABI), brachial-ankle pulse wave velocity (ba-PWV), and tongue manifestation datas. An automatic tongue diagnosis system was used to capture tongue images and characterize clinical tongue manifestations. Results A bluish or petechiae tongue was assoicated with a significant decrease in high-density lipoprotein level, and bluish tongue was associated with significant increase in blood triglyceride in patients with type 2 DM. On assessing arterial stiffness, patients with a petechiae tongue had a higher ba-PWV for both sides (L:1938.41 ± 469.54 cm/sec v.s.1723.99 ± 302.16, p = 0.02; R:1937.28 ± 405.55 v.s.1741.99 ± 325.82, p = 0.03). Conclusion Blood stasis, particularly a tongue with petechiae, may be associated with arterial stiffness in patients with type 2 DM. Furthermore, tongue diagnosis could detect blood stasis relevant to DM and could serve as a feasible predictor for DM. Keywords Traditional Chinese medicine (TCM)Diabetes mellitus (DM)Tongue diagnosisBlood stasisArterial stiffness,and brachial-ankle pulse wave velocity (ba-PWV)Taiwan Ministry of Health and Welfare Clinical Trial and Research Center of ExcellenceMOHW105-TDU-B-212-133019Huang Yu-Chuen issue-copyright-statement© The Author(s) 2016 ==== Body Background As the global population increases and ages, diabetes mellitus (DM) has become a major public health concern worldwide [1–3]. DM increases the risk for disability and premature death, and imparts a substantial socioeconomic burden due to the micro- and macro-vascular complications [3]. The long-term micro-vascular complications of DM include retinopathy [4], nephropathy, neuropathy and macro-vascular complications. DM is a hypercoagulable state and is associated with an increased risk of ischemic events; it is also associated with accelerated atherothrombosis [5, 6]. Arterial stiffness is closely related to the progression of DM complications [7]. Consequently, patients with DM have shown a 2- to 4-fold greater risk for coronary artery disease and cerebrovascular disease than those without DM [8]. Thus, vascular complications and arterial stiffness due to poor blood circulation should be closely monitored in patients with DM. Blood stasis is one of the most important pathological concepts in traditional Chinese medicine (TCM) [9]. Blood stasis is characterized as a disorder of blood circulation with hallmarks including extravagated or sluggish blood circulation and viscous or congested blood; all of these hallmarks may contribute to various disease pathologies [10]. Many diseases lead to blood stasis, such as cardiovascular disease, cerebral vascular accidents, and DM [11]. Blood stasis is often accompanied by characteristic symptoms, such as pain in a fixed position, a dark-purple colored face, infraorbital darkness, a bluish tongue, an engorged sublingual varicosis, petechiae tongue, or an astringent pulse [12]. Tongue diagnosis is important in TCM [13]. The tongue is connected to the internal organs through the meridians; thus the conditions of the organs, qi, blood, and bodily fluids, as well as the degree and progression of disease, are manifested in the tongue [14]. Clinically, practitioners observe tongue characteristics, such as tongue color and shape, fur color and thickness, and the amount of saliva, to deduce the primary ailment of a patient [15]. A bluish tongue, petechiae and engorged sublingual collateral vessels are potential tongue manifestations of blood stasis [16]. Tongue diagnosis is helpful in detecting blood stasis of rheumatoid arthritis (RA) and could serve as a feasible predictor of RA [17]. However, to the best of our knowledge, no study has focused on tongue diagnosis in patients with type 2 DM, despite the theoretical and clinical applications. This study aimed to investigate the tongue characteristics of and relationship between blood stasis and vascular disorders in patients with type 2 DM. Methods Patients We conducted across-sectional study and recruited patients with type 2 DM from the Department of Chinese Medicine, Changhua Christian Hospital, between January 2012 and December 2013. We excluded patients with type 1 DM or type 2 DM who had cancer. One hundred and forty eligible patients with type 2 DM were enrolled. The purpose, procedures, potential risks, and benefits of the study were thoroughly explained to the patients. The personal details and photographs of the patients were kept confidential, and all participants signed consent for publication. This study was approved by the Institutional Review Board of Changhua Christian Hospital (IRB#:111106). Data collection Patient metadata were collected (i.e., sex, age, weight, height, history of DM, and any micro-vascular complications). Physical examinations included blood pressure, body mass index(BMI), waistline, hipline, foot examination, ankle brachial index (ABI), and brachial-ankle pulse wave velocity (ba-PWV). Ba-PWV is a direct measurement of aortic stiffness and is the gold standard of arterial stiffness measurements [18]. ABI is a non-invasive method that assesses the patency of peripheral occlusive arterial disease [19]. Routine biological blood tests included hemoglobin A1c (HbA1c), fasting blood-glucose level, cholesterol, triglyceride (TG), high-density lipoprotein (HDL), low-density lipoprotein (LDL), creatinine (Cr), glutamate oxaloacetate transaminase (GOT) and glutamate pyruvate transaminase (GPT) levels. Tongue photographs and procedures An automatic tongue diagnosis system was developed to capture tongue images. The consistency and stability of image capturing relies on the brightness and color calibration to compensate for variations, such as intensity and color temperature of light source as well as imaging hardware [15, 16]. Analysis of tongue images was conducted by Chinese medical physicians who had 3–5 years of clinical experience in the Chinese medicine department of Changhua Christian Hospital (CCH), Taiwan. The physicians attended regularly meetings over the past two years and examined over 1000 tongue images from CCH outpatients. Tongue images were identified according to nine primary tongue features: tongue body shape (i.e., small, median, enlarged), tongue body color (i.e., pale, pink red, red), tongue characteristics (i.e., spots, petechiae, teeth-marks, fissures), bluish tongue (i.e., yes, no), fur color (i.e., white, yellow), fur thickness (i.e., peeled, thin, thick), saliva (i.e., dry mouth, normal, wet mouth), and sublingual collateral vessels (i.e., normal, engorged) [15]. Tongue diagnosis followed a standardized protocol (Fig. 1).Fig. 1 Standard processing flowchart for tongue diagnosis Data analysis The statistical analysis of datawas performed with IBM SPSS Statistics 19 (IBM Co., New York, NC, USA). After determining if that data was normal (Kolmogorov-Smirnov test), Student’s t tests were used to determine differences between continuous variables. Regression models were used to analyze the relationship between the change in ba-PWV and patients possessing tongue manifestations of blood stasis. A p-value less than 0.05 was considered statistically significant. Result The characteristics of the 140 patients with type 2 DM were 40 % males (56 patients) and 60 % females (84 patients), the average age was 62.95 ± 11.10 years old (range of 33 to 87 years); the mean BMI was 25.94 ± 3.93 kg/m2 , and mean HbA1C was 7.03 ± 1.08 % (Table 1). The duration of DM course in the patients was 2 to 35 years, with an average of 13.57 ± 8.54 years. Twenty-eight (20.0 %) of study patientshad retinopathy; fifty patients (35.7 %) had nephropathy and seventeen patinets (12.1 %) had neuropathy.Table 1 Characteristics of 140 patients with type 2 diabetes Characteristics Age (years, mean ± SD) 62.95 ± 11.10 Gender  Male (n, %) (56, 40.0 %)  Female (n, %) (84, 60.0 %) BMI (kg/m2, mean ± SD) 25.94 ± 3.93 DM history (years, mean ± SD) 13.57 ± 8.54  ≤ 10 years (n, %) (39, 38.6 %)  10 < age ≤ 20 years (n, %) (42, 41.6 %)  > 20 years (n, %) (20, 19.8 %) HbA1c (mean ± SD) 7.03 ± 1.08  ≤ 7 (n, %) (72, 54.6 %)  7 < HbA1c ≤ 8 (n, %) (42, 31.8 %)  > 8 (n, %) (18, 13.6 %) Micro-vascular complication  Retinopathy (n, %) (28, 20.0 %)  Nephropathy (n, %) (50, 35.7 %)  Neuropathy (n, %) (17, 12.1 %) Tongue inspection refers to the visual examination of tongue body shape, tongue color, fur color, and fur thickness, as well as other characteristics. The tongue body shape was classified as median (n = 123, 87.9 %), enlarged (n = 15, 10.7 %), and small (n = 2, 2.4 %) (Table 2). The tongue body color was pink red (n = 86, 65.2 %), red (n = 31, 23.5 %), or pale (n = 15 ,11.4 %). Other characteristics observed on the tongue surface included teeth-marks (n = 80, 57.1 %), fissures (n = 33, 23.6 %), spots (n = 32, 22.9 %) and petechiae (n = 22,15.7 %). A bluish toned tongue may indicate a problem with blood circulation (n = 34, 24.3 %). The classifications of fur color and thickness were white (n = 97, 73.5 %) or yellow (n = 35, 26.5 %) and thin fur (n = 64, 47.1 %), thick fur (n = 64, 47.1 %), or peeled fur (n = 8, 5.9 %), respectively. The amount of saliva observed was normal (n = 110, 80.9 %), dry (n = 20, 14.7 %), or wet (n = 6, 4.4 %). There were 99 patients (73.3 %) with engorged sublingual collateral vessels.Table 2 Tongue manifestations of patients with type 2 diabetes Tongue manifestations Types N (%) Tongue body median 123 (87.9 %) enlarged 15 (10.7 %) small 2 (2.4 %) Body color pink red 86 (65.2 %) red 31 (23.5 %) pale 15 (11.4 %) Tongue characteristics spots 32 (22.9 %) petechiae 22 (15.7 %) teeth-marks 80 (57.1 %) fissures 33 (23.6 %) Bluish tongue no 106 (75.7 %) yes 34 (24.3 %) Fur color white 97 (73.5 %) yellow 35 (26.5 %) Fur thickness thin 64 (47.1 %) thick 64 (47.1 %) peeled 8 (5.9 %) Saliva normal 110 (80.9 %) dry mouth 20 (14.7 %) wet mouth 6 (4.4 %) Sublingual collateral vessels engorged 99 (73.3 %) normal 36 (26.7 %) According to the TCM theory, a bluish tongue, petechiae, and engorged sublingual collateral vessels are potential manifestations of blood stasis (Fig. 2). Therefore, the physical examinations and laboratory data were further examined to address the relationship between vascular disorders and blood stasis related to tongue characteristics (i.e., bluish tongue, petechiae, or engorged sublingual collateral vessels; Table 3). A bluish tongue was correlated with significant decrease in HDL (p = 0.03) and a significant increase in TG (p = 0.04) in the lipid profile (Table 3). Interestingly, both the left- and right-side ba-PWV (L: 1938.41 ± 469.54 cm/sec v.s. 1723.99 ± 302.16 cm/sec, p = 0.02; R: 1937.28 ± 405.55 cm/sec v.s. 1741.99 ± 325.82 cm/sec, p = 0.03) were significantly higher in patients with a petechiae tongue than in patients with type 2 DM.Fig. 2 Representative images of tongue manifestations: (a) normal tongue; (b) petechiae tongue; (c) bluish tongue, and; (d) engorged sublingual collateral vessels Table 3 Comparison of physical examinations and laboratory data of patients with type 2 diabetes with and without tongue manifestations of blood stasis Variables Bluish tongue p-value Petechiae p-value Engorged sublingual collateral vessels p-value Without (n = 106) With (n = 34) Without (n = 118) With (n = 22) Without (n = 36) With (n = 99) BMI(kg/m2) 25.53 ± 3.56 27.01 ± 4.69 0.09 25.85 ± 4.10 26.30 ± 3.23 0.65 26.04 ± 3.25 25.96 ± 4.25 0.94 Waistline(cm) 87.10 ± 9.96 89.50 ± 11.36 0.33 87.13 ± 10.26 89.85 ± 10.55 0.30 87.87 ± 7.90 87.76 ± 11.17 0.97 Hipline(cm) 95.86 ± 7.75 97.84 ± 9.57 0.31 96.33 ± 8.70 96.51 ± 6.49 0.93 96.36 ± 4.43 96.59 ± 9.22 0.88 Systolic pressure(mmHg) 122.44 ± 13.84 129.28 ± 13.39 0.06 123.04 ± 13.96 131.50 ± 11.98 0.10 121.61 ± 13.60 124.18 ± 14.04 0.45 Diastolic pressure(mmHg) 71.44 ± 4.04 73.72 ± 4.86 0.04 71.61 ± 4.24 74.75 ± 3.81 0.05 71.87 ± 4.71 71.79 ± 4.18 0.94 Foot examination score 1.49 ± 1.62 1.26 ± 1.18 0.53 1.41 ± 1.51 1.50 ± 1.54 0.81 1.21 ± 1.35 1.46 ± 1.53 0.51 ABI(L) 1.10 ± 0.11 1.12 ± 0.09 0.36 1.10 ± 0.11 1.15 ± 0.09 0.07 1.06 ± 0.15 1.12 ± 0.09 0.05 ABI(R) 1.13 ± 0.09 1.12 ± 0.12 0.66 1.12 ± 0.11 1.15 ± 0.06 0.17 1.10 ± 0.13 1.13 ± 0.07 0.28 ba-PWV(L)(cm/sec) 1743.10 ± 360.03 1817.88 ± 316.40 0.36 1723.99 ± 302.16 1938.41 ± 469.54 0.02 1725.53 ± 368.85 1761.38 ± 348.82 0.71 ba-PWV(R)(cm/sec) 1760.39 ± 366.31 1831.88 ± 305.58 0.38 1741.99 ± 325.82 1937.28 ± 405.55 0.03 1710.88 ± 383.36 1787.58 ± 346.15 0.43 HbA1c(%) 7.06 ± 1.11 6.94 ± 1.01 0.59 7.05 ± 1.08 6.94 ± 1.13 0.66 7.16 ± 1.33 6.97 ± 0.97 0.38 AC sugar(mg/dl) 137.05 ± 34.33 136.29 ± 25.85 0.91 137.33 ± 33.33 134.41 ± 27.10 0.70 139.09 ± 44.13 135.90 ± 27.37 0.69 Cholesterol(mg/dl) 168.97 ± 34.17 168.03 ± 34.90 0.89 170.01 ± 34.69 162.09 ± 31.57 0.32 168.97 ± 37.21 169.10 ± 33.33 0.98 TG(mg/dl) 116.70 ± 62.92 145.47 ± 86.96 0.04 118.84 ± 67.86 149.95 ± 79.17 0.06 139.74 ± 76.76 119.10 ± 69.06 0.14 HDL(mg/dl) 50.17 ± 13.88 44.32 ± 11.97 0.03 49.90 ± 14.19 42.59 ± 7.88 0.02 47.83 ± 12.89 49.45 ± 14.09 0.55 LDL(mg/dl) 96.05 ± 27.10 91.19 ± 24.12 0.35 95.68 ± 26.37 90.50 ± 26.72 0.40 98.17 ± 29.54 93.59 ± 25.37 0.38 Cr(mg/dl) 0.93 ± 0.50 0.99 ± 0.43 0.56 0.94 ± 0.50 0.96 ± 0.40 0.89 0.90 ± 0.30 0.97 ± 0.54 0.47 GOT(U/L) 27.88 ± 15.69 32.03 ± 21.16 0.30 28.66 ± 17.61 30.32 ± 15.62 0.68 29.22 ± 19.24 28.92 ± 16.99 0.93 GPT(U/L) 27.66 ± 15.76 33.59 ± 31.47 0.30 28.76 ± 20.72 31.09 ± 21.85 0.63 29.79 ± 19.02 29.30 ± 21.90 0.91 Microalbumin(mg/day) 286.79 ± 982.78 374.99 ± 1094.05 0.67 352.04 ± 1096.11 95.14 ± 205.03 0.03 244.97 ± 731.47 339.18 ± 1111.65 0.67 Values represented as mean ± SD. p-values performed by independent t test HbA1c hemoglobin A1c, AC sugar fasting blood-glucose level, TG triglyceride, HDL High-density lipoprotein, LDL low-density lipoprotein, Cr Creatinine, GOT glutamate oxaloacetate transaminase, GPT glutamate pyruvate transaminase, BMI body mass index, ABI ankle brachial index, ba-PWV brachial-ankle pulse wave velocity Discussion The core of assessment in Chinese medicine is “pattern identification/syndrome differentiation and treatment” based on inspection, listening and smelling examinations, inquiry, and palpation. Inspection is the most important of the four assessments, and tongue assessment is a crucial part of observation. Tongue appearance is a crucial indicator of physiological and pathological changes to the internal organs [19]. Studies have shown that tongue diagnosis plays an important role in clinical prognosis of RA and DM [15, 16, 20–22]. To the best of our knowledge, this is the first attempt to apply TCM tongue diagnosis to the survey of patients with type 2 DM. Tongue inspection refers to the shape, color, and fur color, and fur thickness, as well as other characteristics [23]. In patients with DM, buccal alterations can be easily observed with adequate glycemic control. Dry mouth is generally associated with decreased saliva production and is present in 10 to 30 % patients with DM; in these patients, a coated tongue is also observed [24]. In TCM, diabetes-related symptoms are referred to as “Xiaoke”, which means increased thirst (or polydipsia), since as long as 2000 years [25]. Furthermore, we show that 47.1 % of patients possessed a coated tongue (i.e., thick fur). According to the TCM theory, tongue fur indicates the Yang organs, especially the digestive system. Thick fur is usually associated with phlegm-dampness and patterns of blood stasis [26]. Thus, understanding and interpreting these tongue manifestations of DM by TCM are important for both in theoretical and clinical applications. Pulse wave velocity (PWV) is a noninvasive clinical index of arterial stiffness. Arterial pulse wave velocity reflects the stiffness of arteries, and serves as an indicator of atherosclerosis [27, 28]. Arterial stiffness is an age-related process that is present in numerous diseases, including DM. The PWV of patients with DM is higher than that of healthy subjects [29]. According to previous studies, ba-PWV ≥ 1600 cm/sec is an independent risk factor for cardiovascular disease and vascular complications [30]. Here, we observed an average ba-PWV above 1700 cm/sec in patients with type 2 DM; this implies arterial stiffness. Tongue manifestations are important features for detecting blood stasis [17]. Our study revealed that the tongue manifestations of blood stasis (i.e., petechiae tongue, bluish tongue, or engorged sublingual collateral vessels) corresponded to higher ba-PWV, particularly for patients with petechiae tongue. Furthermore, we evaluated the relationship between the change of ba-PWV and the number of blood stasis tongue manifestations. The results showed that patients possessing increased blood stasis tongue manifestations had significantly increased mean ba-PWV (78.8 cm/sec; p = 0.037). This suggests that patients with type 2 DM have increased blood stasis tongue manifestations that are correlated with severe arterial stiffness. There were several limitations of our study. First, the sample size was relatively small. Second, we did not enroll healthy controls or patients with type 1 DM; this is because most patients with type 1 DM only use conventional medicine (i.e., insulin injection) and do not utilized TCM as a complementary therapy. Therefore, further studies with larger sample sizes, including healthy subjects as well as type 1 and type 2 DM groups, are required to determine the relationship between tongue manifestations and disease. This proposed study can provide a rationale for a wider use of tongue diagnosis in clinical practice. Conclusion Blood stasis of the tongue, particularly petechiae tongue, is associated with arterial stiffness in patients with type 2 DM. Tongue diagnosis is helpful for detecting blood stasis and could serve as a feasible predictor of DM. Abbreviations ABIAnkle brachial index AC sugarFasting blood-glucose level baPWVBrachial-ankle pulse wave velocity BMIBody mass index CCHChanghua Christian Hospital CrCreatinine GOTGlutamate oxaloacetate transaminase GPTGlutamate pyruvate transaminase HbA1cHemoglobin A1c HDLHigh-density lipoprotein LDLLow-density lipoprotein TCMTraditional Chinese medicine TGTriglyceride The authors would like to thank all of the colleagues who contributed to this study. Funding This study was supported in part by China Medical University Hospital, Taichung Taiwan (DMR-105-094), and in part by Taiwan Ministry of Health and Welfare Clinical Trial and Research Center of Excellence (MOHW105-TDU-B-212-133019) and CMU under the Aim for Top University Plan of the Ministry of Education, Taiwan. Availability of data and materials All the data is contained in the manuscript. Authors’ contributions PCH, JYC, and LCL conceived the study. PCH, PYL, and LCL conducted the study. YCH performed the statistical analysis. PCH ,HHC and LCL led the writing of the manuscript. All authors commented on the analytic plan and interpretation, and contribution to the editing and final approval of the manuscript. Competing interests All authors declare that they have no conflict of interest. Consent for publication Informed consent documents were obtained for publication of these figures and photographs in the article. Ethics approval and consent to participate This study was approved by the Institutional Review Board of the Changhua Christian Hospital. All participants in the study signed informed consent documents. ==== Refs References 1. Yang W Lu J Weng J Jia W Ji L Xiao J Shan Z Liu J Tian H Ji Q Prevalence of diabetes among men and women in China N Engl J Med 2010 362 12 1090 1101 10.1056/NEJMoa0908292 20335585 2. Engelgau MM Geiss LS Saaddine JB Boyle JP Benjamin SM Gregg EW Tierney EF Rios-Burrows N Mokdad AH Ford ES The evolving diabetes burden in the United States Ann Intern Med 2004 140 11 945 950 10.7326/0003-4819-140-11-200406010-00035 15172919 3. Wild S Roglic G Green A Sicree R King H Global prevalence of diabetes: estimates for the year 2000 and projections for 2030 Diabetes Care 2004 27 5 1047 1053 10.2337/diacare.27.5.1047 15111519 4. Chen SY Hsu YM Lin YJ Huang YC Chen CJ Lin WD Liao WL Chen YT Lin WY Liu YH Yang JS Sheu JC Tsai FJ Current concepts regarding developmental mechanisms in diabetic retinopathy in Taiwan Biomed (Taipei) 2016 6 2 1 8 10.7603/s40681-016-0001-9 5. Jung JH Tantry US Gurbel PA Jeong YH Current antiplatelet treatment strategy in patients with diabetes mellitus Diabetes Metab J 2015 39 2 95 113 10.4093/dmj.2015.39.2.95 25922803 6. Creager MA Luscher TF Cosentino F Beckman JA Diabetes and vascular disease: pathophysiology, clinical consequences, and medical therapy: Part I Circulation 2003 108 12 1527 1532 10.1161/01.CIR.0000091257.27563.32 14504252 7. Prenner SB Chirinos JA Arterial stiffness in diabetes mellitus Atherosclerosis 2015 238 2 370 379 10.1016/j.atherosclerosis.2014.12.023 25558032 8. Emerging Risk Factors C Sarwar N Gao P Seshasai SR Gobin R Kaptoge S Di Angelantonio E Ingelsson E Lawlor DA Selvin E Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease: a collaborative meta-analysis of 102 prospective studies Lancet 2010 375 9733 2215 2222 10.1016/S0140-6736(10)60484-9 20609967 9. Kaptchuk TJ Chinese medicine : the web that has no weaver 2000 London Rider 10. Lee JA Park TY Lee J Moon TW Choi J Kang BK Ko MM Lee MS Developing indicators of pattern identification in patients with stroke using traditional Korean medicine BMC Res Notes 2012 5 136 10.1186/1756-0500-5-136 22410195 11. Park B You S Jung J Lee JA Yun KJ Lee MS Korean studies on blood stasis: an overview Evid Based Complement Altern Med 2015 2015 316872 12. Park YJ Yang DH Lee JM Park YB Development of a valid and reliable blood stasis questionnaire and its relationship to heart rate variability Complement Ther Med 2013 21 6 633 640 10.1016/j.ctim.2013.08.019 24280472 13. Luo L, Jiang Y. Lin chuang wang she cai se tu jie=Traditional Chinese medicine tongue diagnosis. Zhang hua shi: Luo lun qian, jiang yi wu; 2010. 14. Lo LC Chen YF Chen WJ Cheng TL Chiang JY The study on the agreement between automatic tongue diagnosis system and traditional Chinese medicine practitioners Evid Based Complement Alternat Med 2012 2012 505063 22924055 15. Liao PY Hsu PC Chen JM Chiang JY Lo LC Diabetes with pyogenic liver abscess--A perspective on tongue assessment in traditional Chinese medicine Complement Ther Med 2014 22 2 341 348 10.1016/j.ctim.2013.12.009 24731906 16. Lo LC, Chiang JY. Traditional Chinese medicine tongue diagnosis. Changhua city: Lun-Chien lo, John. Y. Chiang; 2010. 17. Lo LC Chen CY Chiang JY Cheng TL Lin HJ Chang HH Tongue diagnosis of traditional Chinese medicine for rheumatoid arthritis Afr J Tradit Complement Altern Med 2013 10 5 360 369 24311851 18. Laurent S Cockcroft J Van Bortel L Boutouyrie P Giannattasio C Hayoz D Pannier B Vlachopoulos C Wilkinson I Struijker-Boudier H Expert consensus document on arterial stiffness: methodological issues and clinical applications Eur Heart J 2006 27 21 2588 2605 10.1093/eurheartj/ehl254 17000623 19. Zheng ZJ Sharrett AR Chambless LE Rosamond WD Nieto FJ Sheps DS Dobs A Evans GW Heiss G Associations of ankle-brachial index with clinical coronary heart disease, stroke and preclinical carotid and popliteal atherosclerosis: the Atherosclerosis Risk in Communities (ARIC) Study Atherosclerosis 1997 131 1 115 125 10.1016/S0021-9150(97)06089-9 9180252 20. Jiang M Zha Q Lu C He Y Lu A Association between tongue appearance in Traditional Chinese Medicine and effective response in treatment of rheumatoid arthritis Complement Ther Med 2011 19 3 115 121 10.1016/j.ctim.2011.05.002 21641515 21. He Y Lu A Zha Y Tsang I Differential effect on symptoms treated with traditional Chinese medicine and western combination therapy in RA patients Complement Ther Med 2008 16 4 206 211 10.1016/j.ctim.2007.08.005 18638711 22. Jiang WY Therapeutic wisdom in traditional Chinese medicine: a perspective from modern science Trends Pharmacol Sci 2005 26 11 558 563 10.1016/j.tips.2005.09.006 16185775 23. Lo LC Chiang JY Cheng TL Shieh PS Visual agreement analyses of traditional chinese medicine: a multiple-dimensional scaling approach Evid Based Complement Alternat Med 2012 2012 516473 23024693 24. Negrato CA Tarzia O Buccal alterations in diabetes mellitus Diabetol Metabol Syndr 2010 2 3 10.1186/1758-5996-2-3 25. Tong XL Dong L Chen L Zhen Z Treatment of diabetes using traditional Chinese medicine: past, present and future Am J Chin Med 2012 40 5 877 886 10.1142/S0192415X12500656 22928822 26 Anastasi JK Currie LM Kim GH Understanding diagnostic reasoning in TCM practice: tongue diagnosis Altern Ther Health Med 2009 15 3 18 28 19472861 27 Laurent S Boutouyrie P Asmar R Gautier I Laloux B Guize L Ducimetiere P Benetos A Aortic stiffness is an independent predictor of all-cause and cardiovascular mortality in hypertensive patients Hypertension 2001 37 5 1236 1241 10.1161/01.HYP.37.5.1236 11358934 28 O'Rourke M Mechanical principles in arterial disease Hypertension 1995 26 1 2 9 10.1161/01.HYP.26.1.2 7607724 29 Taniwaki H Kawagishi T Emoto M Shoji T Kanda H Maekawa K Nishizawa Y Morii H Correlation between the intima-media thickness of the carotid artery and aortic pulse-wave velocity in patients with type 2 diabetes. Vessel wall properties in type 2 diabetes Diabetes Care 1999 22 11 1851 1857 10.2337/diacare.22.11.1851 10546019 30 Ha BK Kim BG Kim DH Lee SI Jung SM Park JY Lee CW Kim SS Kim BH Kim IJ Relationships between brachial-ankle pulse wave velocity and peripheral neuropathy in type 2 diabetes Diab Metabol J 2012 36 6 443 451 10.4093/dmj.2012.36.6.443

PMC005xxxxxx/PMC5002209.txt

==== Front BMC Infect DisBMC Infect. DisBMC Infectious Diseases1471-2334BioMed Central London 177610.1186/s12879-016-1776-8Research ArticlePrevalence, risk factors for infection and subtype distribution of the intestinal parasite Blastocystis sp. from a large-scale multi-center study in France El Safadi Dima dima.elsafadi@hotmail.com 1Cian Amandine amandine.cian@pasteur-lille.fr 1Nourrisson Céline c_nourrisson@chu-clermontferrand.fr 23Pereira Bruno bpereira@chu-clermontferrand.fr 4Morelle Christelle christelle.morelle@univ-montp1.fr 5Bastien Patrick p-bastien@chu-montpellier.fr 5Bellanger Anne-Pauline apbellanger@chu-besancon.fr 6Botterel Françoise francoise.botterel@hmn.aphp.fr 7Candolfi Ermanno candolfi@unistra.fr 8Desoubeaux Guillaume guillaume.desoubeaux@univ-tours.fr 9Lachaud Laurence laurence.LACHAUD@chu-nimes.fr 10Morio Florent Florent.MORIO@chu-nantes.fr 11Pomares Christelle pomares.c@chu-nice.fr 12Rabodonirina Meja meja.rabodonirina@chu-lyon.fr 13Wawrzyniak Ivan ivan.wawrzyniak@univ-bpclermont.fr 3Delbac Frédéric Frederic.DELBAC@univ-bpclermont.fr 3Gantois Nausicaa Nausicaa.Gantois@pasteur-lille.fr 1Certad Gabriela Gabriela.Certad@pasteur-lille.fr 1Delhaes Laurence laurence.delhaes@gmail.com 114Poirier Philippe ppoirier@chu-clermontferrand.fr 23Viscogliosi Eric eric.viscogliosi@pasteur-lille.fr 11 Université de Lille, CNRS, Inserm, CHU Lille, Institut Pasteur de Lille, U1019 – UMR 8204 – CIIL – Centre d’Infection et d’Immunité de Lille, 1 rue du Professeur Calmette, BP 245, 59019 Lille cedex, France 2 Laboratoire de Parasitologie-Mycologie, CHU Gabriel-Montpied, Clermont-Ferrand, France 3 Clermont Université, Université Blaise Pascal-Université d’Auvergne - CNRS UMR 6023 Laboratoire Microorganismes: Génome et Environnement, Clermont-Ferrand, France 4 CHU Clermont-Ferrand, Unité de Biostatistiques, Direction de la Recherche Clinique (DRCI), Clermont-Ferrand, France 5 Laboratoire de Parasitologie-Mycologie, CHU de Montpellier, CNRS UMR 5290/IRD 224/UM1, Université de Montpellier 1, Montpellier, France 6 Laboratoire de Parasitologie-Mycologie, CHU de Besançon, Besançon, France 7 Laboratoire de Parasitologie-Mycologie, AP-HP Hôpital Henri Mondor, Créteil, France 8 Institut de Parasitologie et de Pathologie Tropicale de Strasbourg, Université de Strasbourg, Hôpitaux Universitaires de Strasbourg, Strasbourg, France 9 Service de Parasitologie-Mycologie-Médecine Tropicale, CHU de Tours/CEPR Inserm U1100 Equipe 3, Université François-Rabelais de Tours, Tours, France 10 Département de Parasitologie-Mycologie, Faculté de Médecine de Montpellier-Nîmes, Université de Montpellier I, CHU de Montpellier, Montpellier, France 11 Département de Parasitologie et Mycologie Médicale, Laboratoire de Parasitologie-Mycologie, Institut de Biologie, CHU de Nantes, EA1155-IICiMed, Université de Nantes, Nantes, France 12 Laboratoire de Parasitologie-Mycologie CHU de Nice, C3M INSERM U1065, Université de Nice Sophia Antipolis, Nice, France 13 Service de Parasitologie, Hospices Civils de Lyon, Lyon, France 14 Département de Parasitologie-Mycologie, CHU de Lille, Faculté de Médecine, Lille, France 26 8 2016 26 8 2016 2016 16 1 4516 6 2016 11 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Blastocystis sp. is the most common intestinal parasite of humans. Despite its potential public health impact, epidemiological data regarding the prevalence and molecular subtype distribution of Blastocystis sp. in Europe are rarely reported. Therefore, the first multi-center epidemiological survey performed in Europe was conducted in France to diagnose and subtype Blastocystis sp. and to identify risk factors for infection. Methods Stool samples from 788 patients were collected either in summer or winter in 11 hospitals throughout France together with patient data. All stool samples were tested for the presence of Blastocystis sp. by quantitative PCR targeting the SSU rDNA gene. Positive samples were sequenced to determine the distribution of the subtypes in our cohort. Statistical analyses were performed to identify potential risk factors for infection. Results Using quantitative PCR, the overall prevalence of Blastocystis sp. was shown to reach 18.1 %. The prevalence was significantly higher in summer (23.2 %) than in winter (13.7 %). Travellers or subjects infected with other enteric parasites were significantly more infected by Blastocystis sp. than non-travellers or subjects free of other enteric parasites, respectively. Different age-related epidemiological patterns were also highlighted from our data. The prevalence of Blastocystis sp. was not significantly higher in patients with digestive symptoms or diagnosed with chronic bowel diseases. Among symptomatic patients, Blastocystis sp. infection was significantly associated with abdominal pain. Gender, socioeconomic status, and immune status were not identified as potential risk factors associated with infection. Among a total of 141 subtyped isolates, subtype 3 was predominant (43.3 %), followed by subtype 1 and subtype 4 (20 %), subtype 2 (12.8 %), subtype 6 and subtype 7 (2.1 %). No association between ST and clinical symptoms was statistically evidenced. Conclusions A high prevalence of Blastocystis sp. infection was found in our French patient population. Seasonal impact on the prevalence of Blastocystis sp. was highlighted and recent travels and age were identified as main risk factors for infection. Most cases were caused by subtypes 1 to 4, with a predominance of subtype 3. Large variations in both prevalence and ST distribution between hospitals were also observed, suggesting distinct reservoirs and transmission sources of the parasite. Electronic supplementary material The online version of this article (doi:10.1186/s12879-016-1776-8) contains supplementary material, which is available to authorized users. Keywords Blastocystis spIntestinal parasiteMolecular epidemiologyPCRSubtypingRisk factors for infectionUniversité de Lille 2Programme Orientations StratégiquesViscogliosi Eric issue-copyright-statement© The Author(s) 2016 ==== Body Background Blastocystis sp. is a common protozoan intestinal parasite with worldwide distribution that inhabits the digestive tract of humans and a large variety of animal hosts [1–3]. In numerous epidemiological surveys, this cosmopolitan enteric parasite was frequently identified as the most common unicellular eukaryote found in human fecal samples [1, 2, 4]. Indeed, its prevalence may reach 20 % in industrialized countries, including the European population [5] and 50 % in developing countries [6]. Recently, the prevalence of Blastocystis sp. was shown to be 100 % in a cohort of children living in a rural area in Senegal, highlighting the impact of blastocystosis mainly in developing countries with poor healthcare and hygiene [7]. In this regard, a higher prevalence of this parasite was found among European people with a history of recent travel to tropical countries [5]. At the morphological level, four major forms of Blastocystis sp. have been described, including the infective cyst which is able to survive for a long period in feces and environmental sources and is resilient to standard water chlorination, facilitating waterborne transmission of the parasite [1, 8]. Therefore, the fecal-oral route is considered the main mode of transmission of Blastocystis sp. through the consumption of food or water contaminated by cysts. Blastocystosis is usually diagnosed using direct-light microscopy of fecal smears or possibly short-term xenic in vitro culture of stool samples. However, these methods have a low diagnostic sensitivity compared with molecular tools, i.e. PCR assays, and could greatly underestimate the real prevalence of the parasite [9]. A remarkable genetic diversity has been revealed among Blastocystis sp. isolates from humans and other animals based on the comparison of small subunit (SSU) rRNA gene sequences. Consequently, seventeen lineages of so-called subtypes (ST1 to ST17) (arguably separate species) have been identified among mammalian and avian isolates [10], nine of which (ST1 to ST9) are found in humans with varying prevalence [2, 4]. The other STs (ST10-ST17) are exclusively found in animals [10]. Based on a recent review including all human samples subtyped thus far across various geographic regions worldwide [4], approximately 90 % of human isolates belonged to ST1 to ST4, with a predominance of ST3 (around 60 % of these isolates). Even though these four STs were found in different animal hosts, their predominance in the human population is likely explained by large-scale human-to-human transmission [1, 2]. To our knowledge, ST9 was restricted to humans and until now has been identified in only 3 people from Denmark and Japan [4]. ST5 to ST8 supposedly of animal origin were rarely found in humans and their presence might be linked to zoonotic transmission. Besides, a higher risk of Blastocystis sp. infection was found in people with close animal contact, including zoo keepers [11]. The human health impact of Blastocystis sp. still remains uncertain because the parasite is frequently found in asymptomatic patients and has been associated with a wide range of non-specific symptoms including diarrhea, abdominal pain, bloating, nausea, and vomiting as well as urticarial lesions [1–3]. However, recent findings using in vitro and in vivo approaches combined with in silico analysis of genomic data and clinical reports strongly suggested the pathogenic potential of Blastocystis sp. by identifying putative virulence factors such as cysteine proteases. These proteases are secreted by the parasite and can induce epithelial barrier dysfunction [1–3, 12–14]. The proposed models for pathogenesis of Blastocystis sp. [13–16] mainly involved adhesion of parasites to the intestinal epithelium, apoptosis and degradation of tight junction proteins of intestinal epithelial cells resulting in increased intestinal permeability, degradation of IgA and induction of a pro-inflammatory cytokine response. Blastocystis sp. was also recently associated with Irritable Bowel Syndrome (IBS) [16, 17], a multifactorial functional bowel disorder partly explained by dysbiosis [18]. All these new data provide evidence that the public health burden of Blastocystis sp. continues to be underestimated, hence the interest in conducting large-scale epidemiological surveys in industrialized countries. In France, very little data were available concerning both the prevalence and ST distribution of Blastocystis sp. The parasite was previously reported in two French cohorts with a respective prevalence of 3.0 and 6.1 % by direct-light microscopy of fecal smears [19, 20]. In addition, conflicting ST distributions were observed between two French molecular studies conducted in different geographic areas and both including a limited number of samples [9, 21]. Therefore, the aim of the present study was to reinforce the picture of Blastocystis sp. prevalence and molecular ST distribution in Europe by performing the first multi-center survey conducted in France from a large cohort of patients carried out between December 2012 and September 2013. Methods Cohort of patients and collection of samples This cross-sectional study was conducted in France between December 2012 and September 2013 and involved the parasitology-mycology medical laboratories of 11 teaching hospitals (Besançon, Clermont-Ferrand, Créteil, Lille, Lyon, Montpellier, Nantes, Nice, Nîmes, Strasbourg and Tours) throughout France (Fig. 1). As part of the study, each laboratory randomly selected 21 to 50 stool samples in both winter (from December 2012 to February 2013) and summer (from July 2013 to September 2013) to subsequently evaluate potential seasonal variations in the prevalence (percentage of subjects infected) of the parasite. All these samples (1 sample per patient) were collected at each participating center during routine clinical procedures. A total of 788 subjects followed up for different pathologies, with/without gastrointestinal symptoms were enrolled in this study. A standardized questionnaire was designed to collect information about each participating subject (sex, age, profession, recent travels and destinations, exposure to pets) as well as clinical data especially regarding immune status, presence of digestive symptoms (diarrhea, vomiting, bloating, constipation, and abdominal pain), and diagnosis of IBS or IBD (Inflammatory Bowel Disease). In addition, the observation of intestinal protozoan parasites (Blastocystis sp., amoebas, trichomonads, diplomonads, apicomplexa), fungi (microsporidia), and helminths by direct-light microscopy of fecal smears was also recorded, as well as digestive diseases of bacterial origin. For each subject, about 500 mg of fresh stools was collected and then homogenized by shaking in 1.5 ml of Stool Transport and Recovery (S.T.A.R.) buffer (Roche Diagnostics, Indianapolis, IN) (ratio 1:3 according to the manufacturer’s recommendations). Samples were stored and then transported at −20 °C to Lille for molecular screening and subtyping of Blastocystis sp.Fig. 1 Location of the 11 French centers and seasonal prevalence of Blastocystis sp. by center. Prevalence data were obtained from the 788 enrolled patients. The map of France was obtained from the website Servier Medical Art (http://www.servier.fr/smart/banque-dimages-powerpoint) DNA extraction Stool samples stored in the S.T.A.R. buffer were stirred and then centrifuged for 1 min at 1,000 xg. Total genomic DNA was extracted from 200 μL of the supernatant using the QiaAMP DNA Stool Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s recommended procedures. The DNA was eluted in 100 μL of elution buffer (Qiagen) and stored at −20 °C until use. Detection and molecular subtyping of Blastocystis sp For each sample, 2 μL of extracted DNA was subjected to real-time quantitative PCR (qPCR) assay to detect and subtype Blastocystis sp. qPCR was carried out using the Blastocystis-specific primer pair BL18SPPF1 (5’-AGTAGTCATACGCTCGTCTCAAA-3’)/BL18SR2PP (5’-TCTTCGTTACCCGTTACTGC-3’) targeting the SSU rRNA coding region as previously described [9]. DNA extraction controls (isolation of DNAs without stool and from a Blastocystis sp.-negative stool) subsequently used in qPCR assays and positive (DNA obtained from Blastocystis sp. ST4 or ST7 cultures) and negative (DNA matrix replaced by water) qPCR controls were performed. qPCR product from each positive sample was purified and sequenced in both strands by Genoscreen (Lille, France) or Beckman Coulter Genomics (Essex, United Kingdom). The SSU rRNA gene sequences obtained in this study were deposited in GenBank under accession numbers KU158944 to KU159084 (see Additional file 1). Obtained sequences were compared with all Blastocystis sp. homologous sequences available from the National Centre for Biotechnology Information (NCBI) using the Basic Local Alignment Search Tool (BLAST) program. STs were identified by determining the exact match or closest similarity against all known Blastocystis sp. STs. For two samples, sequence chromatograms analysis revealed the presence of double traces, suggesting mixed infection by different STs. Statistical analysis All analyses were performed using the Stata statistical software (version 13, StataCorp, College Station, US). The categorical data were expressed as the number of patients and associated percentages, whereas the quantitative data were expressed as the mean and associated standard deviation according to the statistical distribution (assumption of normality studied using Shapiro-Wilk’s test). Comparisons between independent groups were performed by Chi-square or Fisher’s exact tests for categorical parameters and by Student’s t or Mann-Whitney tests when t-test conditions were not respected (normality and homoscedasticity studied by the Fisher-Snedecor test) for quantitative variables. The multivariate analysis was performed by stepwise approach according to univariate results and clinical relevance. A random-effects logistic regression was applied to determine parameters associated with the prevalence of Blastocystis sp., taking into account between and within center variability. Results were reported as odds-ratio (noted OR) and 95 % confidence intervals. P-values of 0.05 or below were considered significant (two-sided). Due to multiple comparisons, the type-I-error correction was considered when appropriate. So, the Marascuillo’s procedure was performed after Chi-square test (for example prevalence of Blastocystis sp. among age groups). Results Analysis of the cohort Stool samples were collected from a total of 788 patients (Table 1). Among them, 417 were recruited during winter and 371 in summer. The sex ratio (M/F) was 1.37 and the age of the participants (7 missing data) was between 7 months and 95 years (mean age of 45.7 ± 21.3 years). Epidemiological records revealed that 178 patients had a recent history of travel outside France (i.e. during the last 12 months). Regarding the immune statuses of the patients, 351 were immunocompetent and 378 were immunocompromised (59 missing data). Among immunocompromised patients, 58 were positive for Human Immunodeficiency Virus (HIV) infection, 95 received a solid organ transplant, 131 received immunosuppressive therapy, 65 received a bone marrow transplant, and 34 suffered from tumorigenic processes. Forty patients suffered from chronic intestinal disease with 25 subjects presenting IBD and 15 IBS. Among this cohort, 502 patients presented digestive symptoms, such as diarrhea (52.4 %), abdominal pain (29.9 %), bloating (6.7 %), constipation (3.6 %) and vomiting (3.3 %), and 234 subjects were asymptomatic (52 missing data).Table 1 Prevalence and ST distribution of Blastocystis sp. by center and among the total cohort Centera Samples (n) Positive samples (n) Prevalence (%) Blastocystis sp. STs Mixed infection (n) ST1 ST2 ST3 ST4 ST6 ST7 Besançon (W) 48 7 14.6 - 2 3 2 - - - Besançon (S) 41 9 22.0 1 1 2 4 - - 1 Besançon (T) 89 16 18.0 1 3 5 6 - - 1 Clermont-Ferrand (W) 50 4 8.0 1 - 1 1 1 - - Clermont-Ferrand (S) 50 6 12.0 - - 2 4 - - - Clermont-Ferrand (T) 100 10 10.0 1 - 3 5 1 - - Créteil (W) 29 1 3.4 1 - - - - - - Créteil (S) 29 7 24.1 1 - 6 - - - - Créteil (T) 58 8 13.8 2 - 6 - - - - Lille (W) 30 12 40.0 4 1 3 3 1 - - Lille (S) 21 9 42.9 2 1 3 3 - - - Lille (T) 51 21 41.2 6 2 6 6 1 - - Lyon (W) 38 7 18.4 3 1 2 - - 1 - Lyon (S) 44 8 18.2 1 2 4 - 1 - - Lyon (T) 82 15 18.3 4 3 6 - 1 1 - Montpellier (W) 30 5 16.7 - 2 3 - - - - Montpellier (S) 29 8 27.6 1 2 3 2 - - - Montpellier (T) 59 13 22.0 1 4 6 2 - - - Nantes (W) 49 5 10.2 - - 3 1 - 1 - Nantes (S) 42 13 30.1 3 1 3 4 - 1 1 Nantes (T) 91 18 19.8 3 1 6 5 - 2 1 Nice (W) 33 3 9.1 - 1 2 - - - - Nice (S) 30 7 23.3 - 1 6 - - - - Nice (T) 63 10 15.9 - 2 8 - - - - Nîmes (W) 30 2 6.7 - - 2 - - - - Nîmes (S) - - - - - - - - - - Nîmes (T) 30 2 6.7 - - 2 - - - - Strasbourg (W) 50 9 18.0 3 - 4 2 - - - Strasbourg (S) 48 14 29.2 3 3 7 1 - - - Strasbourg (T) 98 23 23.5 6 3 11 3 - - - Tours (W) 30 2 6.7 2 - - - - - - Tours (S) 37 5 13.5 2 - 2 1 - - - Tours (T) 67 7 10.4 4 - 2 1 - - - Total (W) 417 57 13.7 14 7 23 9 2 2 0 Total (S) 371 86 23.2 14 11 38 19 1 1 2 Grand total 788 143 18.1 28 18 61 28 3 3 2 a W winter, S summer, T total Screening for Blastocystis sp. using qPCR The overall prevalence of Blastocystis sp. was shown to be 18.1 % (143/788) in our study using qPCR (Table 1). The sensitivity of direct-light microscopy of fecal smears was only 45.8 % compared to the molecular method. All samples positive by microscopy were positive by the PCR assay. The overall prevalence determined by qPCR ranged from 10.0 % (Clermont-Ferrand) to 41.2 % (Lille) (Fig. 1 and Table 1). The overall prevalence in summer was significantly higher than in winter (23.2 % versus 13.7 %, p < 0.001) (Fig. 1 and Table 1). The difference in prevalence between males (19.5 %) and females (16.6 %) was not significant (p = 0.29). The mean age of Blastocystis sp.-infected patients was significantly lower when compared to the age in Blastocystis sp.-free subjects (43.0 ± 21.1 years versus 48.9 ± 21.1 years; Student’s t-test, p = 0.003). Subgroup analysis (Fig. 2) revealed that the prevalence of Blastocystis sp. was 26.3 % among subjects aged 0–14 years (15/57), 22.2 % in the group aged 15–49 years (74/333), and 13.6 % in patients aged over 50 years (53/391). However, while the prevalence of Blastocystis sp. in children was higher, with a peak between 5 and 9 years (36 %, Fig. 2), the statistical analysis only confirmed that subjects aged 15 to 49 years showed a higher prevalence than subjects aged over 50 years (p = 0.01, Marascuillo procedure; see also Additional file 2).Fig. 2 Age distribution of patients with blastocystosis (n = 142) Among all the patients, 70 were found to be infected with other enteric parasites, either protozoans or helminths (see Additional file 3). The prevalence of Blastocystis sp. reached 40 % in those patients (28/70), compared to 16.2 % in subjects not infected with other enteric parasites (p < 0.001). Two additional patients presented bacterial infections with Clostridium and Salmonella, respectively. Univariate and multivariate analyses highlighted that patients traveling during the last 12 months had a higher prevalence of Blastocystis sp. than non-travellers (27.5 versus 14.7 %, univariate: p < 0.001, multivariate logistic model: OR = 1.90 [1.18; 3.05], p = 0.009). A history of two or more travels during the previous year was associated with a higher prevalence, reaching 33.3 % (p = 0.43). Even though the travel destination was recorded (Africa, South America, Asia, North America, Oceania, and Europe), no significant difference between destinations and the prevalence of Blastocystis sp. was identified. The socioeconomic level of all subjects was also recorded, and different professional classes were distinguished, including students, workers, executives, the unemployed and pensioners. The prevalence of Blastocystis sp. was not significantly different between the different professional classes. Food handlers and pet owners did not show a significantly higher prevalence than other individuals (22.2 versus 18.8 %, p = 0.57 and 21.9 versus 19.7 %, p = 0.61, respectively). The prevalence of Blastocystis sp. in immunocompromised subjects was significantly lower than in immunocompetent patients (12.4 versus 24.2 %, p < 0.001). Within immunocompromised subjects, patients presenting with HIV, solid organ transplants, immunosuppressive therapy, solid cancer, and bone marrow transplants were distinguished. Subgroup analyses confirmed that the prevalence of Blastocystis sp. was significantly lower in patients receiving immunosuppressive treatment (8.4 %, p = 0.001) and with bone marrow transplant (7.7 %, p = 0.02) but not significantly lower in the other subgroups. The prevalence of Blastocystis sp. in patients suffering from IBD and IBS was 12 and 20 %, respectively, compared to 18.4 % in patients without a history of chronic bowel diseases (p = 0.41 and p = 0.88, respectively). Moreover, all IBS and IBD patients infected with Blastocystis sp. were not infected with other enteric parasites. Blastocystis sp. infection and digestive symptoms The prevalence of Blastocystis sp. was not shown to be significantly higher in symptomatic patients (18.5 %, 93/502) than in asymptomatic carriers (16.2 %, 38/234). The digestive symptoms selected for the study were abdominal pain, bloating, diarrhea, constipation, and vomiting (Fig. 3). Within subjects positive for the parasite and presenting digestive symptoms, Blastocystis sp. infection was significantly associated with abdominal pain (23.3 % in patients with abdominal pain versus 15.7 % in patients without abdominal pain, p = 0.007). However, the most frequent symptom observed in symptomatic carriers was bloating, even if its prevalence in Blastocystis sp.-infected patients compared to non-carriers was not significant (26.4 versus 17.1 %, p = 0.09).Fig. 3 Distribution of digestive symptoms in symptomatic patients infected or not infected with Blastocystis sp. (n = 502) Distribution of Blastocystis sp. STs Among the 143 positive samples, 6 different STs were identified (ST1 to ST4, ST6 and ST7) (Table 1). Two patients had mixed infection (2/143, 1.4 %). Among the 141 patients with a single ST, ST3 was predominant (n = 61, 43.3 %), followed by ST1 and ST4 (both n = 28, 20.0 %), ST2 (n = 18, 12.8 %), ST6 and ST7 (both n = 3, 2.1 %) (Tables 1 and 2). The distribution of STs varied widely between the medical centers (Table 1), and no significant difference in the distribution of STs was found according to the season. Moreover, the distribution of STs between the symptomatic and asymptomatic groups of patients was almost similar. The genetic diversity among isolates belonging to the same ST was very low since 130/141 isolates (92.2 %) belonging to the different STs identified in the present study showed 99 to 100 % identity with homologous sequences available in databases for the same STs. Only 11 isolates belonging either to the ST1 or ST3 exhibited 96 to 98 % identity with homologous sequences. Therefore, numerous sequences of isolates belonging to the same ST were similar or closely related to each other.Table 2 Total and seasonal distribution of Blastocystis sp. STs (n = 141) Blastocystis sp. STs ST1 ST2 ST3 ST4 ST6 ST7 Winter n (%) 14 (24.1) 7 (12.1) 23 (39.7) 9 (15.5) 2 (3.4) 2 (3.4) Summer n (%) 14 (16.7) 11 (13.1) 38 (45.2) 19 (22.6) 1 (1.2) 1 (1.2) Total n (%) 28 (20.0) 18 (12.8) 61 (43.3) 28 (20.0) 3 (2.1) 3 (2.1) Discussion Little data are available in the literature regarding the prevalence of Blastocystis sp. in France and more generally in European countries. In France, the first two studies reported a prevalence of 3 and 6.1 % in two cohorts of 2,581 and 9,700 patients, respectively, by direct-light microscopy of fecal smears [19, 20]. More recently, the prevalence of Blastocystis sp. reached 14.5 % in a single-center study including 186 patients using a molecular assay [9]. Interestingly, this latter value is roughly similar to the prevalence of 18.1 % reported in the current samples using the same qPCR assay. By comparison, the prevalence reported in a few other neighbouring countries was 24.2 % in the Netherlands (n = 442) [5], 23 % in Denmark (n = 93) [22], 7.1 % in Italy (n = 5,351) [23], 7 % in Spain (n = 8,313) [24], and 6.9 and 3.9 % in the United Kingdom (n = 1,390 and n = 1,000, respectively) [25, 26] (Table 3). However, a comparison of the prevalence obtained from these various European studies remains generally uninformative due to the differences in the composition of the cohorts of patients and especially in the diagnostic tools. Indeed, apart from the Dutch and Danish studies, which were conducted using molecular tools and showed a prevalence similar to that of our survey, all other European epidemiological studies were performed using direct-light microscopy or in vitro culture, both methods being shown to be less sensitive than PCR [9]. The present study confirms this observation, since direct-light microscopy showed only 45.8 % sensitivity compared to the qPCR assay.Table 3 Prevalence of Blastocystis sp. in European countries Country Region/city Total number of patients Method of detection Prevalence Reference France Grenoble 2,581 Direct-light microscopy 3.0 % [19] France Paris 9,700 Direct-light microscopy 6.1 % [20] France Clermont-Ferrand 186 qPCR 14.5 % [9] France Multi-center study 788 qPCR 18.1 % Present study The Netherlands Amsterdam 442 PCR 24.2 % [5] Denmark Copenhagen 93 PCR 23.0 % [22] Italy Rome 5,351 Direct-light microscopy 7,1 % [23] Spain Catalonia 8,313 Direct-light microscopy 7.0 % [24] United Kingdom (Wales) Aberystwyth 1,390 Direct-light microscopy 6.9 % [25] United Kingdom (Scotland) Glasgow 1,000 In vitro culture and direct-light microscopy 3.9 % [26] Within the French centers, the prevalence of Blastocystis sp. ranged from 6.7 to 41.2 %. This variation may be naturally explained by differences in the composition of the respective cohorts from each center, but may also reflect differences in food habits and sources of drinking water in various geographic areas, which are also more or less rural and have different climate conditions. In future studies, the analysis of food and environmental samples in these regions, and especially the control of water sources regarding the presence of Blastocystis sp., might help identifying potential primary reservoirs of transmission. A key finding of our study was the seasonal impact on the prevalence of Blastocystis sp., which reached 23.2 % in summer compared to 13.7 % in winter. Interestingly, this seasonal pattern, already described in previous epidemiological surveys [26–29], was observed in 9 of the 10 French centers providing samples during both winter and summer. In France, this difference may be explained by changes in food habits according to the seasons, with an increased consumption of vegetables and fruits, drinks with ice cubes and ice creams in summer. Common water-based recreational activities may also be involved, since human fecal contamination was clearly shown to be correlated with Blastocystis sp. load in recreational rivers, suggesting a greater risk of infection by the parasite in summer [28, 29]. Moreover, frequent trips during the summer holidays and stays for instance in densely populated holiday centers could represent other risk factors for infection. Among our overall population, gender was not identified as a potential risk factor associated with Blastocystis sp. infection, since the finding of a slightly higher prevalence of the parasite in males (19.5 %) than in females (16.6 %) was not statistically significant. By contrast, Blastocystis sp. showed a different age-related epidemiological pattern. The mean age was thus significantly lower in Blastocystis sp.-infected patients (43.0 ± 21.1 years) than in non-carriers of the infection (48.9 ± 21.1 years). In addition, the prevalence of Blastocystis sp. was significantly higher in subjects aged 15 to 49 years compared to those aged over 50 years (22.2 versus 16.6 %). On the other side, the prevalence of Blastocystis sp. was also not significantly higher in the age group 0 to 14 years (26.3 %) compared to older age classes, likely due to the too small number of children included in our study. Interestingly, an infection peak was shown between 5and 9 years of age, suggesting that children in this age category might be more at risk for Blastocystis sp. infection. In this regard, previous studies reported peaking prevalence of the parasite among groups aged under 10 years [27, 30–32]. Such a high rate may be due to inadequate toilet training and hygiene practices of school-children and cross-transmission through close personal contact. In a recent survey conducted in the Netherlands [5], a significantly higher prevalence of Blastocystis sp. was reported among patients with a history of recent travel, suggesting that trips to tropical and low-income countries may increase the risk of parasite infection. A similar conclusion was drawn from our study, since travel during the last 12 months was significantly associated with a higher prevalence of Blastocystis sp. (27.5 versus 14.7 % for non-travellers). Interestingly, the prevalence of the parasite reached 33.3 % in patients reporting at least two travels during the last year in countries at risk. Travellers should therefore follow food and water hygiene recommendations to prevent infection by Blastocystis sp. Intestinal parasitic infections are among the leading causes of morbidity and mortality in patients infected with HIV. Consequently, the Blastocystis sp. detection as a possible pathogenic agent among immunocompromised patients continues to be debated. In this regard, the prevalence of Blastocystis sp. was previously found to be significantly higher in immunocompromised HIV patients, most presenting with diarrhea, than in HIV-seronegative controls [33–35]. Strikingly, Blastocystis sp. was the most commonly occurring parasite among the protozoans searched for in HIV-infected individuals [33, 34, 36, 37], with a prevalence reaching about 70 % in Jakarta, Indonesia [36]. In addition, a statistically significant association was shown between infection with Blastocystis sp. and the presence of digestive disorders among severely immunocompromised HIV-positive patients (with CD4+ T-cell counts < 200/μL) [33]. The prevalence of the parasite was also shown to be negatively correlated with the CD4+ cell count [36] and was significantly higher in HIV patients without antiretroviral therapy than among HIV-positive patients with treatment [38]. In immunocompromised patients presenting haematological malignancies, Blastocystis sp. was more frequently associated with gastrointestinal symptoms than in non-immunocompromised symptomatic patients [39]. This contrasted with a more recent French survey showing no correlation between digestive symptoms and immune status in patients presenting similar pathology [9]. All-in-all, immunodepression seems to be a factor of primary importance in the infection and pathogenic role of Blastocystis sp. However, in the present study, the prevalence of the parasite in immunocompromised subjects was significantly lower than in immunocompetent individuals (12.4 versus 24.2 %, respectively), especially in subgroups of patients receiving immunosuppressive therapy (8.4 %) or with bone marrow transplants (7.7 %). In our opinion, the controlled food diet recommended to these patients for the prevention of potential opportunistic infections along with antibiotic therapy such as metronidazole might have a negative impact on the prevalence of Blastocystis sp. In addition, a history of travel was shown above to be positively correlated with the prevalence of Blastocystis sp.. In fact, only 21.5 % of immunocompromised patients reported having travelled during the last 12 months, compared to 35.4 % of immunocompetent individuals. To clarify the clinical relevance of Blastocystis sp, numerous studies were published related to the comparison of parasite prevalence between symptomatic and asymptomatic individuals [1–3]. If accumulating epidemiological studies suggested that Blastocystis sp. was associated with gastrointestinal disorders, numerous reports did not support this association. From our overall population, the prevalence of Blastocystis sp. was not significantly different between symptomatic and asymptomatic patients, what does not however prejudge the pathogenicity of various isolates. Within the symptomatic group, abdominal pain was reported significantly more frequently in Blastocystis sp. carriers, in agreement with earlier studies recording abdominal pain as one of the most common symptoms of blastocystosis [1]. Bloating was also attributed to blastocystosis in various studies [1]. This symptom was most frequently identified in Blastocystis sp.-positive subjects of our cohort, but was not significantly associated with parasite infection. The prevalence of the parasite in patients suffering from chronic bowel disorders was also investigated, since recent studies suggested an association between Blastocystis sp. and IBS [2, 16], a functional bowel disorder with a prevalence ranging from 5 to 24 % in industrialized countries [18]. Based on a systematic review of the literature and a meta-analysis including previous epidemiological studies in IBS cohorts, it was shown that IBS patients had a relative risk of 2.34 of being infected with Blastocystis sp. when compared to non-IBS subjects [17]. However, in our study, the prevalence of the parasite was not significantly higher in IBS patients (20 %) compared to non-IBS subjects (18.5 %). Since the diagnosis of IBS is still difficult and requires a specific visit to a gastroenterologist, some patients in our study may not have been diagnosed as positive for IBS, which could impact our results. In parallel, according to various studies, the prevalence of Blastocystis sp. in IBD patients was reported to be lower or higher than in non-IBD subjects, which probably depends on the type of IBD (mainly Crohn’s disease or ulcerative colitis) [40, 41]. In our study, the parasite was identified less frequently in IBD subgroups (12 %), but not significantly less than in non-IBD patients (18.4 %). Unfortunately, subgroups analyses from different types of IBD were not performed because of the small size of the IBD cohort. As part of our study, a total of 141 Blastocystis sp. isolates were subtyped to evaluate the ST distribution within our French cohort. ST3 was the most common ST (43.3 %), followed by ST1 and ST4 (20 %), ST2 (12.8 %), ST6 and ST7 (2.1 %). This distribution is nearly similar to that observed in a majority of geographical areas all over the world, including European countries [5, 22, 42], with a predominance of ST3, followed by ST1, ST2 or ST4, highlighting large-scale inter-human transmission [2, 4]. The identification of a few isolates in Lille, Lyon, Nantes, or Clermont-Ferrand belonging to ST6 or ST7 is most likely the result of zoonotic transmission, since both STs are considered avian STs [1–4, 10]. The distribution of some STs in the overall human population showed significant geographical variations, especially in relation to the ST4 [2, 4]. Our data thus confirmed that ST4 is commonly found in Europe [5, 22, 42] and especially in France [9], and is much less frequently detected or absent in Africa, America and Asia. This observation might perhaps be explained by the recently proposed emergence of ST4 in the human population in Europe [2]. The ST distribution was also variable between different French centers, although ST3 was predominant in 8 of the 11 centers. In two previous studies conducted in France, ST3 was the most frequent ST in Lille [21], whereas ST4 showed a higher prevalence in Clermont-Ferrand [9], in agreement with our survey. These patterns in ST distribution, as well as in prevalence of the parasite between centers, suggest that reservoirs and/or sources of contamination may differ from any geographical area of France to another. Until now, and for different reasons, the epidemiological data remain contradictory regarding the correlation between ST and the pathogenesis of Blastocystis sp. [1, 21]. In this regard, our study failed to provide any evidence for ST association with specific symptoms status. Conclusions Our survey provides new insights into the epidemiology of Blastocystis sp. in industrialized countries through the first multi-center study conducted in France. Such a multi-center survey gives a more comprehensive view of the parasite situation in France, by obtaining data regarding the prevalence and ST distribution of the parasite in different geographical areas. From our overall data, age and a history of recent travel represent the principal reliable risk factors for acquiring this infection. A seasonal impact on the prevalence of Blastocystis sp. is also highlighted, with a higher prevalence in summer. Large variations observed in the prevalence and ST distribution of the parasite between French regions suggest various reservoirs and sources of transmission. Further studies in other European countries, with multi-center recruitment of patients, will clearly be required to establish a complete mapping of the prevalence and ST distribution of Blastocystis sp. in this continent and to improve our understanding of the circulation of the parasite within the European population. Additional files Additional file 1: Isolation source (center), season of collect, ST identification and GenBank accession number of Blastocystis sp. isolates characterized in our study. (DOCX 20 kb) Additional file 2: Post hoc comparison of the prevalence of Blastocystis sp. between age groups. (DOCX 14 kb) Additional file 3: Presence of other enteric parasites (protozoans and helminths) and digestive symptoms in the 143 French patients infected with Blastocystis sp. (DOCX 23 kb) Abbreviations HIVHuman immunodeficiency virus IBDInflammatory bowel disease IBSIrritable bowel syndrome OROdds-ratio qPCRReal-time quantitative PCR STSubtype Acknowledgements The authors would like to thank all those who participated in the study and staff at all participating centers: Pr P-Y Hatron, Pr E Hachulla, Pr F Gottrand, Dr F Dubos, Pr I Yakoub Agha and Dr V Coiteux (CHRU of Lille) and Christine Bureau (CHRU of Nantes). Funding This work was supported by grants from the Programme Orientations Stratégiques from the University of Lille 2, the Fonds Hospitalier d’Aide à l’Emergence from the CHRU of Lille, the Centre National de la Recherche Scientifique and the Institut Pasteur of Lille. DES was supported by a PhD fellowship from the Conseil National de la Recherche Scientifique and the AZM & Saade Association and AC by a PhD fellowship from the University of Lille 2 and the Institut Pasteur of Lille. Availability of data and materials The datasets supporting the conclusions of this article are included within the article (and its additional files). Authors’ contributions EV, PP and LD conceived and coordinated the study and designed the protocol of investigation. CN, CM, PB, A-PB, FB, EC, GD, LL, FM, CP and MR collected samples and patient data. DES, AC, NG and GC managed and organized the collection of samples. DES, AC, NG, GC, IW, CN and FD analyzed patient data. BP performed statistical analyses. DES and AC performed molecular experiments and analyzed PCR and sequence data. EV, PP, BP, IW, FD and GC interpreted the data of the study. EV and PP drafted the manuscript. All authors provided comments and approved the final manuscript. Competing interests The authors declare that they have no competing interests. Consent for publication Not applicable. Ethics approval and consent to participate The study was approved by the research ethics committee “Comité de Protection des Personnes Sud-Est 6” (France) with the reference number 2015/CE82, which decided that informed consent from all subjects was not necessary as the experiments did not result in additional constraints for the patients. Patient data were collected anonymously (encoding the identity of patients). This study was conducted in accordance with the Code of Ethics of the World Medical Association (Declaration of Helsinki). ==== Refs References 1. Tan KS New insights on classification, identification, and clinical relevance of Blastocystis spp Clin Microbiol Rev 2008 21 639 65 10.1128/CMR.00022-08 18854485 2. Clark CG van der Giezen M Alfellani MA Stensvold CR Recent developments in Blastocystis research Adv Parasitol 2013 82 1 32 10.1016/B978-0-12-407706-5.00001-0 23548084 3. Wawrzyniak I Poirier P Viscogliosi E Meloni D Texier C Delbac F El Alaoui H Blastocystis , an unrecognized parasite: an overview of pathogenesis and diagnosis Ther Adv Infect Dis 2013 1 167 78 25165551 4. Alfellani MA Stensvold CR Vidal-Lapiedra A Onuoha ES Fagbenro-Beyioku AF Clark CG Variable geographic distribution of Blastocystis subtypes and its potential implications Acta Trop 2013 126 11 8 10.1016/j.actatropica.2012.12.011 23290980 5. Bart A Wentink-Bonnema EMS Gilis H Verhaar N Wassenaar CJA van Vugt M Goorhuis A van Gool T Diagnosis and subtype analysis of Blastocystis sp. in 442 patients in a hospital setting in the Netherlands BMC Infect Dis 2013 13 389 10.1186/1471-2334-13-389 23972160 6. Ramirez JD Sanchez LV Bautista DC Corredor AF Florez AC Stensvold CR Blastocystis subtypes detected in humans and animals from Colombia Infect Genet Evol 2014 22 223 8 10.1016/j.meegid.2013.07.020 23886615 7. El Safadi D Gaayeb L Meloni D Cian A Poirier P Wawrzyniak I Delbac F Dabboussi F Delhaes L Seck M Hamze M Riveau G Viscogliosi E Children of Senegal River Basin show the highest prevalence of Blastocystis sp. ever observed worldwide BMC Infect Dis 2014 14 164 10.1186/1471-2334-14-164 24666632 8. Suresh K Smith H Tan T Viable Blastocystis cysts in Scottish and Malaysian sewage samples Appl Environ Microbiol 2005 71 5619 20 10.1128/AEM.71.9.5619-5620.2005 16151162 9. Poirier P Wawrzyniak I Albert A El AH Delbac F Livrelli V Development and evaluation of a real-time PCR assay for detection and quantification of Blastocystis parasites in human stool samples: prospective study of patients with hematological malignancies J Clin Microbiol 2011 49 975 83 10.1128/JCM.01392-10 21177897 10. Alfellani MA Taner-Mulla D Jacob AS Imeede CA Yoshikawa H Stensvold CR Clark CG Genetic diversity of Blastocystis in livestock and zoo animals Protist 2013 164 497 509 10.1016/j.protis.2013.05.003 23770574 11. Parkar U Traub RJ Vitali S Elliot A Levecke B Robertson I Geurden T Steele J Drake B Thompson RCA Molecular characterization of Blastocystis isolates from zoo animals and their animal-keepers Vet Parasitol 2010 169 8 17 10.1016/j.vetpar.2009.12.032 20089360 12. Denoeud F Roussel M Noël B Wawrzyniak I Da Silva C Diogon M Viscogliosi E Brochier-Armamet C Couloux A Poulain J Segurans B Anthouard V Texier C Blot N Poirier P Ng GC Tan KSW Antiguenave F Jaillon O Aury J Delbac F Wincker P Vivarès CP El Alaoui H Genome sequence of the stramenopile Blastocystis , a human anaerobic parasite Genome Biol 2011 12 R29 10.1186/gb-2011-12-3-r29 21439036 13. Mirza H Wu Z Teo JDW Tan KSW Statin pleiotropy prevents rho kinase-mediated intestinal epithelial barrier compromise induced by Blastocystis cysteine proteases Cell Microbiol 2012 14 1474 84 10.1111/j.1462-5822.2012.01814.x 22587300 14. Ajjampur SSR, Tan KSW. Pathogenic mechanisms in Blastocystis spp. – Interpreting results from in vitro and in vivo studies. Parasitol Int. 2016, http://dx.doi.org/10.1016/j.parint.2016.05.007. 15. Wu Z Mirza H Tan KSW Intra-subtype variation in enteroadhesion accounts for differences in epithelial barrier disruption and is associated with metronidazole resistance in Blastocystis subtype-7 PLoS Negl Trop Dis 2014 8 e2885 10.1371/journal.pntd.0002885 24851944 16. Poirier P Wawrzyniak I Vivares CP Delbac F El Alaoui H New insights into Blastocystis spp.: a potential link with irritable bowel syndrome PLoS Pathog 2012 8 e1002545 10.1371/journal.ppat.1002545 22438803 17. Nourrisson C Scanzi J Pereira B NkoudMango C Wawrzyniak I Cian A Viscogliosi E Livrelli V Delbac F Dapoigny M Poirier P Blastocystis is associated with decrease of fecal microbiota protective bacteria: comparative analysis between patients with irritable bowel syndrome and control subjects PLoS One 2014 9 e111868 10.1371/journal.pone.0111868 25365580 18. Longstreth GF Thompson WG Chey WD Houghton LA Mearin F Spiller RC Functional bowel disorders Gastroenterology 2006 130 1480 91 10.1053/j.gastro.2005.11.061 16678561 19. Pinel C Réjasse C Picot S Brenier-Pinchart MP Grillot R Ambroise-Thomas P Blastocystis hominis : réfléxions épidémiologiques et cliniques à propos de plus de 3500 examens coprologiques Ann Biol Clin (Paris) 1999 57 601 4 10518063 20. Bourée P Blastocystis : commensal ou pathogène? Etude de 590 cas et revue de la littérature Antibiotiques 2007 9 20 4 10.1016/S1294-5501(07)88763-2 21. Souppart L Sanciu G Cian A Wawrzyniak I Delbac F Capron M Dei-Cas E Boorom K Delhaes L Viscogliosi E Molecular epidemiology of human Blastocystis isolates in France Parasitol Res 2009 105 413 21 10.1007/s00436-009-1398-9 19290540 22. Stensvold CR Arendrup MC Jespersgaard C Molbak K Nielsen HV Detecting Blastocystis using parasitologic and DNA-based methods: a comparative study Diagn Microbiol Infect Dis 2007 59 303 7 10.1016/j.diagmicrobio.2007.06.003 17913433 23. Masucci L Graffeo R Bani S Bugli F Boccia S Nicolotti N Fiori B Fadda G Spanu T Intestinal parasites isolated in a large teaching hospital, Italy, 1 May 2006 to 31 December 2008 Euro Surveill 2011 16 24 24. Gonzalez-Moreno O Domingo L Teixidor J Gracenea M Prevalence and associated factors of intestinal parasitisation: a cross-sectional study among outpatients with gastrointestinal symptoms in Catalonia, Spain Parasitol Res 2011 108 87 93 10.1007/s00436-010-2044-2 20862495 25. Windsor JJ Macfarlane L Hughes-Thapa G Jones SK Whiteside TM Incidence of Blastocystis hominis in faecal samples submitted for routine microbiological analysis Br J Biomed Sci 2002 59 154 7 12371057 26. Suresh K Smith H Comparison of methods for detecting Blastocystis hominis Eur J Clin Microbiol Infect Dis 2004 23 509 11 10.1007/s10096-004-1123-7 15168139 27. Haider SS Baqai R Qureshi FM Boorom K Blastocystis spp., Cryptosporidium spp., and Entamoeba histolytica exhibit similar symptomatic and epidemiological patterns in healthcare-seeking patients in Karachi Parasitol Res 2012 111 1357 68 10.1007/s00436-012-2972-0 22763702 28. Amin OM The epidemiology of Blastocystis hominis in the United States Res J Parasitol 2006 1 1 10 10.3923/jp.2006.1.10 29. Ithoi I Jali A Mak JW Sulaiman WYW Mahmud R Occurrence of Blastocystis in water of two rivers from recreational areas in Malaysia J Parasitol Res 2011 2011 123916 10.1155/2011/123916 21772980 30. Nimri LF Evidence of an epidemic of Blastocystis hominis infections in preschool children in northern Jordan J Clin Microbiol 1993 31 2706 8 8253970 31. Dagci H Kurt Ö Demirel M Östan I Azizi NR Mandiracioglu A Yurdagül C Tanyüksel M Eroglu C Ak M The prevalence of intestinal parasites in the province of Izmir, Turkey Parasitol Res 2008 103 839 45 10.1007/s00436-008-1065-6 18604653 32. Pipatsatitpong D Rangsin R Leelayoova S Naaglor T Mungthin M Incidence and risk factors of Blastocystis infection in an orphanage in Bangkok, Thailand Parasit Vectors 2012 5 37 10.1186/1756-3305-5-37 22330427 33. Cirioni O Giacometti A Drenaggi D Ancarani F Scalise G Prevalence and clinical relevance of Blastocystis hominis in diverse patient cohorts Eur J Epidemiol 1999 15 389 93 10.1023/A:1007551218671 10414382 34. Hailemariam G Kassu A Abebe G Abate E Damte D Mekonnen E Ota F Intestinal parasitic infections in HIV/AIDS and HIV seronegative individuals in a teaching hospital, Ethiopia Jpn J Infect Dis 2004 57 41 3 15118206 35. Alemu A Shiferaw Y Getnet G Yalew A Addis Z Opportunistic and other intestinal parasites among HIV/AIDS patients attending Gambi higher clinic in Bahir Dar city, North West Ethiopia Asian Pac J Trop Dis 2011 4 661 5 10.1016/S1995-7645(11)60168-5 36. Kurniawan A Karyadi T Dwintasari SW Sari IP Yunihastuti E Djauzi S Smith HV Intestinal parasitic infections in HIV/AIDS patients presenting with diarrhoea in Jakarta, Indonesia Trans R Soc Trop Med Hyg 2009 103 892 8 10.1016/j.trstmh.2009.02.017 19327806 37. Stensvold CR Nielsen SD Badsberg J-H Engberg J Friis-Møller N Nielsen SS Nielsen HV Friis-Moller A The prevalence and clinical significance of intestinal parasites in HIV-infected patients in Denmark Scand J Infect Dis 2011 43 129 35 10.3109/00365548.2010.524659 20936912 38. Adamu H Wegayehu T Petros B High prevalence of diarrhoegenic intestinal parasite infections among non-ART HIV patients in Fitche Hospital, Ethiopia PLoS One 2013 8 e72634 10.1371/journal.pone.0072634 23991132 39. Tasova Y Sahin B Koltas S Paydas S Clinical significance and frequency of Blastocystis hominis in Turkish patients with hematological malignancy Acta Med Okayama 2000 54 133 6 10925738 40. Cekin AH Cekin Y Adakan Y Tasdemir E Koclar FG Yolcular BO Blastocystosis in patients with gastrointestinal symptoms: a case-control study BMC Gastroenterol 2012 12 122 10.1186/1471-230X-12-122 22963003 41. Petersen AM Stensvold CR Mirsepasi H Engberg J Friis-Møller A Porsbo LJ Hammerum AM Nordgaard-Lassen I Nielsen HV Krogfelt KA Active ulcerative colitis associated with low prevalence of Blastocystis and Dientamoeba fragilis infection Scand J Gastroenterol 2013 48 638 9 10.3109/00365521.2013.780094 23528075 42. Mattiucci S Crisafi B Gabrielli S Paoletti M Cancrini G Molecular epidemiology and genetic diversity of Blastocystis infection in humans in Italy Epidemiol Infect 2016 144 635 46 10.1017/S0950268815001697 26194649

PMC005xxxxxx/PMC5002210.txt

==== Front BMC Pregnancy ChildbirthBMC Pregnancy ChildbirthBMC Pregnancy and Childbirth1471-2393BioMed Central London 103610.1186/s12884-016-1036-3ErratumErratum to: CONCEPTT: Continuous Glucose Monitoring in Women with Type 1 Diabetes in Pregnancy Trial: A multi-center, multi-national, randomized controlled trial - Study protocol Feig Denice S. d.feig@utoronto.ca 12317Asztalos Elizabeth 413Corcoy Rosa 56De Leiva Alberto 56Donovan Lois 7Hod Moshe 8Jovanovic Lois 9Keely Erin 10Kollman Craig 11McManus Ruth 12Murphy Kellie 12Ruedy Katrina 11Sanchez J. Johanna 413Tomlinson George 14Murphy Helen R. 1516on behalf of the CONCEPTT Collaborative Group 1 Mt. Sinai Hospital, Toronto, Ontario Canada 2 Lunenfeld-Tanenbaum Research Institute, Toronto, Ontario Canada 3 Department of Medicine, University of Toronto, Toronto, Canada 4 Sunnybrook Health Sciences Centre, Toronto, Ontario Canada 5 Hospital de la Santa Creu i Sant Pau, Barcelona, Spain 6 CIBER-BBN, Zaragoza, Spain 7 University of Calgary, Calgary, Alberta Canada 8 Helen Schneider Hospital for Women, Rabin Medical Center, Petah Tikva, Israel 9 30-1 Barranca Avenue, Santa Barbara, CA USA 10 The Ottawa Hospital, Riverside Campus, Ottawa, Ontario Canada 11 Jaeb Center For Health Research, Tampa, FL USA 12 St. Joseph Health Care London, London, Ontario Canada 13 Sunnybrook Research Institute, Toronto, Ontario Canada 14 University Health Network, Toronto General Hospital, Toronto, Ontario Canada 15 Institute of Metabolic Science, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK 16 Norwich Medical School, University of East Anglia, Norwich, UK 17 Department of Medicine, Leadership Sinai Centre for Diabetes, Mt. Sinai Hospital, 60 Murray St. #5027, Toronto, Ontario M5T 3L9 Canada 26 8 2016 26 8 2016 2016 16 1 24918 8 2016 18 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.issue-copyright-statement© The Author(s) 2016 ==== Body Erratum After publication of the original article [1], it came to the authors’ attention that an incorrect affiliation was inadvertently added in the Acknowledgements section for the CONCEPTT Collaborative Group. The authors would like to amend the following statement in the CONCEPTT Collaborative Group section as follows: The correct affiliation for Julia Lowe and Anna Rogowsky should read Sunnybrook Health Sciences Centre, Toronto. The online version of the original article can be found under doi:10.1186/s12884-016-0961-5. ==== Refs Reference 1. Feig DS Asztalos E Corcoy R De Leiva A Donovan L Hod M Jovanovic L Keely E Kollman C McManus R Murphy K Ruedy K Sanchez JJ Tomlinson G Murphy HR on behalf of the CONCEPTT Collaborative Group CONCEPTT: Continuous Glucose Monitoring in Women with Type 1 Diabetes in Pregnancy Trial: A multi-center, multi-national, randomized controlled trial - Study protocol BMC Pregnancy Childbirth 2016 16 167 10.1186/s12884-016-0961-5 27430714

PMC005xxxxxx/PMC5002211.txt

==== Front Health Qual Life OutcomesHealth Qual Life OutcomesHealth and Quality of Life Outcomes1477-7525BioMed Central London 52510.1186/s12955-016-0525-4ResearchAssociation of knee pain and different definitions of knee osteoarthritis with health-related quality of life: a population-based cohort study in southern Sweden Kiadaliri Aliasghar A. aliasghar.ahmad_kiadaliri@med.lu.se 1210Lamm Carl Johan 3de Verdier Maria Gerhardsson 4Engström Gunnar 5Turkiewicz Aleksandra 1Lohmander L. Stefan 167Englund Martin 1891 Clinical Epidemiology Unit, Orthopaedics, Department of Clinical Sciences-Lund, Lund University, Lund, Sweden 2 Health Services Management Research Center, Institute for Futures Studies in Health, Kerman University of Medical Sciences, Kerman, Iran 3 Astra Zeneca R&D, Mölndal, Sweden 4 Astra Zeneca R&D, Gothenburg, Sweden 5 Cardiovascular Epidemiology, Department of Clinical Sciences, Lund University, Malmö, Sweden 6 Research Unit for Musculoskeletal Function and Physiotherapy, University of Southern Denmark, Odense, Denmark 7 Department of Orthopedics and Traumatology, University of Southern Denmark, Odense, Denmark 8 Epidemiology and Register Centre South, Skåne University Hospital, Lund, Sweden 9 Clinical Epidemiology Research and Training Unit, Boston University School of Medicine, Boston, MA USA 10 Skånes University Hospital, Clinical Epidemiology Unit, Klinikgatan 22, SE-221 85 Lund, Sweden 26 8 2016 26 8 2016 2016 14 1 1219 2 2016 23 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background While the impact of knee pain and knee osteoarthritis (OA) on health-related quality of life (HRQoL) has been investigated in the literature, there is a lack of knowledge on the impact of different definitions of OA on HRQoL. The main aim of this study was to measure and compare the impact of knee OA and its different definitions on HRQoL in the general population. Methods A random sample of 1300 participants from Malmö, Sweden with pain in one or both knees in the past 12 months with duration ≥4 weeks and 650 participants without were invited to clinical and radiographic knee examination. A total of 1527 individuals with a mean (SD) age 69.4 (7.2) participated and responded to both generic (EQ-5D-3L) and disease-specific (the Knee injury and Osteoarthritis Outcome Score) questionnaires. Knee pain was defined as pain during the last month during most of the days. Knee OA was defined radiographically (equivalent to Kellgren and Lawrence grade ≥2) and clinically according to the American College of Rheumatology (ACR) criteria. Results Of participants with either knee pain or knee OA or both, 7 % reported no problem for the EQ-5D-3L attributes. The corresponding proportion among references (neither knee pain nor OA) was 42 %. The participants with knee pain and OA had all HRQoL measures lower compared to those with knee pain but no OA. The ACR clinical definition of knee OA was associated with lower HRQoL than the definition based on radiographic knee OA (adjusted difference −0.08 in UK EQ-5D-3L index score). Conclusions Applying different definitions of knee OA result in different levels of HRQoL and this is mainly explained by the knee pain experience. These differences may lead to discrepant conclusions from cost-utility analyses. Keywords EQ-5D-3LKnee osteoarthritisKnee painKOOSQuality of lifeSwedenAstraZeneca SwedenSwedish Research CouncilCarfoord FoundationThe King Gustaf V 80-year Birthday FundKock Foundationsissue-copyright-statement© The Author(s) 2016 ==== Body Background The prevalence of knee osteoarthritis (OA) rapidly increases with age [1, 2]. Thus, it is expected that the number of people with knee OA will increase in the future due to the steadily aging population and increasing prevalence of obesity [3, 4]. The Global Burden of Disease (GBD) 2010 study ranked hip and knee OA as the 11th leading cause of years lived with disability (YLD) and the 38th highest contributor to disability-adjusted life years (DALYs) among 291 conditions [5]. The share of total DALYs attributed to hip and knee OA increased from 0.49 % in 1990 to 0.69 % in 2010 (an estimated 64 % increase) [5]. OA is associated with substantial deteriorations in health-related quality of life (HRQoL) [6]. A recent study in the UK ranked OA as the 3rd greatest contributor to loss of HRQoL among eleven long-standing health conditions [7]. HRQoL constitutes a subjective and multidimensional concept that includes the physical, psychological, and social functioning related to a health condition or therapy [8]. HRQoL can be used as a criterion to predict future health care consumption among people with OA [9]. Therefore, measuring HRQoL is important not only to quantify the effects of disease and its treatments, but also to aid informed decision-making in allocation of often limited healthcare resources. The relatively weak association between clinical symptoms of OA and radiographic evidence of the disease has contributed to the existence of several definitions of OA for study purposes. The multiple definitions represent a major challenge in studies on OA [10]. A radiographic definition of OA (most commonly using Kellgrene-Lawrence scale [11] with the cut-off of grade 2 or worse) and the clinical definition proposed by the American College of Rheumatology (ACR) [12] are two most commonly used definitions in the literature. While the impact of knee pain and knee OA on HRQoL has been investigated in the literature [13–15], there is a lack of knowledge on the impact of different definitions of OA on HRQoL. If such differences exist, the different OA definitions used in health economic evaluations may potentially yield different results and conclusions. Thus, in the present study our main aim was to evaluate and compare the impact of these two definitions of knee OA on HRQoL. These were based on patient-reported knee pain, and/or radiographically evident knee OA, as well as clinically defined knee OA. We related our findings to knee-healthy reference participants drawn from the same source population. Method Setting and participants The Malmö OA study (MOA) originated from the Malmö Diet and Cancer Study (MDCS) cohort established between 1991 and 1996 [16]. All men aged 45–73 years and women aged 44–74 years at the time of enrolment living in the city of Malmö were invited to participate in the MDCS (n = 74 138). A cohort of 28 098 had completed baseline examinations [17]. In the year 2007 a postal questionnaire about knee pain was sent to a 10 000 random sample from the MDCS who completed the MDCS baseline examination and were still alive and resident in the Malmö area. Respondents answered a question about whether they had knee pain during the previous 12 months and its duration (<1 week, 1–4 weeks, 1–3 months, >3 months). In the second stage, a random sample of 1300 participants with pain in one or both knees in the past 12 months and duration of at least 4 weeks (group A) and 650 participants without knee pain or with knee pain for shorter duration (group B) were invited to a clinical visit and radiographic examination [16, 18]. Of these, 1028 (79 %) people from group A and 499 (77 %) people from group B participated and were included in the present study. Radiographic evaluation and knee OA definitions Both knees were radiographed in a weight-bearing and semi-flexed position (knees in 10–15° of flexion) using a posterior-anterior beam direction (film focus distance 110 cm, 60 kV and 10 mA) with the aid of fluoroscopy to optimally align the tibia plateau. An independent senior radiologist specialized in musculoskeletal conditions who was blinded to the clinical data assessed joint space narrowing and osteophytes according to the atlas from the Osteoarthritis Research Society International [16, 19]. We classified a knee as having radiographic knee OA if one or more of the following criteria were fulfilled in either the medial, lateral tibiofemoral compartment or patellofemoral compartment: joint space narrowing grade 2 or worse, the sum of marginal osteophyte grades in the same compartment 2 or worse, joint space narrowing grade 1 and osteophyte grade 1 in the same compartment (approximating Kellgren and Lawrence (KL) grade 2 or worse) [20]. All participants answered the 5 questions included in the ACR clinical criteria for knee OA. We classified a participant as having clinical OA if fulfilling ACR clinical criteria according to the recursive partitioning method [12]. Five items are included in the ACR criteria (28): (a) knee pain for most days of the prior month, (b) crepitus on active joint motion, (c) morning stiffness of duration < 30 min, (d) age ≥ 38 years, and (e) bony enlargement of the knee on examination. Knee OA was present if items “a”, “b”, “c”, “d” or items “a”, “b”, “e” or items “a” and “e” were present [12]. We used the ACR clinical criteria “a” (i.e. knee pain for most days of the prior month) to define knee pain in our study. It should be noted that as this definition was different from the definition of knee pain in the first stage and also due to gap between the first and second stages of the study, people who defined as having knee pain in the first stage could have been classified as not having knee pain and vice versa. We classified participants into four exposure groups based on their clinical and radiographic knee status: 1) reference group having neither knee pain nor radiographic or clinically-defined knee OA (n = 744), 2) knee pain without OA, i.e., participants with knee pain but without radiographic or clinically-defined knee OA (n = 169), 3) knee pain with OA, i.e., participants with knee pain having either radiographic knee OA or clinically-defined knee OA (n = 402), and 4) radiographic knee OA but no knee pain, i.e. participants without knee pain but fulfilling the radiographic defined knee OA (n = 186). A total of 26 participants with missing value on knee OA status were excluded. Additionally, to assess the impact of different definition of knee OA on HRQoL, the participants with knee OA (i.e., groups 3 & 4) were collapsed, then classified into three subgroups: those fulfilling the criteria for a) radiographic knee OA (n = 282), b) clinical knee OA (n = 157), and c) both clinical and radiographic knee OA (n = 149). Health-related quality of life measurement The EQ-5D-3L is a generic multi-attribute instrument to elicit health-related preferences. The EQ-5D-3L covers five attributes: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each attribute has three levels: no problems, moderate problems, and severe problems, resulting in 243 (35) possible health states [21]. The responses to these attributes are weighted based on the preferences elicited from a general population/patients sample to calculate an index score. We used the UK [22] and recently developed Swedish [23] sets of preferences to calculate the index score. The UK EQ-5D-3L scores range between −0.594 and 1 (full health), while the Swedish scores range between 0.340 and 0.969. The Knee injury and Osteoarthritis Outcome Score (KOOS) is a validated knee specific instrument [24, 25]. The KOOS is a 42-item self-administered questionnaire covering 5 subscales: pain, other symptoms, function in daily living (ADL), function in sport and recreation (Sport/Rec) and knee related quality of life (QoL). All items have five possible answer options scored from 0 (no problems) to 4 (extreme problems). A normalized score (100 indicating no symptoms and 0 indicating extreme symptoms) is calculated for each subscale. As our participants were an elderly population, we included a sixth answer option (not applicable) into the case report form for the subscale Sport/Rec. If the box “not applicable” was marked, the item was treated as missing data. Since a large number of participants selected this option, we decided not to include this subscale in our analysis. Explanatory variables (confounders) To avoid potential confounder bias, the following explanatory variables were included in our regression analysis: sex, age, body mass index (BMI), comorbidity, years of education, employment, and smoking. Age and BMI were included as continuous variables. Comorbidity was defined as presence of self-reported doctor’s diagnosis of one or more of the following comorbidities: back problems, other joint problems (except knees), asthma/lung disease, hypertension, heart disease, stroke, leg artery disease, neurological disease, diabetes, cancer, gastric ulcer, renal disease, anemia/blood disease, eye disease/visual impairment, balance disorder, depression, and other psychiatric disorder. Three groups were defined: no comorbidity, single comorbidity, and multiple comorbidities. Years of education were categorized in three groups: ≤9 years, 9–12 years, and >12 years. Employment was defined in three categories: employed, unemployed, and retired. We grouped smoking in three groups as: never, current smoker, and ex-smoker. Statistical analysis We applied weighting to account for a possible selection bias that might arise from non-responses in the first or second part in the MOA study [26]. A logistic regression model with sex, age on 1 January 2007, education, smoking, and BMI as covariates was used to estimate the probability of response in the first stage of survey. Similar logistic regression models including knee pain during last 12 months as an additional covariate were applied to estimate the probability of participation and attendance at the clinical examination. The sampling weights (the reciprocal of the sampling probability for those with and without knee pain ≥ 4 weeks duration) were multiplied by the weights for response, participation and attendance to construct the final weights used in analyses. We present continuous variables using means and standard deviations. The mean differences between groups were tested using one-way analysis of variance (ANOVA) and Bonferroni post-hoc tests. Proportions were compared using Pearson Chi-2 test. Adjusted analysis was conducted using Ordinary least squares regression with robust standard errors. Due to the skewed distribution of the EQ-5D-3L data, several different methods have been applied to these data in the literature [27, 28], but ordinary least squares (OLS) regression is considered as a simple and valid approach [29]. The linearity of the continuous variables was checked using design variables and residual plots and non-linearity was modelled using fractional polynomial. The possible reasonable linear and non-linear interactions were also checked using the same approach [30]. Potential confounders (age, sex, body mass index, comorbidity, years of education, employment, and smoking) were included in the models regardless of their significance level. STATA version 13 (StataCorp LP, College Station, TX, USA) was used for statistical analysis. Results The mean (SD) age and BMI of the participants included was 69.4 (7.2) and 27.7 (4.9), respectively, and 63.8 % were women (Table 1). Except sex and smoking, there were no statistically significant differences in other explanatory variables across the study groups.Table 1 Characteristics of participants included in the Malmö osteoarthritis study (MOA) Reference group (n = 744) Knee pain without knee OA (n = 169) Knee pain with knee OA (n = 402) Radiographic knee OA without pain (n = 186) Women, % 63.0 70.4 63.7 61.3 Age, years (SD) 68.7 (7.2) 68.4 (7.3) 70.1 (7.2) 71.5 (6.7) BMI (SD) 26.7 (4.3) 27.8 (4.8) 28.9 (5.5) 28.8 (5.1) Comorbidity, %  Single 32.5 26.6 23.1 28.0  Multiple 46.1 59.2 64.9 53.2 Years of education, %   ≤ 9 years 35.0 30.3 43.3 34.4   > 9 years &  ≤ 12 years 38.3 43.2 34.7 44.8  >12 years 26.7 26.5 22.0 20.8 Smoking, %  Current smoker 14.1 12.7 12.3 11.8  Ex-smoker 45.1 46.4 46.0 42.5 Employment, %  Employed 25.3 23.7 14.7 15.6  Retired 72.2 72.8 80.9 81.7 The mean UK EQ-5D-3L index scores were 0.69 (95 % CI 0.67 to 0.70) and 0.85 (95 % CI 0.83 to 0.86) in non-reference and reference groups, respectively. Across non-reference groups, the participants with radiographic knee OA without knee pain had generally less problems on the EQ-5D-3L dimensions and reported statistically significantly higher HRQoL scores compared with the participants who had knee pain (Table 2). Among the participants with knee pain, the participants with both knee pain and knee OA reported lower HRQoL scores.Table 2 Percentage (95 % CI) of participants with moderate/severe problems in the EQ-5D-3L attributes, mean (95 % CI) EQ-5D-3L index and KOOS scores in the full sample Reference group (n = 744) Knee pain without knee OA (n = 169) Knee pain with knee OA (n = 402) Radiographic knee OA without knee pain (n = 186) EQ-5D-3L dimension  Mobility 11.5 (8.9 to 14.8) 32.6 (23.1 to 43.8)a 52.2 (44.8 to 59.5)ab 34.7 (25.4 to 45.5)ac  Self-care 0.2 (0.1 to 0.5) 2.7 (1.3 to 5.6)a 7.3 (3.5 to 14.5)a 3.6 (1.0 to 12.3)a  Usual activities 5.6 (3.9 to 8.0) 21.9 (14.3 to 32.0)a 24.8 (20.1 to 30.3)a 10.0 (5.1 to 18.7)bc  Pain/discomfort 35.9 (31.8 to 40.3) 88.1 (76.1 to 94.5)a 96.7 (92.1 to 98.7)ab 70.4 (60.1 to 78.9)abc  Anxiety/depression 17.5 (14.3 to 21.1) 30.2 (20.9 to 41.3)a 32.2 (25.9 to 39.4)a 21.4 (13.9 to 31.4) No problem at any dimension (full health) 54.2 (50.0 to 58.6) 7.7 (2.9 to 19.0)a 3.1 (1.2 to 7.8)a 22.6 (15.2 to 32.2)abc Mean UK EQ-5D-3L index score 0.89 (0.87 to 0.90) 0.71 (0.67 to 0.75)a 0.67 (0.64 to 0.69)ab 0.77 (0.72 to 0.82)abc Mean Swedish EQ-5D-3L index score 0.93 (0.93 to 0.94) 0.87 (0.85 to 0.89)a 0.84 (0.83 to 0.85)ab 0.89 (0.87 to 0.91)abc KOOS subscale  Pain 92.8 (91.8 to 93.8) 68.9 (64.3 to 73.4)a 55.4 (52.9 to 57.8)ab 81.4 (78.2 to 84.6)abc  Symptoms 66.1 (65.4 to 66.8) 56.7 (54.4 to 58.9)a 48.2 (46.5 to 50.0)ab 61.6 (59.5 to 63.8)abc  ADL 91.3 (90.1 to 92.6) 70.3 (65.6 to 74.9)a 56.2 (53.1 to 59.3)ab 80.1 (76.4 to 83.9)abc  QoL 86.4 (84.9 to 87.9) 56.9 (51.6 to 62.1)a 40.1 (37.4 to 42.9)ab 66.4 (61.9 to 70.8)abc ADL function in daily living, QoL knee-related quality of life a P < 0.05 compared with reference group b P < 0.05 compared with knee pain without knee OA c P < 0.05 compared with knee pain with knee OA Among the participants with knee OA, those with radiographic knee OA generally experienced less problems in the EQ-5D-3L dimensions and reported statistically significantly higher HRQoL scores than the participants fulfilled clinical knee OA definition (Table 3). The participants with both radiographic and clinical knee OA reported lower scores on three KOOS subscales compared with the participants with clinically defined knee OA.Table 3 Percentage (95 % CI) of participants with moderate/severe problems in the EQ-5D-3L attributes, mean (95 % CI) EQ-5D-3L index and KOOS scores among participant with knee osteoarthritis (OA) Radiographic knee OA (n = 282) Clinical knee OA (n = 157) Clinical and radiographic knee OA (n = 149) EQ-5D-3L dimension  Mobility 40.5 (32.1 to 49.5) 38.9 (29.3 to 49.5) 59.7 (48.7 to 69.8)ab  Self-care 5.5 (2.1 to 13.6) 5.0 (2.7 to 9.4) 5.3 (2.6 to 10.5)  Usual activities 12.2 (7.6 to 18.9) 27.4 (20.2 to 36.0)a 26.5 (19.3 to 35.2)a  Pain/discomfort 76.3 (67.8 to 83.2) 98.6 (95.6 to 99.5)a 94.4 (82.2 to 98.4)a  Anxiety/depression 23.4 (16.7 to 31.8) 36.0 (25.5 to 48.1) 29.7 (22.0 to 38.7) No problem at any dimension (full health) 18.3 (12.4 to 26.1) 0.9 (0.2 to 3.5)a 5.6 (1.6 to 17.8)ab Mean UK EQ-5D-3L index score 0.75 (0.71 to 0.79) 0.67 (0.63 to 0.71)a 0.66 (0.61 to 0.70)a Mean Swedish EQ-5D-3L index score 0.88 (0.87 to 0.90) 0.84 (0.82 to 0.86)a 0.84 (0.81 to 0.86)a KOOS subscale  Pain 75.6 (72.4 to 78.9) 58.6 (54.8 to 62.4)a 52.0 (47.5 to 56.6)ab  Symptoms 59.5 (57.6 to 61.4) 48.2 (45.6 to 50.7)a 45.0 (42.4 to 47.6)a  ADL 74.6 (70.8 to 78.4) 59.5 (54.4 to 64.6)a 53.6 (49.2 to 58.1)ab  QoL 60.5 (56.5 to 64.5) 44.2 (40.1 to 48.3)a 35.9 (30.3 to 41.5)ab ADL function in daily living, QoL knee-related quality of life a P < 0.05 compared with radiographic knee OA b P < 0.05 compared with clinical knee OA Controlling for potential confounders in the regression analysis did not alter our unadjusted findings in the full sample (Table 4). Generally the outcome measures were ranked in following order: reference group > radiographic knee OA without knee pain > knee pain without knee OA > knee pain with knee OA. Among the participants with knee OA, while the participants with radiographic knee OA reported higher scores on all outcome measures, the difference between two other groups (i.e., clinical knee OA only and clinical and radiographic knee OA) was statistically significant only on the KOOS-Pain subscale.Table 4 The effects of knee osteoarthritis (OA) on health-related quality of life: results from regression analysis UK EQ-5D-3L index score Swedish EQ-5D-3L index score Pain (KOOS) Symptoms (KOOS) ADL (KOOS) QoL (KOOS) Full samplea  Reference group 0.00 (ref)  Knee pain without knee OA −0.15*** −0.06*** −22.69*** −8.78*** −19.67*** −27.41***  Knee pain with knee OA −0.19*** −0.08*** −35.14*** −16.92*** −31.94*** −43.48***  Radiographic knee OA without pain −0.09*** −0.03** −10.04*** −4.27*** −8.72*** −18.47*** Participants with knee OAa  Radiographic knee OA 0.00 (ref)  Clinical knee OA −0.08*** −0.04*** −17.92*** −10.91*** −17.40*** −18.40***  Clinical and radiographic knee OA −0.10** −0.05*** −23.79*** −14.08*** −22.50*** −24.42*** ADL function in daily living, QoL Knee-related quality of life ***,** p < 0.001, and p < 0.01 significance level compared to the reference group aBoth models were additionally adjusted for age, sex, body mass index, comorbidity, years of education, employment, and smoking Discussion We measured and compared the impact of knee pain and knee OA on HRQoL, using both generic and disease-specific scales, in a large population-based cohort from southern Sweden. First, our results confirmed the expected notion that participants with either knee pain or knee OA (either clinically or radiographically defined) reported lower HRQoL scores than reference participants free of such knee pain and with no knee OA. We also found that knee pain without knee OA had more profound negative impact on HRQoL than radiographic knee OA without knee pain. Moreover, among participants with knee OA (either radiographically or clinically defined), those with clinical knee OA according to the ACR clinical definition reported lower scores of HRQoL than those who had radiographic knee OA not fulfilling the ACR clinical criteria. The effect of different OA definitions in the study of prevalence and incidence of OA is well documented, where it was reported that radiographic definition of OA resulted in the highest prevalence estimates [10]. The effect of different OA definitions on HRQoL is however much less well investigated, which may have important implications for the interpretation of findings from health economic evaluations. Our results suggest that the knee OA definition based on the ACR clinical criteria was associated with lower HRQoL than the definition based on radiographic knee OA. The presence of knee pain in the ACR clinical definition is a possible explanation for this finding. Interestingly, in our full sample, people with knee pain had lower HRQoL compared with people with radiographic features of knee OA without knee pain. In our sample, only 96 (22 %) of participants in the radiographic knee OA group experienced pain in most days during last month and interestingly the EQ-5D-3L index score for these 96 participants (0.66) was equal to participants in the clinical knee OA group (0.66). Our finding of differences in HRQoL measures between persons fulfilling different definitions of knee OA (i.e., radiographic or clinical ACR criteria) has an important implication for cost-utility analyses. The difference in the mean HRQoL according to knee OA definition will result in different estimates of quality-adjusted life years (QALYs) which might yield discrepant and potentially conflicting conclusions from these analyses. For example, in our study, if a hypothetical intervention improved HRQoL of participants with knee OA to the level of the reference group (0.85), then QALYs gained from using clinical knee OA to define participants with knee OA (0.85 minus 0.66) would be 1.9 times higher than applying radiographic signs for definition of knee OA (0.85 minus 0.75). This translates into less favorable cost-effectiveness ratios using radiographic signs to define knee OA. This should be taken into account by policy makers when using cost-utility analysis to make decisions on funding of knee OA interventions. In line with previous studies [31–33], we found that pain was the most affected dimension in the EQ-5D-3L questionnaires among participants with knee pain and knee OA. In our study the participants suffering from both knee pain and knee OA reported lower scores for all KOOS subscales compared to people with either knee pain or radiographic knee OA only. In addition, among participants with knee OA, radiographic knee OA without pain was associated with higher HRQoL than radiographic knee OA with pain. These findings suggest that the combination of knee pain and knee OA results in additional negative impact on HRQoL compared with knee pain or radiographic knee OA alone. Mean EQ-5D-3L index score among participants with knee pain and knee OA in our study (0.65) was higher than people with knee OA in Singapore (0.49) [31] and UK (0.44) [34] using the UK weights. Differences in severity of knee OA, with a substantial portion of participants in those studies having severe disease and waiting for TKR might be an explanation. However, our EQ-5D-3L score is comparable to values reported among Swedish people waiting for total hip replacement (0.73) [35] and ACL surgery (0.69) [36]. This similarity between Swedish patients with different severity highlights another potential explanation for higher scores among the Swedish patients than patients in other countries, i.e., clinical, environmental, organizational, and cultural differences between patients and countries in these studies. The results of this study should be interpreted in light of some limitations. The MOA study originated from the MDCS whose participants were shown to have a slightly lower mortality than non-participants [17]. Eleven percent of people with knee pain in the past 12 months and 30 % of participants without knee pain in past 12 month did not agree to participate in our study. If non-participation was associated with both knee pain/knee OA and HRQoL through unmeasured factors in our weighting exercise, then this is a potential source of selection bias and also could limit the generalizability of our findings. Another limitation of the current study is the self-reported nature of some explanatory variables (e.g., self-reported doctor-diagnosed comorbidity and smoking) that might be prone to recall bias or underreporting. The cross-sectional design of our study implies that any causal inference should be avoided. Conclusion The current study showed that participants with knee pain (with and without knee OA) have poorer HRQoL, measured by both generic and disease-specific scales, than the references. The presence of knee OA has additional negative impact on HRQoL above the knee pain alone. Importantly, we found that applying different definitions of knee OA resulted in different levels of HRQoL and this was mainly explained by the knee pain experience. These differences are important to take into account when assessing the impact of knee OA on HRQoL and interpreting the results of cost-utility analyses. Acknowledgements We would like to thank the Malmö Diet Cancer Study cohort steering committee and their data managers for their assistance with the data transfer, and Dr Inga Redlund-Johnell for scoring of the MOA radiographs. Funding This work was supported by Astra Zeneca, the Swedish Research Council, Crafoord Foundation, The King Gustaf V 80-year Birthday Fund, Kock Foundations, the Faculty of Medicine Lund University, Governmental Funding of Clinical Research within National Health Service (ALF) and Region Skåne. The funding sources had no influence on the study design, collection, analysis and interpretation of data, in the writing the manuscript, or in the decision to submit the manuscript for publication. Availability of data and materials The dataset supporting the conclusions of this article will not be available in a public repository because consent/approval was not obtained for the sharing of subject data from participants or the Regional Ethics Committee in Lund. Authors’ contributions AAK participated in the design, analysis, and interpretation of results and drafting the manuscript. CJL, MGdV, SL and GE participated in conception of the study, acquisition of data, and revision of the manuscript for important intellectual content. AT and ME participated in acquisition of data, interpretation of results, and revision of the manuscript for important intellectual content. All authors approved the final manuscript. Competing interest MGdV is employed by AstraZeneca and is an owner of AstraZeneca shares as part of the bonus system. GE and CJL were formerly employed by AstraZeneca R&D. All other authors have declared no conflicts of interest. Consent for publication Not applicable. Ethics approval and consent for participation The study was approved by the Regional Ethics Committee in Lund and informed consent for participation was obtained from all participants in accordance with the Declaration of Helsinki. ==== Refs References 1. Felson DT The epidemiology of knee osteoarthritis: results from the Framingham Osteoarthritis Study Semin Arthritis Rheum 1990 20 42 50 10.1016/0049-0172(90)90046-I 2287948 2. Busija L Bridgett L Williams SR Osborne RH Buchbinder R March L Osteoarthritis Best Pract Res Clin Rheumatol 2010 24 757 68 10.1016/j.berh.2010.11.001 21665124 3. Reijman M Pols HA Bergink AP Hazes JM Belo JN Lievense AM Body mass index associated with onset and progression of osteoarthritis of the knee but not of the hip: the Rotterdam Study Ann Rheum Dis 2007 66 158 62 10.1136/ard.2006.053538 16837490 4. Zhang Y Jordan JM Epidemiology of osteoarthritis Clin Geriatr Med 2010 26 355 69 10.1016/j.cger.2010.03.001 20699159 5. Cross M Smith E Hoy D Nolte S Ackerman I Fransen M The global burden of hip and knee osteoarthritis: estimates from the global burden of disease 2010 study Ann Rheum Dis 2014 73 1323 30 10.1136/annrheumdis-2013-204763 24553908 6. Farr Ii J Miller LE Block JE Quality of life in patients with knee osteoarthritis: a commentary on nonsurgical and surgical treatments Open Orthop J 2013 7 619 23 10.2174/1874325001307010619 24285987 7. Wu M Brazier JE Kearns B Relton C Smith C Cooper CL Examining the impact of 11 long-standing health conditions on health-related quality of life using the EQ-5D in a general population sample Eur J Health Econ 2014 16 2 141 51 10.1007/s10198-013-0559-z 24408476 8. Revicki DA Health-related quality of life in the evaluation of medical therapy for chronic illness J Fam Pract 1989 29 377 80 2677212 9. Ethgen O Kahler KH Kong SX Reginster JY Wolfe F The effect of health related quality of life on reported use of health care resources in patients with osteoarthritis and rheumatoid arthritis: a longitudinal analysis J Rheumatol 2002 29 1147 55 12064827 10. Pereira D Peleteiro B Araujo J Branco J Santos RA Ramos E The effect of osteoarthritis definition on prevalence and incidence estimates: a systematic review Osteoarthritis Cartilage 2011 19 1270 85 10.1016/j.joca.2011.08.009 21907813 11. KELLGREN JH LAWRENCE JS Radiological assessment of osteo-arthrosis Ann Rheum Dis 1957 16 4 494 502 10.1136/ard.16.4.494 13498604 12. Altman R Asch E Bloch D Bole G Borenstein D Brandt K Development of criteria for the classification and reporting of osteoarthritis. Classification of osteoarthritis of the knee. Diagnostic and Therapeutic Criteria Committee of the American Rheumatism Association Arthritis Rheum 1986 29 1039 49 10.1002/art.1780290816 3741515 13. Hill CL Parsons J Taylor A Leach G Health related quality of life in a population sample with arthritis J Rheumatol 1999 26 2029 35 10493687 14. Picavet HS Hoeymans N Health related quality of life in multiple musculoskeletal diseases: SF-36 and EQ-5D in the DMC3 study Ann Rheum Dis 2004 63 723 9 10.1136/ard.2003.010769 15140781 15. Muraki S Akune T Oka H En-yo Y Yoshida M Saika A Association of radiographic and symptomatic knee osteoarthritis with health-related quality of life in a population-based cohort study in Japan: the ROAD study Osteoarthritis Cartilage 2010 18 1227 34 10.1016/j.joca.2010.06.001 20633679 16. Turkiewicz A Gerhardsson de Verdier M Engstrom G Nilsson PM Mellstrom C Lohmander LS Prevalence of knee pain and knee OA in southern Sweden and the proportion that seeks medical care Rheumatology (Oxford) 2015 54 827 35 10.1093/rheumatology/keu409 25313145 17. Manjer J Carlsson S Elmstahl S Gullberg B Janzon L Lindstrom M The Malmo Diet and Cancer Study: representativity, cancer incidence and mortality in participants and non-participants Eur J Cancer Prev 2001 10 489 99 10.1097/00008469-200112000-00003 11916347 18. Kiadaliri AA, Gerhardsson de Verdier M, Turkiewicz A, Lohmander LS, Englund M. Socioeconomic inequalities in knee pain, knee osteoarthritis, and health-related quality of life: a population-based cohort study in southern Sweden. Scand J Rheumatol. 2016:1–9. DOI:10.1080/03009742.2016.1181203 19. Altman RD Hochberg M Murphy WA Jr Wolfe F Lequesne M Atlas of individual radiographic features in osteoarthritis Osteoarthritis Cartilage 1995 3 Suppl A 3 70 8581752 20. Englund M Roos EM Lohmander LS Impact of type of meniscal tear on radiographic and symptomatic knee osteoarthritis: a sixteen-year followup of meniscectomy with matched controls Arthritis Rheum 2003 48 2178 87 10.1002/art.11088 12905471 21. Brooks R EuroQol: the current state of play Health Policy 1996 37 53 72 10.1016/0168-8510(96)00822-6 10158943 22. Dolan P Modeling valuations for EuroQol health states Med Care 1997 35 1095 108 10.1097/00005650-199711000-00002 9366889 23. Burstrom K Sun S Gerdtham UG Henriksson M Johannesson M Levin LA Swedish experience-based value sets for EQ-5D health states Qual Life Res 2014 23 431 42 10.1007/s11136-013-0496-4 23975375 24. Roos EM Roos HP Lohmander LS Ekdahl C Beynnon BD Knee Injury and Osteoarthritis Outcome Score (KOOS)--development of a self-administered outcome measure J Orthop Sports Phys Ther 1998 28 88 96 10.2519/jospt.1998.28.2.88 9699158 25. Roos EM Roos HP Ekdahl C Lohmander LS Knee injury and Osteoarthritis Outcome Score (KOOS)--validation of a Swedish version Scand J Med Sci Sports 1998 8 439 48 10.1111/j.1600-0838.1998.tb00465.x 9863983 26. Little RJ Vartivarian S On weighting the rates in non-response weights Stat Med 2003 22 1589 99 10.1002/sim.1513 12704617 27. Hernandez Alava M Wailoo AJ Ara R Tails from the peak district: adjusted limited dependent variable mixture models of EQ-5D questionnaire health state utility values Value Health 2012 15 550 61 10.1016/j.jval.2011.12.014 22583466 28. Basu A Manca A Regression estimators for generic health-related quality of life and quality-adjusted life years Med Decis Making 2012 32 56 69 10.1177/0272989X11416988 22009667 29. Pullenayegum EM Tarride JE Xie F Goeree R Gerstein HC O’Reilly D Analysis of health utility data when some subjects attain the upper bound of 1: are Tobit and CLAD models appropriate? Value Health 2010 13 487 94 10.1111/j.1524-4733.2010.00695.x 20230549 30. Royston P Sauerbrei W Multivariable Model-Building: A pragmatic approach to regression analysis based on fractional polynomials for modelling continuous variables 2008 UK Wiley 31. Xie F Li SC Luo N Lo NN Yeo SJ Yang KY Comparison of the EuroQol and short form 6D in Singapore multiethnic Asian knee osteoarthritis patients scheduled for total knee replacement Arthritis Rheum 2007 57 1043 9 10.1002/art.22883 17665466 32. Turner AP Barlow JH Heathcote-Elliott C Long term health impact of playing professional football in the United Kingdom Br J Sports Med 2000 34 332 6 10.1136/bjsm.34.5.332 11049141 33. Krupic F Garellick G Gordon M Karrholm J Different patient-reported outcomes in immigrants and patients born in Sweden: 18,791 patients with 1 year follow-up in the Swedish Hip Arthroplasty Registry Acta Orthop 2014 85 221 8 10.3109/17453674.2014.919556 24803309 34. Brazier J Roberts J Tsuchiya A Busschbach J A comparison of the EQ-5D and SF-6D across seven patient groups Health Econ 2004 13 873 84 10.1002/hec.866 15362179 35. Gordon M Greene M Frumento P Rolfson O Garellick G Stark A Age- and health-related quality of life after total hip replacement: decreasing gains in patients above 70 years of age Acta Orthop 2014 85 244 9 10.3109/17453674.2014.916492 24786908 36. Ageberg E Forssblad M Herbertsson P Roos EM Sex differences in patient-reported outcomes after anterior cruciate ligament reconstruction: data from the Swedish knee ligament register Am J Sports Med 2010 38 1334 42 10.1177/0363546510361218 20410376

PMC005xxxxxx/PMC5002212.txt

==== Front BMC Med EducBMC Med EducBMC Medical Education1472-6920BioMed Central London 74910.1186/s12909-016-0749-3Research ArticleResearch interest and activity among medical students in Gothenburg, Sweden, a cross-sectional study http://orcid.org/0000-0002-4438-4849Stockfelt Marit marit.stockfelt@gu.se 1Karlsson Lars lars.karlsson@neuro.gu.se 2Finizia Caterina caterina.finizia@orlss.gu.se 31 Institute of Medicine, Department of Rheumatology and Inflammation Research, Sahlgrenska Academy, University of Gothenburg, Guldhedsgatan 10A, 405 30 Gothenburg, Sweden 2 Institute of Neuroscience and Physiology, Department of Clinical Neuroscience and Rehabilitation, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden 3 Department of Otorhinolaryngology, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden 26 8 2016 26 8 2016 2016 16 1 22630 6 2016 19 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background The proportion of physicians undertaking doctoral studies is decreasing. Early recruitment of medical students could counteract this trend. This follow-up survey investigated research interest and activity among medical students at the Sahlgrenska Academy, Gothenburg, Sweden. Methods A questionnaire was administered to all medical students at the Sahlgrenska Academy, as a follow-up to a 2006 survey. The Mann-Whitney U test was used for ordinal variables and the Fisher exact test for categorical variables. Data from Statistics Sweden was used to analyse the number of PhDs awarded to individuals who earned a medical degree in 2000–2012. Results Of the students, 16 % were already conducting and another 36 % wanted to conduct research during their studies. The interest was at the same level compared to 2006. The main reasons for conducting research consisted of an interest in scientific problems or the research topic, a wish for personal development or intellectual stimulation. Students engaged in research reported lack of time, increased workload and less time to study as hindering factors. Conclusions Recruitment could be improved by offering improved and regular information, clarifying career paths, broadly announcing available projects, and creating new and expanding existing research programmes. The potential for recruitment of Gothenburg medical students to research is substantial, but students are hampered by lack of time, lack of supervisors and lack of information. Electronic supplementary material The online version of this article (doi:10.1186/s12909-016-0749-3) contains supplementary material, which is available to authorized users. Keywords Medical studentResearchRecruitmentPhysicianScientistissue-copyright-statement© The Author(s) 2016 ==== Body Background The proportion of physicians undertaking doctoral studies has decreased progressively over the last years, both in Sweden [1] and internationally [2, 3]. Studies from the US report that the number of US medical students interested in research has decreased, as has the percentage of US medical doctors among those receiving National Institute of Health (NIH) government grants [4, 5]. The number of US physicians with research as their main professional activity is also decreasing [6]. In Sweden, the proportion of physicians among PhDs, and PhD students at Swedish medical faculties, has declined progressively since the beginning of the last decade. Moreover, the median age for Swedish physicians to finish their doctoral studies is currently 41 years, as compared with 34 years for other doctoral degrees [1]. Physicians trained in the scientific disciplines and the field of clinical medicine are essential for bringing patient-oriented research questions into focus, and bridging the gap between basic and clinical sciences [7, 8]. A shortage of physicians in research could in the future have negative consequences for academia, clinical research and health care [1], undermining the translation of basic research into patient care [8, 9]. To counteract this trend, early recruitment of medical students has been proposed, and medical students engaged in research during medical studies conduct more postgraduate research compared with their peers in the US [10–12] and the Netherlands [13]. Research conducted by medical students has been shown to be productive in terms of publications, as reported from Norway [14] and Germany [15]. The Swedish medical education spans 5.5 years, of which half a year at the end of the medical programme is dedicated to writing a research Master’s thesis mandatory for all students as part of the curriculum. An additional 1.5 years of internship is required to obtain a licence to practise medicine. With the PhD education in Sweden currently lasting 4 years, this adds up to a total of 11 years to become a licensed physician holding a PhD. One of six medical schools in Sweden is located in Gothenburg at the Sahlgrenska Academy. For extracurricular research, a part of the students are involved in the ‘Research assistant programme’ which was initiated in 2009 to stimulate research among medical students. Currently the programme offers ten positions per year [16]. Students are accepted for a 3-year period with scheduled research activities, such as lectures and presentations, and financial compensation for part-time research in parallel with their medical studies. Similar research stimulating projects can be found at other Swedish universities, in the forms of summer research programmes and research preparatory courses. In light of this and other efforts made to stimulate research engagement among medical students in Gothenburg, we have conducted a follow-up survey to investigate the progression of research interest and extracurricular research activity among medical students. Methods Study design and participants An anonymous questionnaire composed of both open and closed questions was administered electronically to all medical students at the Sahlgrenska Academy during the fall of 2012 and the spring of 2013 (Additional file 1). The questionnaire was partly based on a previous survey conducted in Gothenburg in 2006 that examined research interest and extracurricular research activity among medical students [17]. Research was defined as participating in a medical research project on scholarship, as a PhD student or as a researcher during and/or between semesters, but excluding the Master’s thesis at the end of the medical programme. Answers were required for all questions except for the open response questions. Parts of the data have previously been published in Swedish [18]. The research was performed in compliance with the Helsinki Declaration and approval from the Faculty was given. Questions were graded on a 10 point scale, with one representing to a very low degree, and ten representing to a very high degree. To questions with alternatives, the respondents were asked to choose one or multiple alternatives they agreed with, as well as to add their own alternatives. For open response questions, answers were systematically analyzed and assigned into mutually exclusive categories based on the respondents’ answers. From the most common categories, quotes were chosen as examples. Statistics A 10-point numerical scale was used for graded answers. When comparing categorical variables, the Mann-Whitney U test was used for ordinal variables and the Fisher exact test for categorical variables. Data from Statistics Sweden (SCB), that were based on Swedish Register of Education data, were used to analyse the total number of awarded PhD degrees in medicine (orientation code 731 according to the Swedish Educational Terminology, SUN-2000) for persons holding a medical degree in Sweden for each calendar year from 2000 to 2012. Results The participants The questionnaire was completed by 471 students, resulting in a response rate of 42 %. Of the respondents, 264 were female and 207 male. The distribution was comparable across the semesters. The proportion of women and men answering the questionnaire (56 and 44 % respectively) was consistent with the gender distribution in the medical programme in the autumn of 2012 (53 and 47 % respectively). The median age was 25 years, with 14 % of respondents being older than 30 years. Most of the students were living alone without children (54 %) or were married/cohabiting without children (35 %) and 10 % of respondents had children (Table 1).Table 1 Gender, age and family situation of students conducting/not conducting active research projects Gender Median age (yrs) Active research Family situation Yes N (%) No N (%) Single without children N (%) Married/Cohabiting without children N (%) Single with children N (%) Married/Cohabiting with children N (%) Female 24 38 (14 %) 226 (86 %) 139 (53 %) 106 (40 %) 1 (0 %) 18 (7 %) Male 26 37 (18 %) 170 (82 %) 117 (57 %) 61 (29 %) 3 (1 %) 26 (13 %) Total 25 75 (16 %) 396 (84 %) 256 (54 %) 167 (35 %) 4 (1 %) 44 (9 %) Are medical students interested in research? The proportion of students with active research projects was 16 %. Another 36 % reported interest in conducting active research during their medical studies. The interest in scientific issues was substantial and did not vary between semesters. Respondents scored an average of 7.4 on a scale from 1 (very little interest) to 10 (very great interest). Male students indicated a slightly greater interest, scoring 7.7 compared with 7.1 for female students (p < 0.01). The major reasons for conducting active research were an interest in scientific problems, an interest in the research topic, and a desire for personal development or for intellectual stimulation (Fig. 1). Students who expressed interest but were not actively conducting research indicated similar reasons. Among them, a higher proportion wanted to contribute to better health care and acquire critical thinking skills, while fewer expressed a wish for an extra income.Fig. 1 Reasons for medical students to conduct research Statistics produced by SCB show that the number of physicians finishing their PhD per year was largely unchanged between the years 2000–2012. By contrast, statistics from the Swedish National Board of Health and Welfare show that the number of Swedish medical licences issued per year almost doubled during the same period, from 1185 to 2186 (Fig. 2).Fig. 2 Ratio between PhD degrees awarded to persons holding a medical degree and number of medical licences issued per year in Sweden for the period 2000–2012 What are the disincentives for doing research? Lack of time (23 %), increased workload (22 %) and less time to study (16 %) were disincentives reported by students engaged in research. Inadequate financial compensation was considered by a minority (9 %) to be a disadvantage. Other reported disincentives were delayed income rise, slower career development, and research not being as meriting as clinical work. Students who were interested in research but did not have an active research project mainly identified lack of time, followed by not knowing how to start and not having found a research group or supervisor, as hindrance to starting a research project. Many students also indicated lack of information, difficulty in combining research with medical studies (9 %) and inadequate financial compensation, and a small percentage also indicated lack of finance, ‘more interested in clinical work’ , and ‘other’ (Fig. 3).Fig. 3 Hindering factors for medical students who wished to conduct research The students who were not interested in conducting research often identified time constraints (40 %) as the reason for not conducting research, followed by lack of interest (15 %), economic aspects (9 %) and family situation (9 %). These students felt that improved information (37 %), better financial compensation (27 %) and better work conditions in the future (23 %) would encourage more medical students to engage in research. How do students get involved in research? Most students actively engaged in research established contact by approaching the research team on their own. Some students continued after having conducted research on summer breaks, or after attended the Research assistant programme, while a smaller portion continued after completing the half year Master’s thesis for the medical degree, or were approached by the group or a teacher, or stated “other” way of being recruited than those suggested (Table 2). The lecturers’ commitment and personal contacts were important factors that got the students to do research, as were scholarships or having fellow students involved in research. Half of the students (49 %) reported that a lecturer had attempted to recruit them to research during lecture, while 11 % reported that this had occurred in private conversation. Only 14 % felt that they had received enough information and knowledge about conducting research during medical school. The Research assistant programme was known and had an impact on research interest for three quarters of the students. The positive influence was regardless of whether the students were conducting active research or not, with responses averaging at 5.7 and 4.3 respectively on a scale from 1 (very low influence) to 10 (very high). The proportion of students conducting active research increases during later semesters, with 8 % in the second semester and 19 % in the eleventh semester. Of the students who were not conducting active research, most reported willingness to engage in research in the future, and of those who were conducting active research, a majority of 80 % considered undertaking PhD studies in the future.Table 2 Pathways for involvement in research during medical studies Count Percentage Approached the research group 45 40.2 % Other 25 22.3 % Continued after research activity on summer breaks 20 17.9 % Participated in the Research assistant programme 14 12.5 % Continued after completing thesis project for medical degree 5 4.5 % Approached by research group or recruited by teacher 3 2.7 % Total 69 100.0 % What kind of research do the students wish to pursue? Basic science research projects, defined as being performed mainly in a laboratory environment, were most common (45 %), while a third of the students worked on clinical projects (32 %), defined as research projects that mainly involved human participants. The remaining respondents were involved in projects with both basic science and clinical aspects (23 %). Among the students who wished to do research actively, only one-tenth were interested in basic science research (9 %), while one-third preferred clinical research (30 %) and half favoured a project that included both basic science and clinical aspects (51 %). How do students combine research and medical studies? In general, students engaged in research found it intermediately difficult to combine medical studies with active research, with an average response of 4.5 on a scale of 1 (very easy to combine) to 10 (very difficult). Common forms of compensation for extracurricular research activity were salary (47 %), scholarship (33 %) or no compensation at all (15 %). Students willing to engage in research were interested in conducting research to a high extent between semesters, with an average score of 7.3 on a scale of 1 (very little extent) to 10 (very high). How should we improve recruitment to research? Students were asked in an open response question how recruitment could be improved. Students conducting research suggested that recruitment could be improved by offering financial compensation, informing and inspiring students, and clarifying career paths. Students interested in engaging in research requested information about research during lectures, and public announcements of research projects and groups. Other suggestions included expanded programmes and opportunities such as the Research assistant programme, and scholarships for summer research. Financial compensation, information and transparency in the recruitment process were also desired. Discussion Strong interest in research among medical students The present study shows that a large percentage of medical students in Gothenburg are interested in conducting research. The interest in scientific issues in our study population was at the same level as reported in the previous questionnaire study in 2006 [17] with a remaining gender difference, where male students indicated a slightly higher interest compared with female students, as also seen elsewhere [19]. Our survey showed that the proportion of students conducting active research increases during later semesters, and of the students who were not conducting extracurricular research, most reported willingness to engage in research in the future. Lack of time, information, and research group are hindering factors The main obstacles to engaging in research identified in this study are lack of time, lack of information, and lack of supervisors. However, lack of compensation, which has previously been proposed as a major obstacle [20], did not come up as one of the more important factors (Fig. 3). Possibly this becomes important once a research project is established, but by that time the majority of the students with active research projects in our study already had some form of financing. This study confirms the importance of continuously offering information about what research involves and how to begin a research career. Lack of time came up as an important barrier, and students not interested in conducting research often identified time constraints as the reason. Students at the Sahlgrenska Academy do not get any protected time for research. However, recently half a year has been dedicated for pursuing a Master’s thesis at the end of the medical education as a part of the curriculum. It might be interesting to follow whether this has an impact on extracurricular research activity among medical students. Difficulty in finding a research group In our survey, a strong interest in research was reported, as 36 % of those respondents not already conducting research reported willingness to engage in research during medical studies. However, we identified a gap between interested students and accessible supervisors. As one student wrote in free text, ‘Give us the chance to start doing research. The whole class would probably come running.’ The potential for improvement here lies primarily in improved contact between research groups and interested students. A recurring suggestion from the students was to have a central website where available projects and research groups could be publicly announced for students to apply to, a feature that has been requested by students in a Hawaiian study as well, where a vast majority of students were positive to an online searchable database of research projects [21]. A first step in the mediation of research projects has been introduced at the Sahlgrenska Academy in the form of a project database for Master’s thesis. This tool could easily be adapted for the use of students looking for a long-term engagement in research. The Research assistant programme enhances research interest throughout the medical programme This survey included 24 out of the 30 medical students who at the time were active in the Research assistant programme. This means that two-thirds of the students conducting research were doing research outside of the Research assistant programme. In our study, many of the students requested an expansion of the programme. It is also interesting to note that the Research assistant programme seems to have a positive influence on research interest both among students doing active research and among other students. A similar medical student research programme in Norway has been reported to have increased the recruitment of physicians to research [14]. The University of Queensland in Australia has implemented the Clinical Scientist Track, a research intensive programme for medical students to pursue a research Master’s degree alongside their medical degree, which has led to a majority continuing research as PhD students [22]. In Canada, apart from the recently cancelled joint MD/PhD program [23], there is since 1995 a Clinician Investigator Program through which clinicians engage in research training concurrently with their postgraduate medical education. This program has been recognized as an important mechanism for producing highly qualified clinicial investigators [24]. Furthermore, participation in a research training programme for medical students at the Howard Hughes Medical Institute, in Maryland in the US, has been shown to increase the likelihood of receiving NIH post-doctoral support [11]. Research experience during studies at three different medical schools in the US has also been strongly associated with post-graduate research involvement [12]. Basic science and clinical research An acute shortage of medical students at the basic science stage has previously been described [1, 25]. We found a decreased interest in basic science projects during the early semesters, which is consistent with a previous study from 2006 [17]. Almost half the students conducting active research were conducting research for a basic science project, whereas only one-tenth of the students wishing to do research were interested in basic science research. Most students were especially interested in a combination of basic science and clinical research. Clinical research seems to be difficult to initiate during the early semesters, with the proportion of students doing clinical research projects increasing with time (from 0 % in the second semester to 11 % in the eleventh semester). The reverse is true for the proportion of students in this study working on basic science projects (8 % in the second semester and 2 % in the eleventh semester). Several students commented that an enhanced opportunity for clinical research during the early semesters would increase their interest to start doing research. The Swedish Research Council reported in 2003 that the research interest among medical students increases during later semesters. However, in our study, the interest in engaging in research did not seem to be dependent on the semester. Importance of early recruitment A previous study of the Research assistant programme shows that medical students who begin research in pre-clinical semesters express stronger long-term research interest compared with students who start later [25], and the same pattern has been seen in early recruitment of medical students to surgery [26]. In light of substantial retirement of senior researcher physicians and a high median age for medical doctors gaining their PhD degree, efforts should be made for early recruitment of students to research. Lack of information about research and lack of availability of projects and research groups risks delaying recruitment to clinical research. This study reveals an opportunity for targeted and expanded efforts during the early semesters of medical school to improve information on current basic science and clinical research. High demand for recruitment Increased recruitment of young physicians involved in research is required. Of the more than 6000 physicians in Sweden with a PhD, only 9 % are under 40 years [1]. We are facing retirements of senior research physicians, and the annual number of new PhD physicians is not sufficient for re-growth of qualified supervisors and clinical researchers. In a German study, medical student research activity has been shown to significantly increase output of publications at the faculty level [15]. Suggestions from our respondents as to how recruitment can be increased were to stimulate interest in research, guide interested students into active research, and support students who are already engaged in research. Students who were not interested in research indicated that lack of time is a hindering factor, and one way to address this would be to increase acceptance in the medical programme of participation in research group meetings and laboratory work during the studies. Students doing active research likewise considered lack of time as a disadvantage, with the extracurricular research activity resulting in increased workload and decreased study time. Furthermore, the financial compensation for research participation may be seen as inadequate, and scholarships for summer research and the Research assistant programme may be important incentives. It should be noted that although financial compensation did not feature as a crucial hindering factor, more structured programmes such as the Research assistant programme were requested and such programmes could constitute important incentives. Substantial recruitment potential This questionnaire was sent to all students enrolled at the Sahlgrenska Academy. The response rate was low, with 42 % of students responding. Self-selection bias is possible, with students more interested in research being more likely to answer a questionnaire about the subject, meaning that the actual interest and engagement in this population may be overestimated. However, from the answers we can conclude that there is a substantial recruitment potential among Gothenburg medical students, with a large number of medical students across all semesters interested in conducting active research during their medical studies. The students are hindered mainly by a lack of information about what research entails and how to get in contact with research groups, and not so much by lack of financial compensation. Through targeted initiatives to stimulate research engagement, such as improved information, we may have an opportunity early on to recruit a large number of medical students to research and thus secure the need for physicians in research in the future. Conclusion The proportion of physicians undertaking doctoral studies has decreased progressively over the last years, both in Sweden and internationally. To counteract this trend, it is important to investigate and improve recruitment of medical students to research. This study was designed to investigate research interest and engagement of medical students in Sweden in 2012, as a follow-up to a similar study done in 2006. For increased and early recruitment of medical students, efforts could be directed towards offering improved and regular information about conducting research, publicly announcing available research projects, and creating and expanding research programmes for motivated medical students. Along with improved conditions and financial resources, these measures could help to accommodate the growing need for recruitment of medical students and physicians to research. Additional files Additional file 1: English version of the questionnaire used in the study. (DOCX 109 kb) Additional file 2: Raw data from the questionnaire. (XLSX 214 kb) Additional file 3: Data from Statistics Sweden (SCB). (PDF 6 kb) Acknowledgements The authors would like to thank Atefeh Hariri for help with designing the electronic questionnaire, and the AT-kansli at Sahlgrenska University Hospital, Gothenburg, for financing the project. Funding The AT-kansli at Sahlgrenska University Hospital, Gothenburg. Availability of data and materials The data analysed during this study are included in this published article [and its Additional files 2 and 3]. Authors’ contributions MS, LK and CF were involved in the conception and design of this study. MS and LK performed the acquisition, analysis and interpretation of data. MS and LK drafted the manuscript, and CF revised it critically. All authors have read and approved the final manuscript. Authors’ information MARIT STOCKFELT, MD, is a resident physician in Rheumatology and a PhD student in Medicine at the Sahlgrenska Academy, University of Gothenburg. LARS KARLSSON, MD, MSc(Pharm), BSc(Econ), is a licensed physician, and a PhD student in Neuroscience and Physiology at the Sahlgrenska Academy, University of Gothenburg. He is also the president of the PhD Student Council at the Academy. CATERINA FINIZIA, Professor, MD, PhD, with credits in pedagogy, is working as an ENT specialist at Sahlgrenska University Hospital (SU), a position she has held since 1994. Present postgraduate appointment: ENT specialist SU 20 %, director of research and development SU 60, and 20 % of working hours assigned to research work as a professor. Competing interest The authors declare that they have no competing interest. Consent for publication Not applicable. Ethics approval and consent to participate The research was performed in compliance with the Helsinki Declaration. Written consent to participate was given by all participants. Approval was given to this study by the Institutional review board, Faculty of Medicine at the The Sahlgrenska Academy, University of Gothenburg represented by Professor, Vice Dean of the faculty Eric Hanse, Institute of Neuroscience and Physiology. The Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden. No registers or patient journals were used and no patients will be followed up. The questionnaire was anonymous and voluntary. No individual patient data are reported. ==== Refs References 1. Stendahl O Världsklass! Åtgärdsplan för den kliniska forskningen SOU 2008 Stockholm 7 1 271 2. Cech TR Egan LW Doyle C Gallin E Lichtman MA Queenan CJ Sung N The biomedical research bottleneck Science (New York, NY) 2001 293 5530 573 10.1126/science.293.5530.573 3. Wyngaarden JB The clinical investigator as an endangered species Bull N Y Acad Med 1981 57 6 415 426 6941836 4. Rosenberg LE Physician-Scientists--Endangered and Essential Science (New York, NY) 1999 283 5400 331 332 10.1126/science.283.5400.331 5. Gottesman MM The role of the NIH in nurturing clinician-scientists N Engl J Med 2013 368 24 2249 2251 10.1056/NEJMp1302969 23675642 6. Garrison HH Deschamps AM NIH research funding and early career physician scientists: continuing challenges in the 21st century FASEB J 2014 28 3 1049 1058 10.1096/fj.13-241687 24297696 7. Zemlo TR Garrison HH Partridge NC Ley TJ The physician-scientist: career issues and challenges at the year 2000 FASEB J 2000 14 2 221 230 10657979 8. Sung NS Crowley WF Jr Genel M Salber P Sandy L Sherwood LM Johnson SB Catanese V Tilson H Getz K Central challenges facing the national clinical research enterprise Jama 2003 289 10 1278 1287 10.1001/jama.289.10.1278 12633190 9. Varki A Rosenberg LE Emerging opportunities and career paths for the young physician-scientist Nat Med 2002 8 5 437 439 10.1038/nm0502-437 11984577 10. Solomon SS Tom SC Pichert J Wasserman D Powers AC Impact of medical student research in the development of physician-scientists J Investig Med 2003 51 3 149 156 11. Fang D Meyer RE Effect of two Howard Hughes Medical Institute research training programs for medical students on the likelihood of pursuing research careers Acad Med 2003 78 12 1271 1280 10.1097/00001888-200312000-00017 14660432 12. Segal S Lloyd T Houts PS Stillman PL Jungas RL Greer RB 3rd The association between students’ research involvement in medical school and their postgraduate medical activities Acad Med 1990 65 8 530 533 10.1097/00001888-199008000-00010 2383337 13. Reinders JJ Kropmans TJ Cohen-Schotanus J Extracurricular research experience of medical students and their scientific output after graduation Med Educ 2005 39 2 237 10.1111/j.1365-2929.2004.02078.x 15679693 14. Hunskaar S Breivik J Siebke M Tommeras K Figenschau K Hansen J-B Evaluation of the medical student research programme in Norwegian medical schools. A survey of students and supervisors BMC Med Educ 2009 9 1 43 10.1186/1472-6920-9-43 19602226 15. Cursiefen C Altunbas A Contribution of medical student research to the MedlineTM-indexed publications of a German medical faculty Med Educ 1998 32 4 439 440 10.1046/j.1365-2923.1998.00255.x 9743810 16. Hanse E Wahlqvist M Sturm A Lindholm C Haraldsson B Andersson R Manhem K The medical education in Gothenburg is being reformed. Professional development, research and internationalization Lakartidningen 2011 108 12 669 673 21534321 17. Giglio D Bergh N Lack of time and money prevent medical students from conducting research Lakartidningen 2008 105 14 1011 1014 18478752 18. Karlsson L, Stockfelt M, Finizia C. Information på läkarprogrammet viktig för rekrytering till forskning. Lakartidningen. 2014, 111. 19. Guelich JM Singer BH Castro MC Rosenberg LE A Gender Gap in the Next Generation of Physician‐Scientists: Medical Student Interest and Participation in Research J Investig Med 2002 50 6 412 418 10.1136/jim-50-06-01 20. Giglio D Difficult for postgraduate medical students to find time and money for research. Still the degree of publishing and theses is rather high Lakartidningen 2009 106 14 999 1003 19485032 21. Grace T Arakawa R Hsiao YS Hirose-Wong S Hoshide R Medical school hotline: promoting medical student research using a searchable database of research projects at the John A.Burns School of Medicine Hawai’i J Med Public Health 2013 72 12 437 439 24377079 22. Eley DS Wilkinson D Building a teaching-research nexus in a research intensive university: rejuvenating the recruitment and training of the clinicial scientist Med Teach 2015 37 2 174 180 10.3109/0142159X.2014.947937 25155753 23. Twa DD, Squair JW, Skinnider MA, Ji JX. The Canadian clinician-scientist training program must be reinstated. J Clin Invest. 2015; 2015. 24. Hayward CP Danoff D Kennedy M Lee AC Brzezina S Bond U Clinician investigator training in Canada: a review Clin Invest Med 2011 34 4 E192 21810376 25. Boren J Adiels M Wickstrom Y Olofsson SO Bostrom P Recruitment of medical students to research requires early engagement. A research assistant program shows that early initiation yields better results Lakartidningen 2009 106 46 3072 3074 19998609 26. Haggerty KA Beaty CA George TJ Arnaoutakis GJ Baumgartner WA Increased exposure improves recruitment: early results of a program designed to attract medical students into surgical careers Ann Thorac Surg 2014 97 6 2111 2114 10.1016/j.athoracsur.2014.02.029 24725839

PMC005xxxxxx/PMC5002213.txt

==== Front BMC Public HealthBMC Public HealthBMC Public Health1471-2458BioMed Central London 354810.1186/s12889-016-3548-1Research ArticleA systematic review of published interventions for primary and secondary prevention of ischaemic heart disease (IHD) in rural populations of Australia Alston Laura V. 03 5247 9426laura.alston@deakin.edu.au Peterson Karen L. karen.peterson@deakin.edu.au Jacobs Jane P. jane.jacobs@deakin.edu.au Allender Steven steven.allender@deakin.edu.au Nichols Melanie melanie.nichols@deakin.edu.au WHO Collaborating Centre for Obesity Prevention, Faculty of Health, Deakin University, Locked Bag 20001, Geelong, VIC 3220 Australia 27 8 2016 27 8 2016 2016 16 1 89520 5 2016 18 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Rural Australians are known to experience a higher burden of ischaemic heart disease (IHD) than their metropolitan counterparts and the reasons for this appear to be highly complex and not well understood. It is not clear what interventions and prevention efforts have occurred specifically in rural Australia in terms of IHD. A summary of this evidence could have implications for future action and research in improving the health of rural communities. The aim of this study was to review all published interventions conducted in rural Australia that were aimed at the primary and/or secondary prevention of ischaemic heart disease (IHD) in adults. Methods Systematic review of the peer-reviewed literature published between January 1990 and December 2015. Search terms were derived from four major topics: (1) rural; (2) ischaemic heart disease; (3) Australia and; (4) intervention/prevention. Terms were adapted for six databases and three independent researchers screened results. Studies were included if the published work described an intervention focussed on the prevention or reduction of IHD or risk factors, specifically in a rural population of Australia, with outcomes specific to participants including, but not limited to, changes in diet, exercise, cholesterol or blood pressure levels. Results Of 791 papers identified in the search, seven studies met the inclusion criteria, and one further study was retrieved from searching reference lists of screened abstracts. Typically, excluded studies focused on cardiovascular diseases without specific reference to IHD, or presented intervention results without stratification by rurality. Larger trials that included metropolitan residents without stratification were excluded due to differences in the specific needs, characteristics and health service access challenges of rural populations. Six interventions were primary prevention studies, one was secondary prevention only and one included both primary and secondary intervention strategies. Two interventions were focussed exclusively on Aboriginal and Torres Strait Islander (Australian Indigenous) populations. Conclusions Few interventions were identified that exclusively focussed on IHD prevention in rural communities, despite these populations being at increased risk of IHD in Australia, and this is consistent with comparable countries, internationally. Although limited, available evidence shows that primary and secondary interventions targeted at IHD and related risk factors can be effective in a rural setting. Electronic supplementary material The online version of this article (doi:10.1186/s12889-016-3548-1) contains supplementary material, which is available to authorized users. Keywords RuralAustraliaIschaemic heart diseaseInequalitiesInterventionDeakin University (AU)issue-copyright-statement© The Author(s) 2016 ==== Body Background Globally, more people die from cardiovascular diseases (CVD), than any other cause [1]. In 2013, 29.5 % of Australian deaths were attributed to CVD, making it the most common cause of death [2]. ischaemic heart disease (IHD) is the most prevalent CVD and is defined clinically as acute myocardial infarction (AMI) or angina pectoris [2], and it is estimated that, globally, 7.4 million people die from IHD each year [1]. IHD has been the leading single cause of death in Australia since 2000 [2]. These conditions appear to affect some populations more than others [3], particularly those living in rural areas, people of Aboriginal and Torres Strait Islander (ATSI) heritage and people of lower socio-economic status (SES) [4, 5]. Modelled estimates of Australian mortality figures between 2009 and 2011 suggest that more than 1200 lives would have been saved annually, if people living in rural areas had the same IHD mortality rate as metropolitan counterparts [6]. Overall IHD mortality rates in Australia decreased substantially between 2001 and 2010, though these decreases were smaller in more remote areas than in major cities (−4.1 % for males and −4.3 % for females in major cities, compared to −2.4 % and −3.9 % in remote areas) [5]. The increased burden of IHD in rural areas of Australia, despite overall mortality decline, is comparable to patterns observed in high income countries internationally including in rural Scotland, Norway and the United States (US) [7–9]. Decreases in IHD have been observed in high-income countries, such as in the US and UK, and have been largely attributed to primary and secondary prevention efforts that have led to the reduction in modifiable risk factors such as hypertension, cholesterol and smoking, as well as advances in medical therapies [9–11]. Risk factors for IHD are interlinked, with modifiable factors including tobacco smoking, poor nutrition, physical inactivity, obesity, high blood pressure and high blood cholesterol [5]. These risk factors are also common to other major non-communicable diseases (NCDs), including stroke, cancer, respiratory disease and diabetes [12]. The importance of these risk factors is emphasised by the World Health Organisation’s 25x 25 goal, which identifies them as significant targets to achieve the goal of reducing premature mortality from NCDs by 25 % by the year 2025 [12, 13]. CVD, including IHD and stroke, has been identified as a high priority in rural Australia with particular reference to primary prevention strategies focused on improving nutrition and physical activity and reducing tobacco smoking [14]. The disparity between rural and metropolitan mortality and disease rates represents an important equity target for any prevention strategy and there is some limited evidence internationally for the effectiveness of community level prevention efforts in rural communities, when they are tailored specifically to the needs of the target population [11, 15]. Interventions attempting to address the increased burden of IHD in rural areas need to take into account the ways in which rural and metropolitan populations differ, which include health care access, education, income and risk factor prevalence. The aim of this study was to systematically review all published literature since 1990 reporting interventions that focussed on reducing the IHD burden in rural Australia, through primary or secondary prevention, and to synthesise the available evidence on the efficacy of such prevention efforts. Methods Data sources We sought to identify studies within the published peer-review literature that were focussed on rural populations and that aimed to prevent or reduce IHD burden or risk factors. This systematic review was registered with Prospero, (number CRD42016033431). The term ‘rural’ used throughout this paper, refers to all areas classified as being outside of major cities of Australia, by the Australian Bureau of Statistics, Accessibility Remoteness Index of Australia (ARIA) [16]. ARIA has five categories of remoteness, which are, in increasing order of remoteness major cities, inner regional, outer regional, remote and very remote. These definitions are based on remoteness scores derived from relative road distance to population localities and services [16]. Search terms used were related to four major topics including, ‘ischaemic heart disease’, ‘rural’, ‘intervention or prevention,’ and ‘Australia’. The search was conducted in November and December of 2015. The six databases included in the search were CINAHL, Medline, Academic Search complete, Rural and Remote Health Database, Health and Society Database and Embase. An additional hand search was undertaken of reference lists from included studies. The following were the Inclusion criteria:Studies had to be published in peer review journals from 1990 to 2015. Population: The study had to be focussed on a population of adults living exclusively in a rural area of Australia. Larger trials that included both rural and metropolitan residents without stratification by rurality were excluded. Intervention: Interventions reporting an explicit aim of primary or secondary prevention of heart disease, with specific mention of IHD as a target. For example, if a study referred only to CVD as a whole, and not specifically to IHD, it was excluded. Comparator: Comparisons between intervention groups and control group (preferably), or relevant health survey data or baseline results. Comparison to a non-rural population was not necessary for inclusion. Outcomes: Including but not limited to: changes in behavioural risk factors (including exercise, diet, alcohol, smoking and stress management), knowledge of heart disease, health assessment measures (e.g. blood pressure, cholesterol, blood glucose levels, obesity or weight), and rates of mortality, morbidity, case fatality, hospital admissions, or complications. Study design: All types of intervention designs were considered in this review. Studies describing intervention models (study design/protocol papers) that did not present intervention results were excluded. Study selection, data extraction and analysis The lead researcher (LA) reviewed all results from the six databases, removed duplicates and screened all results based on titles and, abstracts against the review criteria (see Fig. 1). Two additional researchers (KP & JJ) each screened a 50 % sample of titles and abstracts as a second reviewer. Any discrepancies were identified and resolved by consensus among the three researchers producing a list of papers for full text assessment for eligibility against the review criteria. Reference lists of all full texts were then searched for additional potentially eligible studies.Fig. 1 PRISMA diagram of the systematic review process for this review Data were extracted into a spreadsheet from the full texts by the lead researcher. The details collected included the publication details of the study, years of intervention, intervention type, follow up period, outcome measures (such as changes in clinical and modifiable risk factors), results and authors’ conclusions. Each intervention was then then categorised as either primary or secondary prevention, or both, and by the broad type of intervention (e.g. delivery through initial screening/education/exercise or whole community programs). The studies were then synthesized into a narrative analysis, with a focus on changes in outcome measures. We applied a narrative analysis because quantitative meta-analysis was deemed inappropriate due to the small sample size and heterogeneity of the interventions returned by the search strategy. Quality analysis Two researchers independently assessed each study using the Cochrane Collaboration’s tool for assessing risk of bias [16]. The tool is used to assess the risk of bias within each individual study based on five different types of bias including: selection bias (randomisation of participants), performance bias (blinding of participants), detection bias (blinding of outcome measures), attrition bias (incomplete outcome data) and complete reporting. Studies were rated as either high, low or unclear risk against each of the criteria. The Cochrane tool does not use a total score to assess overall risk of bias, so each type of bias is assessed individually. Results Of the initial 791 papers returned by the database search, 33 full texts were screened, and of these, seven studies met the inclusion criteria. Major reasons for exclusion at full text stage included that the study did not specifically refer to IHD, (usually only reporting on CVD as a whole). One further eligible peer reviewed study was identified through hand searching of reference lists, resulting in a total of eight studies included in the review. Details of interventions, outcome measures, results and conclusions of the studies are described in Table 1. Across these eight studies, five were conducted in ‘inner regional areas’ [17–21], two in ‘outer regional areas’ [22, 23] and one in a ‘very remote’ area [24]. No studies included here were conducted in remote areas of Australia. Two included interventions focussed exclusively on Aboriginal and Torres Strait populations [22, 24], four included a screening component part of the intervention [18, 20–22] one evaluated the effectiveness of a long term, whole community intervention [19] and one study included an assessment of cardiac rehabilitation [17].Table 1 Characteristics of prevention programs aimed at reducing ischaemic heart disease burden in rural Australia Author, year of publication Year(s) of study Intervention strategies Participants, follow up Outcome measures Results Conclusions Aoun & Rosenberg, 2004 [17] 2000–2001 7 week cardiac rehabilitation program N = 203 patients with current CVD diagnosis, n = 159 controls. Followed up at post program, 3, 6 and 12 months Self-reported changes in: Cardiac Rehab programs in rural areas are successful in reducing risk factors for IHD and improving quality of life -WT -WT: ↓ 0.5 kg -PA (6 min walk test) (p = 0.004) -BP, -PA: 431.6 m to 469.6 m (p < 0.001) -Quality of life scores (QoL) -BP: NS, p value not reported -QoL: 80.69 (15.9) to control 71.6 (18.86) (p = 0.04) Burgess et al., 2015 [22] 2012–2014 Cardiac prevention screening services within primary health teams Aboriginal clients aged 20 years and over, N = 2586 identified as high risk. Followed up every 3 months for two years Achievement of target (not compared to baseline for significance): Achieved target post program: This type of program is a feasible way of reducing IHD risk factors in rural indigenous populations -BP -BP: 57 % -TC -TC : 40 % No control group -% Stopped smoking -Stopped smoking: 50 % Carrington and Stewart, 2015 [18] 2009–2010 Nurse-led screening and education program N = 530, pre/post follow up design, no control group. Followed up at 6 months Mean change in -BP diastolic: ↓ 4 mmHg Systolic: ↓ 1 mmHg Feasibility of a nurse-led screening and intervention was shown for a rural population -BP -TC -WT (kg) -TC: ↓ 0.6 mmol/L -BMI -WT: ↓ 1.0 kg -BMI: ↓ 0.3mkg2 Higginbotham et al., 1999 [19] 1980–1990s (exact years not specified) Whole community intervention N = 359, no control group, but rates compared to nearby region Change in Intervention area: Whole community interventions can have multiple positive impacts in rural communities and possibly reduce IHD burden if implemented with consideration of community needs and subgroups -IHD Mortality (age standardised rates (per 100,000)) Women (35-64y) Fatal MI: −14.2 (95 % CI: −26.0, −2.4) 9 year data collection phase -Non-fatal MI rates, Non-fatal MI: 1.7 (95 % CI: −4.4, 7.9) -Case fatality compared to non-intervention region Men (35-64y) Fatal MI: −10.9 (95 % CI: −18.2, −3.6) Non-fatal MI: 3.2 (95 % CI: −0.6, 7.0) Rates declined faster in intervention population compared to than non-intervention region Krass et al., 2003 [20] Year(s) of intervention not specified Pharmacy screening and education program N = 389 adults in regional area, followed up from baseline to 3 months, no control group From baseline to 3 months: % Inactive Community Pharmacies have the potential to increase resource provision in rural areas and can be effective at reducing risk factors for IHD Cohort 1 Change in 57 % to 44 % (p < 0.0001) Cohort 2 -BP -TC -% Current smokers 50 % to 44 % (p = 0.01) -% Not meeting PA recommendations % Smokers = No change -% Of people by BMI category Both Cohorts: Mean TC: ↓ 0.26 mmol/L (95 % CI 10–0.42) (p < 0.003). BP: ↓ 10.5 mmHg (95 % CI 4.0-16.9) in mean systolic BP within Cohort 1 (p = 0.012), no difference for cohort 2. BMI = NS (p value NR) Kerr et al., 2008 [23] Year(s) of intervention not specified Exercise and cardiovascular monitoring program N = 164 war veterans, followed up at 3, 6, 12 months 3 monthly follow up: 12 months: This type of program was shown to be effective at reducing risk factors in a high risk, regional population of males -Diastolic and systolic BP (mmHg) Resting HR:↓ 4.0 bmp - HR (bpm) Diastolic BP: ↓ 6.4 mmHg Systolic BP: ↓ 8.4 mmHg (p = <0.05). Weight (kg) :NS Ray, 2001 [21] Year(s) of intervention not specified Once-off mobile heart screening program N = 135 adults aged 30–69 years followed p 6 months post intervention Self-report change in health behaviour after screening Self-report health behaviours: Heart risk screening can be a motivator for health behaviour change 76 = positive change 59 = no change Rowley et al., 2000 [24] 1993–1995 Lifestyle education program Aboriginal community participants Change in risk factors overtime (Intervention group either compared BL or to control): -no significant change in dietary and physical activity when compared to controls. Some short term changes were not sustained in metabolic profiles from this intervention, however this program was found to be sustainable for this type of rural community N = 32 intervention, N = 17 controls followed up at, 6 months, 2 years -BMI -Fasting glucose -BMI: ↓from BL at 6 months (to control: p = 0.012), 12 months: NS (p = NR) -Fasting glucose: Positive changes in awareness and behavioural risk factors were noted 6 months:↓ 0.9 mmol (intervention to baseline p = 0.021) - Glucose tolerance (oral glucose tolerance test (OGTT)) Intervention to control : NS (p = 0.132) −2 h post -OGTT: -plasma insulin 6 months: ↓ 1.6 mmol/l (p = 0.01 to BL) -triglyceride concentration Intervention to control: NS p = 0.154 -Fasting insulin: Intervention to control NS (p = 0.103) -Fasting triglycerides: NS (p = 0.158) Abbreviations: BL baseline, BMI body mass index, BP blood pressure, HR heart rate, bpm beats per minute, IHD ischaemic heart disease, MI myocardial infarction, NS not significant, NR Not reported, OGTT oral glucose tolerance test, PA physical activity, TC total cholesterol, QoL quality of life, WT weight (kg), ↓: decrease Primary prevention Six primary prevention studies were identified; five in inner and outer regional areas [18, 20–23], and one in a remote area [24]. Two reported on interventions in ATSI populations [22, 24]. Most studies [17, 20, 21, 24] were published more than 8 years ago. The intervention activities included an exercise program for a high risk population [23], cardiac rehabilitation [17], a full community intervention [19], and five risk factor screening and/or subsequent education or treatment programs within small communities [18, 20–22, 24]. Generally, the studies reviewed showed that IHD prevention efforts in rural communities are feasible and were effective in either reducing one or more risk factors, or IHD mortality, however the studies were limited by short follow up periods, small population numbers and a lack of inclusion of control groups in study designs. Kerr et al. [23] measured the effect of a 12-month exercise program on IHD risk factors in a population of war veterans living in regional Queensland (n = 164), without a control group. The main outcome measures included measurement of heart rate (HR), blood pressure (BP), skinfold and girth measurements, exercise heart rate response and estimated aerobic capacity. The results were used to determine if the program could be effective in this rural, high risk population. The study showed that an organised exercise group could be feasible in a rural setting for high risk clients, with positive effects shown for resting HR (−4 bpm), diastolic (−6.4 mmHg) and systolic BP (−8.4 mmHg) by the end of the program (p = <0.05). Weight was unchanged at 12 months, however there were some improvements in body composition. The generalisability of results from this study is limited because only 54 % of participants completed the final 12 month follow up assessment. Three studies [18, 20, 21] used primary risk factor screening as the start point of the intervention, screening patients’ BP, cholesterol and BMI, and when compared to baseline, showed that these types of programs are potentially feasible in rural areas. Krass and colleagues [20], assessed the impact of a pharmacy screening and health promotion program (n = 389) on the risk of IHD and stroke in two towns in regional New South Wales. The health promotion program included individual education on lifestyle improvements including diet, exercise and smoking cessation advice. After three months, significant changes were observed in mean total cholesterol for both towns (−0.26 mmol/l, 95 % CI 0.10-0.42, p = <0.003), while BP was reduced in participants from one town (−10.5 mmHg, 95 % CI 4.0-16.9, p = 0.012), changes in physical activity and smoking prevalence were reported, with increases in activity reported by participants at 3 months. The authors noted that there was little change in smoking prevalence, and attributed this to the relatively short period of intervention and follow up of this study design, which with no control group comparison, would also make it difficult to draw concrete conclusions on the overall effectiveness of this intervention. Carrington and Stewart [18], describe a similar, yet nurse-led intervention in regional Victoria in which 530 self-selecting patients were screened for cardiovascular risk, then provided with counselling and advice tailored to their risk level [18]. Just over 60 % of the patients (n = 326) had clinically significant improvements in risk factor levels at 6 months post-intervention, with BP, total cholesterol and weight all decreasing from baseline levels, yet these results were not compared to a control group. No further follow up was undertaken after 6 months post-intervention, making it difficult to determine either the sustainability of this design or the long term impact on risk factors in the rural community studied. A third screening study published in 2001 [21], without inclusion of controls, assessed the effect of a one-time screening session delivered by the mobile ‘Heart Bus’ on self-reported behaviour changes in inner regional Queensland. The study followed up 135 participants and found that 76 % of participants self-reported they had made positive changes to their diet and exercise behaviours 6 months later. The major limitations of this study include the small sample design, lack of control and the self-reported nature of all of the outcome measures. Two studies assessed IHD interventions in ATSI populations only [22, 24] and showed that IHD primary prevention efforts in these populations have the potential to be effective. One study was a clinical audit of cardiac prevention and screening services [22], and the other was an assessment of a diet and exercise program with comparison made to a self-selected control group [24]. Burgess et al. [22] analysed the effectiveness of cardiac prevention services through clinical audits every three months of cardiovascular risk assessments, and level of pharmaceutical prescription delivered through primary health care services over 2 years. The study focussed on results from 2586 participants, who were identified to have a five year CVD risk of 16 % or greater. Blood pressure medication was prescribed for 67 % of participants and lipid lowering medications for 55 %, with clinical follow-up every three months to assess if target levels were achieved. By the end of the two year evaluation, the number of participants who achieved clinical targets for BP was 1366 (56 %) while 989 (40 %) reached targets for cholesterol, however changes in the proportion of participants reaching targets did not change significantly over time for either outcome. Rowley and colleagues [24] assessed the effectiveness of a primary health care service providing diet and/or exercise education and support in a small rural ATSI population. The study included an intervention group (n = 32) and a self-selected control group (n = 17). There were significant differences (p = 0.03) observed between the intervention and control arms, for mean change in 2 h plasma glucose, and triglyceride levels at two years post-intervention. There were also changes in dietary behaviour and physical activity in the intervention group, however these did not appear to be significant when compared to controls. Response rates ast the two year follow up were low for younger participants aged 15–34, with 43 % responding, compared to 80 % for those aged 35 and over, however results shown here are clearly limited by the small study sample. Secondary prevention Only one secondary prevention intervention [17] was identified. The study evaluated a 7-week bi-weekly education and exercise cardiac rehabilitation program in a rural area of Western Australia (Heart Smart), and included 203 participants with a current CVD diagnosis. The evaluation compared quality of life and cardiac knowledge scores of these participants with to 159 non-participants (who were eligible, but did not wish to participate in the intervention). All follow up data, including clinical measures were self-reported. After 6 months, the intervention group had reportedly increased their physical activity (p < =0.001), and reduced their weight (−0.5 kg, p < 0.05) from baseline. Higher quality of life and cardiac knowledge scores were observed for the intervention group, however the non-intervention group had a low response rate to the follow up survey (42 %). Self-reported cholesterol levels were 3.6 mmol/L at baseline, and these reduced to 2.8 mmol/L by 6 months post follow up, however no changes were observed for self-reported BP. The authors did not specify if these measures were taken by the same medical clinic, or if any clinical documentation was collected with self-report results. Primary and secondary prevention Higginbotham et al. [19] was the only study to include both a primary and secondary prevention program. The Coalfields Healthy Heartbeat was a 10-year community intervention, which employed multiple strategies and included health promotion and awareness advertising, mobilisation of community resources, school health-promotion programs, exercise, cooking and education groups, and a cardiac rehabilitation program. The remainder of the Hunter Valley region served as a comparison population for the program evaluation. The authors found a larger reduction in fatal AMI cases in the intervention area relative to the comparison population over a 9-year monitoring period [19], however there was no reduction in non-fatal AMI rates. There were no significant differences in risk factors changes between the intervention and non-intervention areas. Table 1: Characteristics of prevention programs aimed at reducing IHD burden in rural Australia. Quality assessment Due to the lack of randomization used in the design of all studies and minimal use of control groups, all of the included studies were assessed to have a high risk of bias for the selection, performance and detection criteria of the Cochrane tool, and low or unclear risk for attrition and reporting bias criteria. Only one study reported using a behavioural science theory in the design of their study, identifying the health promotion and behavioural change models in the design of their follow up questionnaire [16]. There was also a lack of detail around the length of the intervention, and follow up periods for 4 of the 8 studies, and generally follow up periods were small. Discussion Eight studies were identified that met the inclusion criteria, indicating that there is little published work about IHD prevention efforts occurring in rural communities in Australia. All studies based in inner and outer regional areas [16], found that primary and secondary prevention activities can be effective at reducing IHD and related risk factors, although many reported modest and/or mixed results. The paucity of studies in remote and very remote areas [16] particularly in light of the relatively high burden of IHD in these areas compared to metropolitan counterparts [25], shows this is an understudied population in Australia. This is consistent with findings from the US, where rural populations are also identified as disadvantaged, and understudied in terms of IHD burden and prevention [9]. Worldwide, there is a lack of published research on comprehensive interventions to reduce IHD, especially in rural populations [26], despite the recognition that prevention efforts at population level, aimed at modifiable risk factors, can be both cost-effective and sustainable approaches to reducing IHD burden in high risk communities [11, 26, 27]. A recent review by Papadakis and Moroz [11] of high quality international studies of IHD prevention efforts, found only one study focussed on rural areas [27], out of 15 included in the review. All studies included in this review employed a non-randomized design, mostly without a control group (six of eight studies), and all were found to have a high risk of bias. Half of the studies did not provide clear details about the exact timeframe of the intervention period, and subsequent follow up, or plans to follow up. Lack of comprehensive follow-up has been identified as an issue with the quality of evidence around IHD prevention efforts in rural areas, internationally [26]. Research in rural areas is likely to be challenging due to limited resources, small population numbers, and geographical remoteness. The lack of eligible studies retrieved for this review may reflect either a true lack of action or a lack of research reporting and publishing of prevention programs operating in rural communities. It is possible that effective and targeted interventions are taking place in many rural settings; however, they are not being published in the academic literature for others to learn from. It is known that rural health professionals feel ill-equipped to undertake, complete, and publish research, and this is due to a lack of resources, supervision, and perceived skills in rural areas [28, 29]. Research with larger samples and more rigorous study designs are required to progress strategies for reducing IHD burden in rural Australia, and worldwide. Relatively few of the studies included in this review, reported extensively on behaviour change outcomes. Those studies that did included behaviour measurements [17, 18, 20, 22–24], used self-reported data only and findings were mixed. Changes in smoking prevalence at follow up were reported by one study [20], but did not change. Changes in BMI or weight were reported in five studies [17, 18, 20, 23, 24], however all but one study [17] found significant differences from pre and post intervention. Heterogeneity among these studies also made it difficult to draw concrete conclusions on the effectiveness of these interventions in improving risk behaviours. Reductions in BP through medication prescription and monitoring, were demonstrated in four of the five studies [17, 18, 20, 22, 23] that assessed BP as an outcome, suggesting that perhaps, if similar interventions were implemented more broadly in rural areas, significant population-level impacts on blood pressure levels could be achieved. A reduction of 5 mmHg in diastolic BP level has been suggested to reduce population mortality from IHD by approximately 21 % [30, 31], and BP reduction is widely accepted as an important target risk factor, along with smoking and cholesterol levels, when aiming to prevent IHD in high risk populations internationally [7, 9, 30]. The results of this review showed a strong emphasis on clinical measurements rather than behavioural measures as outcomes when evaluating IHD prevention efforts. This pattern is evident in other international research, including a large scale intervention in rural Maine in the US, that was spread across 23 rural communities from 1970 to 2010 [27]. Although the intervention consisted of nurse delivered education focussed on behaviour changes, prescription medication and monitoring, cholesterol and BP were the main measured outcomes [18, 22]. This study provided some evidence that nurse-led education programs can reduce IHD risk factors in rural communities, but again, did not compare to a control group [27]. The focus on clinical risk factors as outcomes, as opposed to behavioural changes may be due to the possible bias that can arise from these self-reported measures, with clinical risk factor measures providing more concrete evidence of the effectiveness of the IHD prevention programs. Higginbotham et al. [19] identified challenges with implementing a whole community intervention, and noted resistance when trying to engage the whole community, finding that only 35 % of people in the region thought IHD was of high concern [19]. Conclusions from this study included that when designing interventions for rural communities, use of existing structures and knowledge of the needs and interests of local subgroups is fundamentally important [19]. The authors of studies in the US and UK have also emphasized the importance of early assessment of rural community needs, interests and subgroups prior to implementation, when focussing on reducing IHD risk factors [7, 9, 32]. Conclusions from a review of international population-level interventions to reduce IHD showed that there is no ‘one size fits all’ program, and success of interventions relies heavily on due consideration of the needs, interests, characteristics and location of the community [11]. This is certainly applicable to the Australian rural context, which comprises of many diverse and different communities across a very large landscape. Notable from this review was that only two of the selected studies were published in the past 5 years, with the remainder being eight or more years old, possibly limiting the relevance of these studies to when considering the current rural health context. The age of the evidence may be of concern given that Australia has seen both significant changes in the burden of IHD, and substantial shifts in urbanisation and population characteristics over recent decades [33]. Further, advances in technology, including e-health and telecommunications, may alter the experience and health implications of living in a non-metropolitan area. Therefore, rural populations presented in older studies, and the issues they faced, may not be comparable to those currently residing in rural areas. Strengths of this research include that it employed a systematic methodology with broad search terms and included all Australian-based studies from 1990 to 2015. The comprehensive search strategy and use of three researchers to screen results increased the chances of all appropriate results being identified and included. The limitations of this review are that there was also only a small sample of studies that met the inclusion criteria, which had high heterogeneity and risk of bias in the methods, and this made solid conclusions on the effectiveness of interventions in rural areas difficult to decipher. The review was also limited to Australian studies only. While there are important similarities between rural populations across comparable countries, rural populations may differ significantly in demographic characteristics and disadvantage between different countries [34]. Implications for future research A major finding from this review is the lack of high quality studies assessing the effectiveness of interventions to reduce IHD burden in rural areas of Australia. Priority must be placed on action-orientated prevention research in rural communities, with a focus on methodologies such as community-based participatory research to empower communities, create tailored strategies and address inequalities [35] Future research must have specific focus on rigorous methodology, including comparison to control groups, more comprehensive measurements of changes in modifiable risk factors, and longer follow-up time frames in order to assess sustainability of such programs in the rural context. . There is significant potential for countries with comparable rural health inequalities (for example, geographically large, high income countries such as Australia, Canada and the US) to learn from international examples [7, 9, 26, 27] and create a stronger body of evidence for the prevention of IHD among rural populations globally international examples and create a stronger body of evidence for the prevention of IHD among rural populations globally. Conclusions There are very few studies on interventions to prevent and reduce IHD in regional and remote areas of Australia, despite the higher burden in rural areas being well documented and this is consistent with international observations. Published interventions have generally shown encouraging results in reducing IHD risk factors or outcomes, although there significant limitations to quality and external validty among the studies identified. More research is needed to determine appropriate, feasible, effective, and cost-effective strategies for improving IHD rates in rural areas, and the best ways to tailor interventions to the specific needs of rural communities. Additional files Additional file 1: PRISMA checklist. (DOC 62 kb) Additional file 2: Planned search terms. (DOCX 22 kb) Abbreviations ATSIAboriginal and Torres Strait Islander CVDCardiovascular disease IHDIschaemic heart disease NCDsNon-communicable diseases 95 % CI95 % confidence interval Acknowledgements There are no further acknowledgements. Funding MN, KP, LA and JJ are supported by funding from the National Heart Foundation of Australia and Deakin University. SA is supported by funding from an Australian National Health and Medical Research Council/National Heart Foundation of Australia Career Development Fellowship (APP1045836). He is also a researcher on the US National Institutes of Health grant titled, “Systems Science to Guide Whole-of- Community Childhood Obesity Interventions” (1R01HL115485-01A1) and within a NHMRC Centre for Research Excellence in Obesity Policy and Food Systems (APP1041020). Availability of data and materials As all studies included in this review are published, all data is accessible. A copy of the search strategy, including all search terms, is also available upon request. Authors’ contributions LA led the conception of the research questions, study design, searches, screening, analysis of results and drafting of the manuscript at all stages, MN also contributed to the study design, research questions, and critical revision of the manuscript. KP and JJ participated in the screening of studies, assessment of quality and revision of the manuscript. SA contributed to the conception of research questions and revision of the manuscript. All authors read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. Consent for publication Not relevant for this study. Ethics approval and consent to participate Not relevant for this study. Additional information See supplementary information including a PRISMA checklist (Additional file 1) and Search plan (Additional file 2). ==== Refs References 1. World Health Organisation. Cardiovascular Diseases: The World Health Organisation; 2016 [cited 2016 05/07/2016]. Available from: http://www.who.int/mediacentre/factsheets/fs317/en/. 2. Australian Bureau of Statistics. 3303.0 - Causes of Death, Australia, 2013 Australia: ABS; 2015 [cited 2016 29/02/2016]. Available from: http://www.abs.gov.au/ausstats/abs@.nsf/Lookup/by%20Subject/3303.0~2013~Main%20Features~Summary%20of%20findings~10000. 3. Australia Institute of Health and Welfare. What are cardiovascular diseases? 2013 [cited 2014 04/06/2014]. Available from: http://www.aihw.gov.au/cardiovascular-disease/what-is-cvd/. 4. Australian Bureau of Statistics. HEALTH OUTSIDE MAJOR CITIES Australia: ABS; 2011 [cited 2015 10/06/2015]. Available from: http://www.abs.gov.au/AUSSTATS/abs@.nsf/Lookup/4102.0Main+Features30Mar+2011. 5. Australian Institute of Health and Welfare. Trends in Coronary heart disease mortality: age groups and populations: Cardiovascular series no. 38. Canberra: AIHW C, editor. 2014. 6. Australian Institute of Health and Welfare Mortality inequalities in Australia: bulletin 124 2014 Canberra AIHW 7. Levin KA Leyland AH Urban–rural inequalities in ischemic heart disease in Scotland, 1981–1999 Am J Public Health 2006 96 1 145 51 10.2105/AJPH.2004.051193 16317212 8. Kruger O Aase A Westin S Ischaemic heart disease mortality among men in Norway: reversal of urban–rural difference between 1966 and 1989 J Epidemiol Community Health 1995 49 3 271 6 10.1136/jech.49.3.271 7629462 9. Kulshreshtha A Goyal A Dabhadkar K Veledar E Vaccarino V Urban–rural differences in coronary heart disease mortality in the United States: 1999–2009 Public Health Rep 2014 129 1 19 29 24381356 10. O’Flaherty M Allender S Taylor R Stevenson C Peeters A Capewell S The decline in coronary heart disease mortality is slowing in young adults (Australia 1976–2006): a time trend analysis Int J Cardiol 2012 158 2 193 8 10.1016/j.ijcard.2011.01.016 21288580 11. Papadakis S Moroz I Population-level interventions for coronary heart disease prevention: what have we learned since the North Karelia project? Curr Opin Cardiol 2008 23 5 452 61 10.1097/HCO.0b013e32830c217e 18670256 12. World Health Organisation. NCD Global Monitoring Framework 2015. Available from: http://www.who.int/nmh/global_monitoring_framework/en/. 13. World Heart Federation World Heart Federation: 25 by 2025 Targets 2014 Melbourne World Heart Federation 14. National Rural Health Alliance of Australia. Healthy Horizons: a framework for improving the health of regional, remote and rural Australians, Outlook for 2003–2007. www.ruralhealth.org.au: National Rural Health Alliance, 2002 October 23rd, 2002. Report No. 15. Sellers DE Crawford SL Bullock K McKinlay JB Understanding the variability in the effectiveness of community heart health programs: a meta-analysis Soc Sci Med 1997 44 9 1325 39 10.1016/S0277-9536(96)00263-8 9141165 16. Australian Bureau of Statistics. Remoteness Structure: Australian Statistical Geography Standard Canberra, Australia: ABS; 2016 [cited 2016 06/07/2016]. Available from: http://www.abs.gov.au/websitedbs/D3310114.nsf/home/remoteness+structure#Anchor1. 17. Aoun S Rosenberg M Are rural people getting HeartSmart? Aust J Rural Health 2004 12 2 81 8 10.1111/j.1038-5282.2004.00553.x 15023226 18. Carrington MJ Stewart S Cardiovascular disease prevention via a nurse-facilitated intervention clinic in a regional setting: the Protecting Healthy Hearts Program Eur J Cardiovasc Nurs 2015 14 4 352 61 10.1177/1474515114537022 24847170 19. Higginbotham N Heading G McElduff P Dobson A Heller R Reducing coronary heart disease in the Australian Coalfields: Evaluation of a 10-year community intervention Soc Sci Med 1999 48 5 683 92 10.1016/S0277-9536(98)00384-0 10080368 20. Krass I Hourihan F Chen T Health promotion and screening for cardiovascular risk factors in NSW: a community pharmacy model Health Promot J Austr 2003 14 2 101 7 21. Ray R Self-reported heart health behaviour patterns in a rural context Aust J Rural Health 2001 9 2 59 63 10.1046/j.1440-1584.2001.00344.x 11259957 22. Burgess CP Sinclair G Ramjan M Coffey PJ Connors CM Katekar LV Strengthening cardiovascular disease prevention in remote indigenous communities in Australia’s northern territory Heart Lung Circ 2015 24 5 450 7 10.1016/j.hlc.2014.11.008 25541229 23. Kerr RM Leicht AS Spinks WL Effects of a 12-month exercise program on cardiorespiratory health indicators of Vietnam War veterans resident in the tropics Aust J Rural Health 2008 16 3 132 6 10.1111/j.1440-1584.2007.00934.x 18471182 24. Rowley KG Daniel M Skinner K Skinner M White GA O’Dea K Effectiveness of a community-directed ‘healthy lifestyle’ program in a remote Australian Aboriginal community Aust N Z J Public Health 2000 24 2 136 44 10.1111/j.1467-842X.2000.tb00133.x 10790932 25. Australian Institute of Health and Welfare Trends in coronary heart disease mortality: age groups and populations 2014 Canberra AIHW 26. Record NB Onion DK Prior RE Dixon DC Record SS Fowler FL Community-wide cardiovascular disease prevention programs and health outcomes in a rural county, 1970–2010 JAMA 2015 313 2 147 55 10.1001/jama.2014.16969 25585326 27. Record NB Harris DE Record SS Gilbert-Arcari J DeSisto M Bunnell S Mortality impact of an integrated community cardiovascular health program Am J Prev Med 2000 19 1 30 8 10.1016/S0749-3797(00)00164-1 10865161 28. Australian Bureau of Statistics. Registered Deaths of Aboriginal and Toress Strait Islander Australians Australia: ABS; 2012 [cited 2015 10/12/2015]. Available from: http://www.abs.gov.au/ausstats/abs@.nsf/Lookup/by%20Subject/3303.0~2014~Main%20Features~Deaths%20of%20Aboriginal%20and%20Torres%20Strait%20Islander%20Persons~10005. 29. Pain T Plummer D Pighills A Harvey D Comparison of research experience and support needs of rural versus regional allied health professionals Aust J Rural Health 2015 23 5 277 85 10.1111/ajr.12234 26381792 30. Law M Wald N Morris J Lowering blood pressure to prevent myocardial infarction and stroke: a new preventive strategy Health Technol Assess 2003 7 31 1 94 10.3310/hta7310 14604498 31. Brook RH Ware JE Jr Rogers WH Keeler EB Davies AR Donald CA Does free care improve adults’ health? Results from a randomized controlled trial N Engl J Med 1983 309 23 1426 34 10.1056/NEJM198312083092305 6355851 32. Gold MR Franks P A community-oriented primary care project in a rural population: reducing cardiovascular risk J Fam Pract 1990 30 6 639 44 2345332 33. Australian Institute of Health and Welfare. Remoteness classifications 2015 [cited 2015 1/07/2015]. Available from: http://www.aihw.gov.au/rural-health-remoteness-classifications/. 34. Wakerman J Defining remote health Aust J Rural Health 2004 12 5 210 4 10.1111/j.1440-1854.2004.00607.x 15588265 35. Wallerstein N Duran B Community-based participatory research contributions to intervention research: the intersection of science and practice to improve health equity Am J Public Health 2010 100 Suppl 1 S40 6 10.2105/AJPH.2009.184036 20147663

PMC005xxxxxx/PMC5002214.txt

==== Front Acta Neuropathol CommunActa Neuropathol CommunActa Neuropathologica Communications2051-5960BioMed Central London 36510.1186/s40478-016-0365-9ResearchSpatio-temporal activation of caspase-8 in myeloid cells upon ischemic stroke Rodhe Johanna 1Burguillos Miguel A. 1de Pablos Rocio M. 2Kavanagh Edel 1Persson Annette 3Englund Elisabet 3Deierborg Tomas 4Venero Jose L. 2Joseph Bertrand Bertrand.Joseph@ki.se 11 Department of Oncology-Pathology, Cancer Centrum Karolinska, Karolinska Institutet, Stockholm, Sweden 2 Departamento de Bioquímica y Biología Molecular, Universidad de Sevilla, and Instituto de Biomedicina de Sevilla-Hospital Universitario Virgen del Rocío/CSIC/Universidad de Sevilla, 41012 Sevilla, Spain 3 Department of Clinical Sciences, Division of Oncology and Pathology, Department of Neuropathology, Lund University, Lund, Sweden 4 Experimental Neuroinflammation Laboratory, Department of Experimental Medical Science, Lund University, Lund, Sweden 26 8 2016 26 8 2016 2016 4 1 921 7 2016 11 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Ischemic stroke (caused by thrombosis, embolism or vasoconstriction) lead to the recruitment and activation of immune cells including resident microglia and infiltrating peripheral macrophages, which contribute to an inflammatory response involved in regulation of the neuronal damage. We showed earlier that upon pro-inflammatory stimuli, the orderly activation of caspase-8 and caspase-3/7 regulates microglia activation through a protein kinase C-δ dependent pathway. Here, we present in vivo evidence for the activation of caspase-8 and caspase-3 in microglia/macrophages in post-mortem tissue from human ischemic stroke subjects. Indeed, CD68-positive microglia/macrophages in the ischemic peri-infarct area exhibited significant expression of the cleaved and active form of caspase-8 and caspase-3. The temporal and spatial activation of caspase-8 was further investigated in a permanent middle cerebral artery occlusion mouse model of ischemic stroke. Increasing levels of active caspase-8 was found in Iba1-positive cells over time in the peri-infarct area, at 6, 24 and 48 h after artery occlusion. Analysis of post-mortem brain tissue from human subject who suffered two stroke events, referred as recent and old stroke, revealed that expression of cleaved caspase-8 and -3 in CD68-positive cells could only be found in the recent stroke area. Analysis of cleaved caspase-8 and -3 expressions in a panel of human stroke cases arranged upon days-after stroke and age-matched controls suggested that the expression of these caspases correlated with the time of onset of stroke. Collectively, these data illustrate the temporal and spatial activation of caspase-8 and -3 in microglia/macrophages occurring upon ischemic stroke and suggest that the expression of these caspases could be used in neuropathological diagnostic work. Electronic supplementary material The online version of this article (doi:10.1186/s40478-016-0365-9) contains supplementary material, which is available to authorized users. Keywords MicrogliaMacrophageCaspase-8Caspase-3Ischemic strokeHuman brain tissuepMCAO modelSpatio-temporal activationhttp://dx.doi.org/10.13039/501100004047Karolinska Institutethttp://dx.doi.org/10.13039/501100004359Vetenskapsrådethttp://dx.doi.org/10.13039/501100003792Hjärnfondenhttp://dx.doi.org/10.13039/100008444Parkinsonfondenhttp://dx.doi.org/10.13039/501100004047Karolinska Institutetspanish MINECO/FEDER/UEissue-copyright-statement© The Author(s) 2016 ==== Body Introduction Every year, an estimate of 15 million people worldwide suffers a stroke. As a result, nearly six million people die and an almost equal number of survivors are left with long-term disabilities [1, 2]. Stroke is an acute cerebrovascular accident, which occur due to deranged blood supply to the brain. There are two main types of stroke: ischemic, due to lack of blood flow, and hemorrhagic, due to bleeding. Ischemic stroke, which is caused by a vessel obstructive thrombosis, embolism or vasoconstriction, accounts for over 80 % of all incidents, is the focus of the present study [1–3]. A decrease or reduction in blood flow results in hypoxia and glucose deprivation, which can lead to neuronal damage and cell death. The center of the ischemic area, the ischemic core, is most affected by the reduction in blood flow and also suffers the more instant and severe damage of the tissue. The area surrounding the ischemic region, the penumbra, can receive low levels of blood flow from adjacent vascularized areas, resulting in slower development of neuronal damage. Injured and dying cells release damage-associated molecular patterns (DAMPs), which activate an immune response that is a major contributor to stroke pathophysiology. In fact, the immune response to acute cerebral ischemia triggers an inflammatory reaction that may last up to several months and plays a critical role in mediating post-ischemic damage of the tissue and secondary neurodegeneration in the penumbra [4]. The infiltration of blood-borne immune cells facilitated by disruption of the blood–brain barrier integrity following brain ischemic injury contributes to the neuroinflammation process. Nevertheless, the brain’s initial inflammatory response to ischemic event is primarily thought to be mediated by microglia, the brain resident immune cells. Microglia are highly dynamic cells, which constantly scavenge the brain for potential threats and can get rapidly activated upon detection of insults to the brain, danger-signals or changes in the brain microenvironment [5]. In response to the tissue damage, microglia become activated and migrate to the ischemic area. Microglia are a predominant source of proinflammatory mediators including cytokines (e.g. tumor necrosis factor and interleukin-1β), complement factors, free radicals, nitric oxide (NO), chemokines (e.g. CCL2 and CCL3) and prostaglandins, all of which potentially contribute to further neuronal dysfunction and death [6, 7]. Suppression of neuroinflammation using a variety of drugs were proven to be successful in reducing infarct volume and improving outcomes in experimental models of stroke [8]. Despite these promising preclinical trials, up to date, clinical trials using anti-inflammatory agents have failed to improve clinical outcomes [9]. Therefore, in order to revitalize interest for the therapeutic targeting inflammatory pathways for the treatment of acute ischemic stroke there is a need for comprehensive understanding of the time-dependent recruitment and activation of inflammatory immune cells. Caspases, a family of cysteinyl aspartate-specific proteases, are best known as executioners of apoptotic cell death and their activation are considered as a commitment to cell death. However, caspases also function as regulatory molecules for immunity, cell differentiation and cell-fate determination [10]. We previously reported the existence of a caspase-dependent signalling pathway controlling microglia pro-inflammatory activation and associated neurotoxicity. We showed that the orderly activation of caspases -8 and -3/7, commonly known to have executioner roles in apoptosis, can promote pro-inflammatory activation of microglia in the absence of cell death [11]. Additionally, we recently obtained evidence that the activation of human monocytes/macrophages in response to a pro-inflammatory stimulus also rely on caspase-8 function [12]. In addition, increasing evidence strongly suggests that caspase-8 may play a critical role in IL-1β processing, with special relevance to the NLRP3 inflammasome activation [13]. NLRP3 inflammasome activation usually requires two signals or steps; signal 1 (priming) leading to NF-kB activation and signal 2 (typically ATP or nigericin) leading to NLRP3 machinery assembly (including at least NLRP3, the ASC (PYCARD) adaptor, and caspase-1) [14]. As previously stated, we first demonstrated the involvement of caspase-8 in NF-kB activation (priming) in microglia in response to diverse inflammatory stimuli [11], recently validated using conditional caspase-8 knockout mice specifically in the myeloid system [15]. The activation of NF-kB is critical for upregulating the transcription of both pro-IL-1β and NLRP3, as both are further needed for inflammasome formation and activation. A role of caspase-8 in mediating priming and activation of the NLRP3 inflammasome has also been demonstrated in primary macrophages [16]. Coimmunoprecipitation and confocal studies have demonstrated that caspase-8 is present in the NLRP3 inflammasome complex, where it is believed to be involved in cleavage and processing of procaspase-1 [13]. It is important to note that IL-1β activates T-cell-mediated innate immunity and promotes secondary ischemic damage during the subacute phase of ischemic brain injury [17]. Given the high number of danger-associated molecular patterns (DAMPs) released as a consequence of the ischemic damage, and the key roles of caspases in regulating brain immune functions, it is important to discern between the apoptotic and non-apoptotic roles of caspases upon ischemic stroke. In the present study, we investigated in vivo and in post-mortem tissue from ischemic stroke subjects whether caspase-8 and caspase-3 activation, key players of the caspase-dependent signaling pathways regulating microglia and macrophages (MMs) pro-inflammatory activation, exhibited spatiotemporal features upon ischemic stroke. Materials and methods Human brain tissue Human brain tissue from stroke subjects and controls were used in this study as approved by the Regional Ethical Review Board in Lund, Sweden (Dnr 2010-196). Stroke tissue from regions within the white matter of the frontal and parietal lobe was examined and compared to area- and age-matched controls. Both female and male cases were included in the study with an average age of 75 ± 9 years for stroke cases (n = 9) and 75 ± 11 years for controls (n = 5). Deceased individuals referred for autopsy were examined for stroke or/and diffuse ischemic damage in frontoparietal white matter, judged as relevant for particular ischemia susceptibility in the upper border zone areas. Three of these patients died from a clinically suspected or also radiologically ascertained stroke within the last 24 h. In two other patients, an older stroke was known and also detected radiologically in the basal ganglia and the thalamus, however not reported in the white matter, from where the tissue was macroscopically identified and sampled. In four cases found to have ischemic white matter damage, the autopsy referrals focused on cardiopulmonary disease, including myocardial infarcts in some cases. In the latter individuals, macroscopic examination of the cut brain revealed either clear-cut or suspected infarcts, which were later microscopically verified. For sampling from focal stroke, an area of approximately 10 mm2 and 5 mm thickness in the immediate vicinity/border of the lesion was taken for analysis. For diffuse ischemic damage (non-focal lesion, no complete tissue loss) a similar amount of tissue was sampled in the most damaged area. The tissue was fixed in 4 % formaldehyde solution and embedded in paraffin, in accordance with the standard procedures within the department. Sections of 5 μm thickness were cut with microtome and mounted on glass slides. Tissue sections on regular slides were processed for standard/basic stainings (hematoxylin-eosin (HE) and luxol fast blue/cresyl violet (LFB) for myelin and cell structures. Tissue sections on positively charged glass slides were processed for immunohistochemistry. All brain sections were analyzed under the microscope for clinical diagnostic purpose before inclusion in the study. Animals and surgery C57BL/6 mice used in this study were obtained from the Center of Production and Animal Experimentation (Espartinas, Spain). Experiments were performed at the University of Seville (Spain) in accordance with the Guidelines of the European Union Council (86/609/EU), following Spanish regulation for the use of laboratory animals approved by the Scientific Committee of the University of Seville. Animals were kept at diurnal conditions with ad libitum access to food and water. Experiments were conducted using 3-month-old male mice (n = 13). Electrocoagulation was used to induce a permanent middle cerebral artery occlusion (pMCAO), as described elsewhere [18]. Six to 48 h post pMCAO, animals were perfused with 4 % paraformaldehyde, pH 7.4, during deep isoflurane anesthesia and brains were removed and cryoprotected in sucrose. After freezing in isopentane at −40 °C, brains were cut in 25 μm thick coronal sections with a cryostat and mounted onto gelatin coated slides. Immunohistochemistry Deparaffination of the human tissue sections was done using standard procedures in the department of Neuropathology. The sections were boiled during 15 min in 10 mM Citrate buffer pH 6.0 for antigen retrieval. Stainings were done using an automated immunostainer (TechMateTM 500 Plus, DAKO) with DAKO ChemMate Kit Peroxidase/3-3’diaminobenzidine. Mouse tissue sections were washed and incubated for 20 min in 0.3 % hydrogen peroxide in methanol. After washes, the sections were incubated in blocking buffer containing 1 % goat serum (Vector) in TBS for 1 h. Incubation with primary antibodies diluted in 1 % goat serum and 0.25 % Triton-X 100 in TBS for 24 h. 3 × 10 min TBS washes was followed by incubation with biotinylated secondary antibodies diluted in 0.25 % Triton-X 100 in TBS for 2 h. Sections were thereafter incubated with ExtrAvidin®-Peroxidase solution (Sigma) and visualized with a standard diaminobenzidine/hydrogen peroxidase reaction for 5 min. Immunofluorescence Deparaffination of the human tissue was done using 3 × 5 min incubations in Xylene, followed by 2 × 10 min in 100 % ethanol, 2 × 10 min in 95 % ethanol and 3 × 5 min in H2O. Antigen retrieval was done for all tissue used for immunofluorescent staining by boiling in 10 mM Citrate buffer pH 6.0 for 10 min at 95 °C. When sections reached RT they were washed in TBS, before blocking in 5 % goat serum and 0.1 % Triton-X100 in TBS for 1 h. Incubation of primary antibody diluted in blocking buffer was done overnight at 4 °C. Washes in 3 × 10 min TBS was followed by 1 h by appropriate secondary antibody in 0.1 % Triton-X100 in TBS in RT. Sections were washed in TBS before 15 min incubation in 1 μg/ml Hoechst 33342 in RT, and followed by TBS wash. Autofluorescence removal reagent (#2160 Millipore) was used for the human tissue according to manufacturer’s recommendations. Antibodies Primary antibodies detecting anti-human cleaved caspase-8 (Asp391) (18C8; #9496), anti-mouse cleaved caspase-8 (Asp387) (D5B2; #8592), anti-human and -mouse cleaved caspase-3 (Asp175) (5A1E; #9664), anti-human cleaved PARP (D64E10; #5625) were purchased from Cell Signaling Technology®, anti-human CD68 (PG-M1; #M0876) from Dako and anti-mouse Iba-1 (#NB100-1028) from Novus Biologicals and (#019-19741) from Wako. For immunofluorescence detection, fluorophore conjugated secondary antibodies were purchased from Invitrogen, ThermoFisher Scientific. Goat anti-rabbit Alexa Fluor®488 and goat anti-mouse Alexa Fluor®594 were used for the human tissue analysis, whereas donkey anti-rabbit Alexa Fluor®488 and donkey anti-goat Alexa Fluor®594 were used for mouse tissue analysis. Biotinylated goat anti-rabbit IgG were obtained from Vector Laboratories. Tissue analysis Analysis of human tissue by Immunohistochemistry was done by a neuropathologist with a light microscope and scored for presence of cleaved caspases and upregulation of CD68 in the peri-infarct and ischemic area. Presence of cleaved caspase-8 and -3 as well as increased CD68 was semiquantitative assessed based on staining intensity of each antibody in both peak foci and the entire sampled region. The scoring was done blinded to other clinical information. Additional hematoxylin & eosin (HE) and luxol fast blue/cresyl violet (LFB) staining of all tissue was used to identify the stroke area and to evaluate the age of the ischemic lesions. The same areas were examined for the double immunofluorescent labelling of cells by Zeiss LSM510 and Zeiss LMS700 confocal laser scanning microscopy equipped with inverted Zeiss Axiovert 200 m microscopes using Zeiss LSM 5 and Zeiss ZEN 7.1 software. Immunohistochemistry cell quantification Iba1 and cleaved caspase-8 positive cells were quantified after IHC staining. Analysis was performed using the 48 h post-occlusion time point on four animals. Two sections per animal were quantified and 3 fields per section were counted for each area (ischemic core, peri-infarct area and surrounding area). The peri-infarct area is defined as the region surrounding the ischemic core, approximately 500 μm wide, with a distinct activation of Iba1+ cells (illustrated in Fig. 1). The surrounding area is the region distally outside the peri-infarct area, where areas for quantification are selected approximately 1 mm outside the peri-infarct area. Cells were divided after morphological appearance as ramified, hypertrophic, amoeboid or rounded.Fig. 1 Iba1-positive myeloid cells in the ischemic peri-infarct area express cleaved caspase-8 in a pMCAO mouse model of ischemic stroke. Consecutive frozen tissue sections of the ischemic core and the peri-infarct regions from mice after 48 h of permanent middle cerebral artery occlusion (pMCAO) were analyzed by immunohistochemistry with a mouse specific antibody recognizing active caspase-8 when cleaved at Asp 387, or an Iba1 antibody used to detect myeloid cells (a). Double immunofluorescent staining demonstrated the co-localization of cleaved caspase-8 and the myeloid cell marker Iba1 (b). Scale bars represent 850 μm (a) and 300 μm (b) Results Temporal and spatial activation of caspase-8 in Iba1-positive myeloid cells in a pMCAO mouse model of ischemic stroke To investigate activation of caspase-8 in microglia/macrophages (MMs) upon a stroke event, a mouse model with a permanent middle cerebral artery occlusion (pMCAO) was used to simulate an ischemic stroke. Immunohistochemical labeling of Iba1 was used to detect the MMs present in the ischemic core, peri-infarct area and surrounding area (Fig. 1a). Active caspase-8 was detected using an antibody that recognizes the mouse caspase-8 when proteolytically activated by cleavage at amino acid residue Asp387. Positivity for cleaved caspase-8 was detected in the ischemic core and the peri-infarct area (Fig. 1a) and could be associated with the Iba1-positive cells, as illustrated with immunofluorescence double staining confirming expression of cleaved caspase-8 in Iba1-positive cells upon stroke (Fig. 1b). The brain's inflammatory response post-ischemia is characterized by several stages, which is also reflected in the recruitment and activation state of the involved immune cells. Therefore, morphological changes were analyzed in Iba1-positive cells and related to the proximity to the ischemic area. Early-stage activation is characterized by increased ramification of cytoplasmic processes and cell size and enhanced Iba1 labeling. This is followed by further thickening of processes and retraction of finer ones and increased cell body size, to end with complete retraction of cytoplasmic processes to acquire an amoeboid-shape morphology. Quantifications of Iba1-positive cells revealed highest numbers of ramified or hypertrophic cells in the surrounding area, with increasing numbers of amoeboid cells in the peri-infarct area and of rounded cells in the ischemic core, indicating higher degree of activation of cells in the proximity to the ischemic area (Fig. 2a-c). Interestingly, these morphological changes could even be visualized with the cleaved caspase-8 staining and quantified after morphological appearance. Here the largest number of cleaved caspase-8 positive cells displayed an amoeboid and rounded morphology and were present in the peri-infarct area and in the ischemic core at 48 h post occlusion (Fig. 2a-b and d). The temporal aspect of caspase-8 activation in MMs was investigated during the acute and subacute phase of inflammation, using murine brain tissue samples analyzed at 6, 24 and 48 h post artery occlusion. Morphological changes indicative of an activation of the Iba1-positive MMs were observed already at 6 h in the peri-infarct area, but were even more prominent at 24 and 48 h after occlusion (Fig. 3a). The contralateral side of the brain was used for comparison of the occurring changes in terms of cell numbers and morphological changes of the cells close to the ischemic area. Very low levels of cleaved caspase-8 could be detected at the 6 h’ time point in the peri-infarct area, but those levels were found to be significantly increased at 24 and 48 h’ time point (Fig. 3a). Double immunofluorescence staining for cleaved caspase-8/Iba1 and confocal imaging illustrated that cleaved caspase-8 was rarely detected at the 6 h occlusion, but more frequently in the cytosol of Iba1-positive cells at the 24 and 48 h time points.Fig. 2 Expression of active caspase-8 correlates with the morphological transformation of Iba1-positive myeloid cells. Schematic overview of the distribution of myeloid cells in the 1) ischemic core 2) peri-infarct area and 3) surrounding area (a). Cell morphologies were divided into ramified, hypertrophic, amoeboid and rounded to reflect the states of activation. Examples of the different cell morphologies are given as cartoon and upon immunohistochemical staining for cleaved caspase-8 (b). Morphological quantifications of Iba1-positive cells (c) and cleaved caspase-8-positive cells (d) defined as cells per mm2 in the ischemic core (1), peri-infarct area (2) and surrounding area (3) in tissue from mice (n = 4) upon 48 h of pMCAO Fig. 3 Temporal activation of caspase-8 in Iba1-positive myeloid cells. Frozen tissue sections from mice upon pMCAO (n = 9) were analyzed by immunohistochemistry for expression of active caspase-8 or for expression of the microglia/macrophage marker Iba1. The contralateral side to the ischemic area was used an internal control for each animal. A small increase in body size of Iba1 positive cells was detected in the peri-infarct area at 6 h, with larger increase in cell body and numbers of cells expressing Iba1 detected at 24 and 48 h post occlusion (a). Cleaved caspase-8 was detected at low levels at 6 h, but a significant increase in expression levels was observed at 24 and 48 h in the peri-infarct area, but not at the contralateral side (a). Double immunofluorescent staining revealed that increased active caspase-8 expression was localized in Iba1-positive cells (b). Scale bar represent 100 μm in IHC images (a) and 10 μm in IF images (b) Expression of active caspase-8 and caspase-3 in CD68-positive myeloid cells in stroke subjects correlates with age of the ischemic area We were then interested to examine if activation of the apical caspase-8 and its downstream target caspase-3 could also be detected in MMs in human subjects who had suffered a stroke. Post-mortem brain tissue from subjects with a recent ischemic stroke was examined using immunohistochemical stainings. CD68 was used to detect activated MMs and revealed the presence of large numbers of inflammatory CD68-positive cells with rounded macrophage-like morphology in the peri-infarct area and ischemic core (Fig. 4a and Additional file 1a). Detection of active caspase-8 was performed using an antibody which recognizes human caspase-8 when cleaved at amino acid residue Asp391 (for antibody positive controls, see Additional file 2). Positive signal for cleaved caspase-8 was detected in numerous cells in the stroke peri-infarct area, as well as some positivity in the ischemic core (Fig. 4a and Additional file 1a). Expression of activated caspase-3 was also detected in the same areas using an antibody recognizing human caspase-3 when proteolytically activated by cleavage at amino acid residue Asp175 (Fig. 4a and Additional file 1a). In order to validate the presence of those active caspases in the CD68-positive cells, double immunofluorescent labelling for CD68 and cleaved caspase-8 or caspase-3 were performed. In fact, confocal imaging demonstrated the presence of both cleaved caspase-8 and cleaved caspase-3 in the cytoplasm of CD68-positive cells, confirming the activation for those caspases in the MMs upon stroke (Fig. 4b-c and Additional file 1b-c). Interestingly, for two stroke cases we also got the opportunity to analyze tissue samples from an older stroke area within the same cases. The subjects who had suffered two different stroke events are hereafter referred to as recent and old stroke. In contrast to the recent stroke, investigation of the older stroke area revealed low number and levels of CD68-positive cells still present around the ischemic area (Fig. 4a and Additional file 1a, compare middle “recent stroke area” with “old stroke area” rows). Expression for cleaved caspase-8 and caspase-3 was barely detected in the peri-infarct at this stage. Double staining for the markers confirm the absence/low levels of these active caspases in remaining CD68-positive cells (Fig. 4a and Additional file 1a, compare middle “recent stroke area” with “old stroke area” rows).Fig. 4 Presence of active caspase-8 and caspase-3 in CD68-positive cells correlates with age of the ischemic area. Paraffin imbedded brain tissue sections from subjects who suffered from two strokes, referred as recent and old stroke areas, were analyzed by immunohistochemistry for active caspase-8 and caspase-3 expressions using antibodies raised against caspase-8 when cleaved at Asp391 and human caspase-3 when cleaved at Asp175 and compared to healthy control tissue. Cleaved caspase-8 and -3 were detected in tissue from the recent stroke area, but not in healthy control tissue (a). Numbers and levels of CD68-positive cells were markedly increased in recent stroke tissue as compared to control (a), and found to be positive for cleaved caspase-8 and -3 using double immunofluorescence staining analyses (b-c). In contrast, for the older stroke area, lower levels of CD68 were detected as well as very low level/absence for cleaved caspase-8 and -3 (a). Immunofluorescence analysis confirmed the decrease/absence of active caspase-8 and -3 in the CD68-expressing cells within the older stroke area (b-c). Scale bar for IHC images corresponds to 100 μm (a) and for IF images 10 μm (b-c) Under cell death conditions, activated caspase-8 is known to cleave and activate caspase-3, which in turn can cleave PARP and contribute to the demise of the cell. To demonstrate absence of a caspase-dependent cell death in the CD68-positive myeloid cells, tissue from control subjects and recent stroke areas in patients were labelled with antibodies detecting cleaved PARP and CD68. We could not detect PARP cleavage in CD68 positive cells either in the ischemic core or peri-infarct area after stroke, indicating that the caspase activation in these cells is not linked to cell death (Fig. 5). PARP cleavage was, however, detected in the colon tissue used as positive control (Fig. 5).Fig. 5 Absence of cleaved PARP, marker for apoptosis, in CD68-positive myeloid cells in stroke subject. Tissue from recent stroke case and healthy control subjected to double immunofluorescence staining using antibodies raised against CD68 and cleaved PARP, revealed an absence for the apoptosis marker in CD68-expressing myeloid cells. Tissue from colon was used as positive control for the cleaved PARP staining. Overview of the peri-infarct area of stroke case is depicted on top panel, whereas higher magnification images for both stroke case and controls are presented in the lower panels. Scale bars in images represent 10 μm In a panel of 9 subjects who had suffered a stroke, post mortem tissue from 10 different stroke areas of various ages (i.e. time after stroke onset) were examined. Evaluation of the tissue by hematoxylin and eosin staining was used to detect and determine the age of the stroke areas (not included in manuscript). In addition, five area and age-matched controls were also included in the investigation. Immunohistochemical staining for CD68, active caspase-8 and active caspase-3 was performed as described above. Scoring of the tissue for expression of the different markers was done by an experienced neuropathologist. The following scale was used for scoring: not present (-), present (+) or high levels (++) of active caspases. In the human stroke subjects, highest levels of active caspase-8 and active caspase-3 could be detected the first day after the stroke, and then slowly decreases with time until complete disappearance after 5–29 days after the ischemic event (Table 1). CD68-positivity was reported as upregulated (Yes) or basal/very low levels (No). The high numbers of CD68-positive cells were, as the caspases, detected at the earlier time points after the ischemic event, but declined with time and are back to basal levels (as compared to controls) in most subjects after 5–29 days after onset (Table 1).Table 1 Temporal expression of active caspase-8 and caspase-3 in a panel of stroke subjects Cleaved Caspase-8 Cleaved Caspase-3 CD68 upregulated Days after stroke Case 1 ++ ++ Yes 0–1 Case 2 ++ + Yes 0–1 Case 3 + + Yes 0–1 Case 4 + + Yes 1–4 Case 5 + – Yes 1–4 Case 6 + – Yes 5–29 Case 7a + – No 5–29 Case 8 + – No 5–29 Case 7b – – No 30– Case 9 – – No 30– Control 1 – – No Control 2 – – No Control 3 – – No Control 4 – – No Control 5 – – No Tissue from 10 different areas from stroke subjects (Case; n = 9) and 5 areas from healthy controls (n = 5) were analyzed for CD68, cleaved caspase-8 and cleaved caspase-3 expression by immunohistochemistry (exemplified in Fig. 4a). Cleaved caspase-8 or -3 expression levels were scored as (-) not present, (+) present or (++) high levels. CD68 expression was represented as (No) not present/basal levels or (Yes) increase in CD68 positive cells. Age of stroke area was determined by hematoxylin and eosin staining and is presented as days after stroke event. Highest levels of cleaved caspase-8 and-3 expressions were found within the first days after stroke. They were found to decrease with time, and were completely gone within 30 days. CD68-positive cells can be found at high numbers within the first days after stroke and decrease to basal levels within 30 days after stroke onset Discussion There is compelling evidence that brain injury following ischemic stroke develops from a complex series of pathophysiological events that evolve in time and space [19, 20]. After an ischemic stroke, experimental and clinical data suggest that a prominent inflammatory response develops, propagates, and lasts for many days, and is believed to exacerbate neuronal cell death [21, 22]. The brain’s initial inflammatory response to stroke is proposed to be primarily mediated by microglia, the resident immune cells of the CNS. However, within minutes or hours of the stroke event, the blood–brain barrier is compromised and infiltration of monocytes, neutrophils and lymphocytes occurs [23, 4, 24]. The brain-resident (microglia) and infiltrating peripheral (monocytes) myeloid cells have a prominent role in initiating, sustaining and resolving post-ischemic inflammation. It is therefore of importance to elucidate the molecular mechanism regulating their activation. Our team previously described an unexpected novel function for caspases in the control of microglia activation and thereby neurotoxicity. We showed that orderly activation of caspase-8 and caspase-3 regulates microglia activation, in the absence of cell death [11]. In addition, we recently obtained evidence that caspase-8 regulates the activation of human monocytes [12]. Considering the central role played by these caspases in the activation of microglia/monocytes, and the contribution of these cells in the observed inflammatory response following ischemic stroke, we decided to investigate whether activation of these caspases follow spatial and temporal features. Immunohistochemical staining, as well as immunofluorescence confocal imaging, of post-mortem tissues from subjects who had suffered an ischemic stroke, was used with a CD68-antibody to detect activated myeloid cells. Additional staining with cleaved caspase-8 or cleaved caspase-3 revealed that myeloid cells in the ischemic core and peri-infarct area expressed active caspase-8 and caspase-3. It is believed that non-apoptotic functions of caspases rely on a moderate activity and a restricted subcellular localization. We have demonstrated that a differential processing of caspase-3 zymogen may ultimately lead to apoptosis (caspase-3 subunit p17; nuclear localization) or microglia activation (caspase-3 subunit p19; cytosolic localization) [25]. Our confocal analysis demonstrated a non-nuclear localization of active caspase-3 within myeloid cells early after stroke, a view that fits well with the non-apoptotic role of caspases in regulating myeloid cell activation. Analysis of brain tissue samples from a pMCAO mouse model of ischemic stroke, at 6, 24 and 48 h post artery occlusion, illustrated a temporal and spatial activation for caspase-8 in Iba1-positive myeloid cells. Indeed, increased levels for cleaved caspase-8 staining were found to correlate with morphological changes of the Iba1-positive cells from ramified cells to amoeboid or rounded shapes in proximity to the ischemic core. Notably, this correlation was particularly evident in the peri-infarct area, a region revealing penumbra like conditions and is potentially salvable upon a brain infarct, in contrast to the stroke core where perfusion is completely absent and irreversible loss of tissue (infarction) occurs within minutes [26]. It has been long established that microglia activation is particularly evident in the penumbra region in response to ischemic damage [19]. Although the contribution of the inflammatory response to ischemic brain injury is under debate, increasing evidence points out a deleterious role. For instance, recent data have demonstrated how microglia in the penumbra region is strongly associated with blood vessels; reactive microglia phagocytose endothelial cells, thus contributing to BBB disruption and neurodegeneration [27]. In the present study we demonstrate the potential contribution of caspases in regulating brain immune function in the peri-infarct penumbra-like region. The multifaceted roles of caspase-8 make this caspase very attractive in the context of future stroke strategies aimed at minimizing brain injury. These varied roles include i) it controls microglia activation through a caspase-3/NF-kB-dependent mechanism, with the subsequent release of neurotoxic proinflammatory cytokines [11], ii) recent data suggest a prominent role of this caspase in the priming and activation of inflammasomes [13], iii) inflammasome activation may lead to pyroptosis, a proinflammatory and lytic mode of cell death and iiii) caspase-8 negatively regulates programmed necrosis or necroptosis, which relies of a molecular platform known as necrosome and strongly associated to immune cells of myeloid origin including microglia, monocytes and macrophages [28]. Since the repair process for damaged brain tissues and regeneration of neural cells takes place during resolution of inflammation [17], and having demonstrated that active caspase-8 is mostly confined to reactive myeloid cells in the peri-infarct region, we anticipate that site-directed delivery of caspase-8 inhibitors under stroke conditions may hinder microglia activation and affect microglia survival. Finally, our analysis of post-mortem tissues from subjects who suffered two independent ischemic stroke events, as well as the examination of panel of 10 different stroke areas of various ages (i.e. time after stroke onset) suggested that the presence of active caspase-8 and -3 in CD68-positive cells correlates with the age of ischemic area. The approximate time after stroke onset can be determined by neuropathologists and forensic pathologists by the postmortem analysis of ischemic lesions. Parameters used may include analysis of inflammation and myeloid component [29, 30]. The distinct pattern of cleaved caspase-8 surrounding the infarct in our data suggests that immunohistochemical analysis of cleaved caspase-8 could be considered as a relevant additional diagnostic parameter. Whereas inhibition of caspase-8 and/or caspase-3 could be used as therapeutic strategy to combat the inflammatory response initiated upon ischemic stroke remains controversial. Evidence for neuroprotection have been reported with Quinolyl-valyl-O-methylaspartyl-[-2,6-difluorophenoxy]-methylketone (Q-VD-OPh) a third-generation broad spectrum caspase inhibitor [31], which is able to crosses the blood–brain barrier, in animal models of stroke [32]. However, it remains to be established whether Q-VD-OPh affects microglia/monocytes activation in this disease model and contributes to the neuroprotective effect in that manner. In addition, Q-VD-OPh treatment has been reported to impair the neural stem/progenitor cell response after cortical ischemia in mice, highlighting that caution is warranted using such strategy [33]. Conclusion In conclusion, we revealed the temporal and spatial activation of caspase-8 and -3 in microglia/macrophages occurring upon ischemic stroke in human stroke subjects and in a mouse model of ischemic stroke. Overall, the report of a spatio-temporal activation of caspase-8 and -3 in microglia/macrophages occurring upon ischemic stroke indicates that any attempt to target the molecular signaling regulating the detrimental inflammatory response would have to take into account the time window for intervention. Additional files Additional file 1: Figure S1. Analysis of a second subject whom suffered from two stroke events and a healthy control. Immunohistochemical labelling for CD68, cleaved caspase-8 and cleaved caspase-3 (a), as well as double immunofluorescent labelling for CD68 and cleaved caspase-8 alternatively cleaved caspase-3 (b-c) in recent and old stroke areas are depicted. IHC image scale bar corresponds to 100 μm (a) and IF images to 10 μm (b-c). (TIF 4099 kb) Additional file 2: Figure S2. Positive controls for immunohistochemistry antibodies. Antibodies detecting cleaved caspase-8 and cleaved caspase-3 showed positive signal in sections from human colon tissue. Antibody detecting CD68 showed positive signal in sections from human lymph node. Scale bar represents 100 μm. (TIF 1191 kb) Abbreviations CNSCentral nervous system DAMPsDanger-associated molecular patterns HEHematoxylin-eosin IFImmunofluorescence IHCImmunohistochemistry LFBLuxol fast blue/cresyl violet MMMicroglia/macrophages pMCAOPermanent middle cerebral artery occlusion Q-VD-OPhQuinolyl-valyl-O-methylaspartyl-[-2,6-difluorophenoxy]-methylketone Acknowledgements We thank the CLICK Imaging Facility supported by the Knut and Alice Wallenberg Foundation. Funding J.R. is supported by a doctoral fellowship from the Karolinska Institutet Foundations; M.A.B. is supported by a postdoctoral fellowship from Swedish Research Council. This work has been supported by grants from the Swedish Research Council, the Swedish Brain Foundation, the Parkinson foundation in Sweden, the Spanish MINECO/FEDER/UE and the Karolinska Institutet Foundations. Availability of data and material Not applicable. Authors’ contributions JR performed all the experiments except otherwise noted. TD performed the surgery and further dissection of the animal brains. EE did the neuropathology analysis of the individuals with stroke and control cases. AP prepared tissue and participated in the morphological assessment of human brain specimens. MAB, JLV and RMP contributed with in vivo analyses. MAB and EK were involved in study design. JR, JLV and BJ designed the study, analyzed and interpreted the data. JR and BJ wrote the first draft of the manuscript. All authors discussed the results and commented on or edited the manuscript. All authors read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. Consent for publication Not applicable. Ethics approval Used of human brain tissue from stroke subjects and controls was approved by the Regional Ethical Review Board in Lund, Sweden (Dnr2010-196). Animal experiments were performed in accordance with the Guidelines of the European Union Council (86/609/EU), following Spanish regulation for the use of laboratory animals approved by the Scientific Committee of the University of Seville. ==== Refs References 1. Feigin VL Forouzanfar MH Krishnamurthi R Mensah GA Connor M Bennett DA Moran AE Sacco RL Anderson L Truelsen T O'Donnell M Venketasubramanian N Barker-Collo S Lawes CM Wang W Shinohara Y Witt E Ezzati M Naghavi M Murray C Global Burden of Diseases Ij, and Risk Factors Study 2010 (GBD 2010) and the GBD Stroke Experts Group Global and regional burden of stroke during 1990–2010: findings from the Global Burden of Disease Study 2010 Lancet 2014 383 9913 245 254 10.1016/S0140-6736(13)61953-4 24449944 2. Collaborators GBoDS Global, regional, and national incidence, prevalence, and years lived with disability for 301 acute and chronic diseases and injuries in 188 countries, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013 Lancet 2015 386 9995 743 800 10.1016/S0140-6736(15)60692-4 26063472 3. Donnan GA Fisher M Macleod M Davis SM Stroke Lancet 2008 371 9624 1612 1623 10.1016/S0140-6736(08)60694-7 18468545 4. Iadecola C Anrather J The immunology of stroke: from mechanisms to translation Nat Med 2011 17 7 796 808 10.1038/nm.2399 21738161 5. Kreutzberg GW Microglia: a sensor for pathological events in the CNS Trends Neurosci 1996 19 8 312 318 10.1016/0166-2236(96)10049-7 8843599 6. Jin R Yang G Li G Inflammatory mechanisms in ischemic stroke: role of inflammatory cells J Leukoc Biol 2010 87 5 779 789 10.1189/jlb.1109766 20130219 7. Weinstein JR Koerner IP Möller T Microglia in ischemic brain injury Future Neurol 2010 5 2 227 246 10.2217/fnl.10.1 20401171 8. Nilupul Perera M Ma HK Arakawa S Howells DW Markus R Rowe CC Donnan GA Inflammation following stroke J Clin Neurosci 2006 13 1 1 8 10.1016/j.jocn.2005.07.005 16410192 9. Tuttolomondo A Pecoraro R Arnao V Maugeri R Iacopino DG Pinto A Developing drug strategies for the neuroprotective treatment of acute ischemic stroke Expert Rev Neurother 2015 15 11 1271 1284 10.1586/14737175.2015.1101345 26469760 10. Venero JL Burguillos MA Joseph B Caspases playing in the field of neuroinflammation: old and new players Dev Neurosci 2013 35 2–3 88 101 10.1159/000346155 23445938 11. Burguillos MA Deierborg T Kavanagh E Persson A Hajji N Garcia-Quintanilla A Cano J Brundin P Englund E Venero JL Joseph B Caspase signalling controls microglia activation and neurotoxicity Nature 2011 472 7343 319 324 10.1038/nature09788 21389984 12. Oliva-Martin MJ Sanchez-Abarca LI Rodhe J Carrillo-Jimenez A Vlachos P Herrera AJ Garcia-Quintanilla A Caballero-Velazquez T Perez-Simon JA Joseph B Venero JL Caspase-8 inhibition represses initial human monocyte activation in septic shock model 2016 13. Gurung P Kanneganti TD Novel roles for caspase-8 in IL-1β and inflammasome regulation Am J Pathol 2015 185 1 17 25 10.1016/j.ajpath.2014.08.025 25451151 14. Man SM Kanneganti TD Regulation of inflammasome activation Immunol Rev 2015 265 1 6 21 10.1111/imr.12296 25879280 15. Kavanagh E Burguillos MA Carrillo-Jimenez A Oliva-Martin MJ Santiago M Rodhe J Joseph B Venero JL Deletion of caspase-8 in mouse myeloid cells blocks microglia pro-inflammatory activation and confers protection in MPTP neurodegeneration model Aging (Albany NY) 2015 7 9 673 689 10.18632/aging.100805 26405176 16. Gurung P Anand PK Malireddi RK Vande Walle L Van Opdenbosch N Dillon CP Weinlich R Green DR Lamkanfi M Kanneganti TD FADD and caspase-8 mediate priming and activation of the canonical and noncanonical Nlrp3 inflammasomes J Immunol 2014 192 4 1835 1846 10.4049/jimmunol.1302839 24453255 17. Shichita T Ito M Yoshimura A Post-ischemic inflammation regulates neural damage and protection Front Cell Neurosci 2014 8 319 10.3389/fncel.2014.00319 25352781 18. Lambertsen KL Clausen BH Babcock AA Gregersen R Fenger C Nielsen HH Haugaard LS Wirenfeldt M Nielsen M Dagnaes-Hansen F Bluethmann H Faergeman NJ Meldgaard M Deierborg T Finsen B Microglia protect neurons against ischemia by synthesis of tumor necrosis factor J Neurosci 2009 29 5 1319 1330 10.1523/JNEUROSCI.5505-08.2009 19193879 19. Dirnagl U Iadecola C Moskowitz MA Pathobiology of ischaemic stroke: an integrated view Trends Neurosci 1999 22 9 391 397 10.1016/S0166-2236(99)01401-0 10441299 20. Moskowitz MA Lo EH Iadecola C The science of stroke: mechanisms in search of treatments Neuron 2010 67 2 181 198 10.1016/j.neuron.2010.07.002 20670828 21. Wang Q Tang XN Yenari MA The inflammatory response in stroke J Neuroimmunol 2007 184 1–2 53 68 10.1016/j.jneuroim.2006.11.014 17188755 22. Ceulemans AG Zgavc T Kooijman R Hachimi-Idrissi S Sarre S Michotte Y The dual role of the neuroinflammatory response after ischemic stroke: modulatory effects of hypothermia J Neuroinflammation 2010 7 74 10.1186/1742-2094-7-74 21040547 23. Ritzel RM Patel AR Grenier JM Crapser J Verma R Jellison ER McCullough LD Functional differences between microglia and monocytes after ischemic stroke J Neuroinflammation 2015 12 106 10.1186/s12974-015-0329-1 26022493 24. Benakis C Garcia-Bonilla L Iadecola C Anrather J The role of microglia and myeloid immune cells in acute cerebral ischemia Front Cell Neurosci 2014 8 461 25642168 25. Kavanagh E Rodhe J Burguillos MA Venero JL Joseph B Regulation of caspase-3 processing by cIAP2 controls the switch between pro-inflammatory activation and cell death in microglia Cell Death Dis 2014 5 e1565 10.1038/cddis.2014.514 25501826 26. Dwyer TA Earl DE Wang L The utility of a new in vitro model of the stroke penumbra J Neurosci 2008 28 26 6537 6538 10.1523/JNEUROSCI.1841-08.2008 18579726 27. Jolivel V Bicker F Binamé F Ploen R Keller S Gollan R Jurek B Birkenstock J Poisa-Beiro L Bruttger J Opitz V Thal SC Waisman A Bäuerle T Schäfer MK Zipp F Schmidt MH Perivascular microglia promote blood vessel disintegration in the ischemic penumbra Acta Neuropathol 2015 129 2 279 295 10.1007/s00401-014-1372-1 25500713 28. Pasparakis M Vandenabeele P Necroptosis and its role in inflammation Nature 2015 517 7534 311 320 10.1038/nature14191 25592536 29. Love S Autopsy approach to stroke Histopathology 2011 58 3 333 351 10.1111/j.1365-2559.2010.03614.x 20666847 30. Mena H Cadavid D Rushing EJ Human cerebral infarct: a proposed histopathologic classification based on 137 cases Acta Neuropathol 2004 108 6 524 530 10.1007/s00401-004-0918-z 15517310 31. Chauvier D Ankri S Charriaut-Marlangue C Casimir R Jacotot E Broad-spectrum caspase inhibitors: from myth to reality? Cell Death Differ 2007 14 2 387 391 10.1038/sj.cdd.4402044 17008913 32. Braun JS Prass K Dirnagl U Meisel A Meisel C Protection from brain damage and bacterial infection in murine stroke by the novel caspase-inhibitor Q-VD-OPH Exp Neurol 2007 206 2 183 191 10.1016/j.expneurol.2007.03.032 17585906 33. Osman AM Neumann S Kuhn HG Blomgren K Caspase inhibition impaired the neural stem/progenitor cell response after cortical ischemia in mice 2015

PMC005xxxxxx/PMC5002215.txt

==== Front J NeuroinflammationJ NeuroinflammationJournal of Neuroinflammation1742-2094BioMed Central London 68510.1186/s12974-016-0685-5ResearchmiR-17-92 facilitates neuronal differentiation of transplanted neural stem/precursor cells under neuroinflammatory conditions Mao Susu maosusu@ntu.edu.cn 12Li Xiuhua mg1430040@smail.nju.edu.cn 1Wang Jin dz1430028@smail.nju.edu.cn 1Ding Xin dg1530007@smail.nju.edu.cn 1Zhang Chenyu cyzhang@nju.edu.cn 1http://orcid.org/0000-0002-8230-4598Li Liang 86-25-89681330lijing@sibs.ac.cn 11 State Key Laboratory of Pharmaceutical Biotechnology, Collaborative Innovation Center of Chemistry for Life Sciences, Jiangsu Engineering Research Center for MicroRNA Biology and Biotechnology, NJU Advanced Institute for Life Sciences (NAILS), School of Life Sciences, Nanjing University, 163 Xianlin Road, Nanjing, Jiangsu 210023 China 2 Jiangsu Key Laboratory of Neuroregeneration, Co-Innovation Center of Neuroregeneration, Nantong University, Nantong, 226001 China 27 8 2016 27 8 2016 2016 13 1 20819 3 2016 18 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Neural stem/precursor cells (NSCs) are of particular interest because of their potential application in cell therapy for brain damage. However, most brain injury cases are followed with neuroinflammatory stress, which affects the lineage selection of grafted NSCs by promoting astrocytogenesis, thus hampering the potential for neural replacement. The present study investigated the role of miR-17-92 in protecting against detrimental effects of neuroinflammation on NSC differentiation in cell therapy. Methods NSCs were treated with conditioned medium from lesioned astrocytes with/without neutralizing antibodies of leukemia inhibitory factor (LIF) or/and ciliary neurotrophic factor (CNTF), respectively. Afterward, the levels of p-STAT3 and p-JAK2 were determined by western blotting while expression of glial fibrillary acidic protein (GFAP) and β-tubulin III was assessed by immunostaining. The activation of JAK-STAT pathway and cell differentiation were also evaluated after we overexpressed miR-17-92 in NSCs under different neuroinflammatory conditions. After the transplantation of miR-17-92-overexpressing NSCs into injured mouse cortex, PH3, nestin, GFAP, and NeuN were analyzed by immunostaining. In addition, motor coordination of mice was evaluated by rotarod test. Results Conditioned medium from lesioned astrocytes activated JAK-STAT pathway and facilitated astrocytic differentiation in NSCs while neutralizing antibodies of LIF and CNTF remarkably attenuated such effects. miR-17-92 cluster repressed the expression of multiple proteins including GP130, CNTFR, JAK2, and STAT3 in JAK-STAT pathway. Overexpression of miR-17-92 in NSCs systematically blocked the activation of JAK-STAT pathway mediated by LIF and CNTF, which facilitated neuronal differentiation in vitro. Furthermore, miR-17-92 increased neuronal generation of grafted NSCs and reduced astrogliosis, which resulted in the improvement of motor coordination of brain-injured mice. Conclusions Our results suggest that miR-17-92 promotes neuronal differentiation of grafted NSCs under neuroinflammatory condition via inhibition of multiple proteins in JAK-STAT pathway. Electronic supplementary material The online version of this article (doi:10.1186/s12974-016-0685-5) contains supplementary material, which is available to authorized users. Keywords miR-17-92Neural stem/precursor cellsTransplantationNeuroinflammationJAK-STAT pathwayDifferentiationthe National Natural Science Foundation of China31471019Li Liang issue-copyright-statement© The Author(s) 2016 ==== Body Background Neural stem/precursor cells (NSCs) consist of a heterogeneous population of self-renewing progenitor/precursor cells that can be expanded in vitro, which further give rise to neurons, astrocytes, and oligodendrocytes after transplanting into the central nervous system (CNS) [1, 2]. Recent studies suggest that transplantation of NSCs has become a promising therapy for various neurological disorders including ischemia, traumatic brain injury, and several neurodegenerative diseases of CNS [3, 4]. The tightly regulated cellular processes, especially the neural differentiation, are essential for grafted NSCs to produce enough neurons to restore neural damage in CNS [5]. The differentiation of NSCs is the result of changes in stem-cell properties that are controlled by both extrinsic and intrinsic cues [6]. Proneural genes are intrinsic determinants that control neurogenesis while astrocytogenesis can be initiated by several extrinsic gliogenic signals, such as ciliary neurotrophic factor (CNTF), leukemia inhibitory factor (LIF), and bone morphogenetic proteins (BMPs) [6, 7]. Substantial evidences suggest that neurological disease or injury induces a neuroinflammatory response in the CNS, which has great impact on endogenous neurogenesis as well as differentiation of grafted NSCs [8, 9]. Ideguchi et al. showed that inflammatory response by the host brain suppresses neuronal differentiation of transplanted ES cell-derived neural precursor cells [10]. The low yield of NSCs-derived neurons limits the potential clinical application for neuronal replacement. In the condition of neuroinflammatory response, glia cells are activated which have been proved to affect neuronal survival [8]. In addition, it is also suggested that activated astrocyte can modulate NSCs differentiation [11]. As one of the key players mediating inflammatory response, astrocytes may produce several cytokines including LIF, CNTF, and interleukin 6 (IL-6) that have been proved to modulate neural differentiation in vitro [12]. Particular attention has been paid to LIF and CNTF because of their roles in the developmental switch from neurogenesis to astrocytogenesis [13]. LIF and CNTF promote premature generation of astrocytes in vitro via activation of the Janus kinase (JAK)-signal transducer and activator of transcription (STAT) pathway [14]. Furthermore, cytokine-induced STAT signaling directly activates JAK-STAT pathway in a positive, autoregulatory loop, resulting a potentiation of JAK-STAT-induced astrocytogenesis [15]. Therefore, we believe that the inhibition of JAK-STAT pathway in grafted NSCs under neuroinflammatory condition caused by brain injury may repress astrocytogenesis while enhancing the number of generated neurons, which may further exert beneficial effects on tissue regeneration and functional recovery. MicroRNAs are approximately 21–22 nucleotides which modulate genes expression by targeting 3′ untranslated region (3′ UTR) of mRNAs [16]. Substantial evidences suggest that miRNAs are enriched in CNS indicating their significant role in neural development and physiology [17, 18]. Our previous work demonstrated that miR-17 modulates neuron-astrocyte transition via inhibiting BMPR2 expression [19]. Additionally, there are evidences that a cluster of miRNAs repress multiple targets coordinated in the same pathway, which show that those small RNAs have great potential in modulating cell signals [20–22]. In the present study, we demonstrated that LIF and CNTF secreted by astrocyte in neuroinflammatory condition facilitated astrocytogenesis of NSCs through activation of JAK-STAT pathway. We also showed that miR-17-92 cluster effectively blocked the activation of LIF/CNTF-JAK-STAT signal by repressing multiple protein targets existed in this pathway, leading to the increase of neurogenesis of grafted NSCs in vivo. Such improvement also benefited motor activity in injured mice after NSC transplantation. Methods Embryonic neurosphere culture and lentiviral transduction Embryonic neurospheres (NSCs) were derived from the cortex of E14.5 wild-type C57BL/6J mice as previously described [19]. The cortex was dissected and mechanically dissociated, and the resulting single-cell suspension was placed in Dulbecco’s modified Eagle’s medium: nutrient mixture F-12 (DMEM/F12, Life Technologies, Grand Island, NY, USA) supplemented with N2 plus media supplement (R&D Systems, Minneapolis, MN, USA), 10 ng/ml recombinant bovine fibroblast growth factor (bFGF; R&D Systems), and recombinant human epidermal growth factor (hEGF; R&D Systems). After 5 days in culture, free-floating neurospheres were redissociated and allowed to reform spheres at least three times before further use. For differentiation studies, whole-mount or dissociated neurospheres were plated onto ploy-l-ornithine (Sigma-Aldrich, St. Louis, MO, USA) and human fibronectin (R&D Systems)-coated coverslips and further cultured in the absence of bFGF/hEGF. The lentiviral vector lenti-17-92 used for miR-17-92 cluster overexpression under the control of the elongation factor 1-alpha promoter and the respective control vector driving EGFP expression were generated by GenePharma Inc. (Shanghai, China). For lentiviral tranduction, fourth-passage neurospheres were single-cell dissociated, grown in suspension for 24 h, and then exposed to lenti-17-92 or control lentivirus for 10 h. Transduction efficiency was estimated by quantitative real-time PCR (QPCR). Quantitative real-time PCR Total RNA was isolated using TRIzol reagent (Life Technologies) according to the manufacturer’s instructions. Quantitative real-time PCR of mature miRNAs was performed using TaqMan microRNA probes (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s instructions. U6 snRNA was used for normalization in miRNA expression studies. All of the reactions were run in triplicates. A relative fold change in expression of miRNA was calculated with the equation 2−ΔCT. Astrocyte culture and preparation of conditioned medium Primary astrocytes were purified from neonatal P0 mice cortices and plated in DMEM/F12/10 % FBS in poly-l-lysine (Sigma-Aldrich)-coated 75 cm2 culture flasks, as previously described [23]. Medium was changed after 3 days in culture and thereafter three times per week. After incubating for 2 weeks, to obtain pure astrocytes, the mixed cells were orbitally shaken at 200 rpm for 6 h, and the supernatant was removed. The attached cells were washed twice, digested by trypsin and then replated. When the astrocytes have reached 90 % of confluence at passage 3, cells were washed and a defined serum-free medium was added, containing DMEM/F12 and 1 % N2 plus supplement. After 6 h of equilibration in the serum-free medium, the cultures were mechanically lesioned with a pipette tip in a 2-cm-gird frame. Conditioned medium (CM) was collected 48 h after the injury, filtered at 0.22 μm and immediately stored at −80 °C until later use. Immunocytochemical and immunohistochemical procedures Whole mount neurospheres and dissociated cells were fixed for 20 min in 4 % paraformaldehyde (PFA) followed by 1 h blocking with 5 % normal horse serum in phosphate-buffered saline (PBS)/0.3 % triton X-100. Incubation with primary antibodies was performed in 2 % BSA overnight at 4 °C. Cell nuclei were visualized with DAPI. For immunohistochemistry on tissue sections, mice were perfused transcardially with 50 ml PBS followed by 50 ml 4 % PFA. The brains were removed, post-fixed in the same fixative overnight at 4 °C, and cryoprotected in 15 % sucrose overnight followed by 30 % sucrose overnight. Cryostat sections (30 μm thick) were cut and processed for immunohistochemistry. Free-floating sections were incubated for 1 h in PBS/0.1 % triton X-100/5 % horse serum. The primary antibodies used on cells or tissue sections for multiple label immunofluorescence were as follows: rabbit polyclonal to glial fibrillary acidic protein (GFAP) (ab7779; 1:1000; Abcam, Hong Kong, China), chicken polyclonal anti-nestin (ab81755; 1:1000; Abcam), rabbit polyclonal anti-Sox2 (ab97959; 1:500; Abcam), mouse monoclonal to MAP2 (ab11267; 1:1000; Abcam), mouse monoclonal to CNPase (ab6319; 1:1000; Abcam), rabbit polyclonal anti-oligodendrocyte specific protein (ab53041; 1:1000; abcam), mouse monoclonal to neuron specific β-tubulin III (ab7751; 1:500; Abcam), mouse monoclonal anti-neuronal nuclei (NeuN) (MAB377; 1:100; Millipore, Billerica, MA, USA), rabbit polyclonal anti-p-histone H3 (sc-8656-R; 1:1000; Santa Cruz Biotechnology, Santa Cruz, CA), and rabbit polyclonal anti-Doublecortin (ab18723; 1:1000; Abcam). Following primary antibodies, the appropriate secondary antibodies were used: Alexa Fluor® 488-conjugated AffiniPure donkey anti-chicken IgY (IgG) (H+L) (703-545-155; 1:1000; Jackson ImmunoResearch, West Grove, PA, USA), Alexa Fluor® 594-conjugated AffiniPure donkey anti-chicken IgY (IgG) (H+L) (703-585-155; 1:1000; Jackson ImmunoResearch), Alexa Fluor® 488-conjugated donkey anti-mouse IgG (H+L) (A-21202; 1:1000; Life Technologies, Grand Island, NY, USA), Alexa Fluor® 594-conjugated donkey anti-rabbit IgG (H+L) (A-21207; 1:1000; Life Technologies), Alexa Fluor® 594-conjugated donkey anti-mouse IgG (H+L) (A-21203; 1:1000; Life Technologies). Finally, the cells or brain sections were stained with DAPI. For quantitative analysis, the numbers of each type of cell scored in four random fields were averaged, utilizing the Image-Pro Plus image analysis software. Western blot analysis The cultured cells were lysed using RIPA buffer (Thermo Scientific, Rockford, IL, USA), and the protein was extracted according to the manufacturer’s instructions. Total proteins were determined using a Bicinchoninic Acid Protein Assay Kit (Thermo Scientific, Rockford, IL, USA). The samples were loaded onto 10 % SDS-polyacrylamide gel electrophoresis separating gel and transferred onto polyvinylidene difluoride membranes (Roche Diagnostics, Indianapolis, IN, USA). Primary antibodies used were as follows: goat polyclonal anti-p-JAK2 (sc-21870; 1:500; Santa Cruz), rabbit polyclonal anti-JAK2 (SC-294; 1:500; Santa Cruz), rabbit monoclonal anti-STAT3 (ab109085; 1:1000; Abcam), rabbit monoclonal anti-p-STAT3 (ab76315; 1:1000; Abcam), mouse monoclonal anti-CNTFR (sc-393214; 1:500; santa cruz), rabbit polyclonal anti-GP130 (3732; 1:1000; Cell Signaling Technology, Beverly, MA, USA), rabbit monoclonal to GFAP (ab68428, 1:2000; Abcam), and rabbit monoclonal anti-GAPDH (ab181602; 1:2000; Abcam). Horseradish peroxidase-conjugated anti-mouse, anti-rabbit, and anti-goat secondary antibodies (1:1000; Santa Cruz Biotechnology) were used. Color development was achieved by using the ECL Western Blotting Detection Kit (Thermo Scientific). Quantification of protein expression level was performed using the ImageJ analysis software. Luciferase assay The CNTFR or GP130 3′ UTR containing the predicted target sequence was cloned and inserted into the pMIR-REPORT™ luciferase vector (Ambion, Austin, TX), respectively. The primers used were as follows. For CNTFR, forward primer: 5′CTAGACTAGTCACGAGGACATGCCAGAGC3′, reverse primer: 5′ CCCAAGCTTCCATTGAGTCAGACTAGAAGGGAC3′; for GP130, forward primer: 5′CTAGACTAGTCCTCACTCCCTGAAGATAGGC3′, reverse primer: 5′CCCAAGCTTGCCACCCTTCAACAAACACC3′. Three mutated vectors were generated by Life Technologies by replacing the predicted target region with its reverse sequence (mut CNTFR, from TTTGCAC to AAACGTG; mut1 GP130, from GCACTTT to CGTGAAA; mut2 GP130, from TTTGCAC to AAACGTG). HEK 293T cells were seeded onto 24-well plates for 12 h. Afterwards, 0.2 μg of firefly luciferase reporter plasmid; 0.2 μg of β-galactosidase expression vector (Ambion); 0.2 μg of expression vector pcDNA3.1-overexpressing miR-17-92 cluster; empty vector pcDNA3.1; or 10-, 20-, 50-nM miR-17, miR-20a, miR-19a, and miR-19b mimics were transfected into the cells. Cells were harvested for the luciferase assay (Promega, Madison, WI, USA) 24 h later, and the luciferase activity was normalized to the β-galactosidase activity. Traumatic brain injury and transplantation All of the animal care and experimental procedures were performed in accordance with the Laboratory Animal Care Guidelines approved by the Model Animal Research Center of Nanjing University. Traumatic brain injury was performed as previously described [3]. Female C57BL/6 mice 10–12 weeks old were deeply anesthetized and positioned in a stereotaxic frame. A burr hole was drilled at coordinates AP 1.0, L 1.8 (relative to Bregma = 0) using a dental drill, and stab wound injury was caused to the right hemisphere of the cerebral cortex by inserting a 26-gauge needle 1.0 mm deep from the brain surface (DV 1.0 relative to Bregma = 0). The needle was then retracted and reinserted five times. Immediately after injury, 1 μl of freshly dissociated NSCs (105 cells), either transduced with miR-17-92-overexpressing lentivirus or control lentivirus, was injected 0.3 mm ventrally to the injury site using a 1-μl syringe with a 26-gauge needle (SGE; Syringe perfection). Cells were injected slowly over 5 min, the syringe was left in place for an extra 5 min and then withdrawn gently, and the skin was sutured. After surgery, mice were held on a heated cushion before being returned to their home cages. Rotarod test The rotarod test described by Hamm et al. was modified for use in mice [24]. Briefly, prior to surgeries, mice were trained on the rotarod for three consecutive days (three trials per day) at an accelerating speed increasing from 4 to 40 rpm over 5 min. To test for motor coordination, three trials for fast speed (32 rpm) were performed up to 5 min with 30-min intervals between trials and the best performance value for each animal was recorded. Mice that failed to stay on the rotarod for >30 s at 32 rpm were excluded from the experiment. A trial was terminated if the animal fell off the rotarod or gripped the device and spun around past the lowest point. Post-injury testing was monitored at 1, 4, 6, and 12 weeks. Three different groups of mice were tested: (1) mice with brain injury without cell transplantation (n = 8 mice), (2) mice with brain injury that received control grafts of CON-NSC (n = 8 mice), and (3) mice with brain injury that received grafts of miR-17-92-NSC (n = 8 mice). Mice were cared for in accordance with institutional guidelines. Statistical analysis Data are presented as means ± SEM of at least three independent experiments. Direct comparisons were made using Student’s t tests, and multiple group comparisons were made using one-way analysis of variance (ANOVA) followed by Tukey’s test. Statistical significance was defined as P < 0.05, 0.01, or 0.001 (indicated as *, **, or ***, respectively). PRISM 5.0 software (GraphPad Inc., La Jolla, CA) was used for data analysis. Results Inhibition of LIF and CNTF secreted by reactive astrocyte reduced astrocytogenesis in neuroinflammatory condition To investigate the effects of neuroinflammation on embryonic-derived NSCs differentiation, we first isolated NSCs from E14.5 C57BL/6 mouse cortex. Almost all NSCs were well stained with anti-nestin and Sox2 antibody in the sphere culture (97.4 ± 1.7 and 93.8 ± 2.0 %, respectively) (Fig. 1a), while DCX staining is relatively weak (Additional file 1: Figure S1A). These cultured spheres started to differentiate after withdrawing bFGF and hEGF, in which early neural marker β-tubulin III, mature neural marker MAP2, astrocytic marker GFAP, oligodendrocyte marker CNPase and oligodendrocyte-specific protein (OSP) were observed (Fig. 1b, Additional file 1: Figure S1B, C). Next, NSCs were cultured in astrocytic-conditioned medium (hereafter refers to as CM) from lesioned astrocytes which mimic neuroinflammatory condition in vitro. Compared to control group, NSCs are subjected to CM treatment differentiated into more astrocytes and less neurons (Fig. 1c, d). Since previous study has reported that LIF and CNTF are upregulated in conditioned medium from reactive astrocytes [25], we first confirmed that LIF and CNTF stimulation activated JAK-STAT pathway and significantly increased the ratio of GFAP-positive to β-tubulin III-positive cells in the NSC population under differentiation conditions (Additional file 1: Figure S2). To investigate the role of endogenous LIF and CNTF in CM on NSCs differentiation, we employed LIF or/and CNTF-neutralizing antibodies in CM, which remarkably decreased the proportion of GFAP-positive cells, while increasing that of β-tubulin III-positive cells compared to the CM group (Fig. 1c, d). Additionally, we further evaluated the activation status of JAK2/STAT3 signaling pathway by western blot analysis. We found that CM stimulation significantly increased the levels of p-JAK2, p-STAT3, and GFAP in NSCs, while LIF or/and CNTF-neutralizing antibody treatment blocked such upregulation in those cells (Fig. 1e). These findings suggest that LIF and CNTF secreted by reactive astrocytes activate JAK2/STAT3 pathway, which further induce astrocytogenesis and inhibit neurogenesis of NSCs while inhibiting these signals can significantly reduce astrocytogenesis of NSCs in neuroinflammatory condition.Fig. 1 Effects of reactive astrocytes on NSC differentiation. a Immunofluorescence labeling of whole-mount embryonic cortical neurospheres showing the expression of NSC marker nestin (green); DAPI (blue) was used for nuclear staining. Scale bar, 100 μm. b The whole-mount neurospheres allowed to differentiate for 7 days in the absence of bFGF and hEGF were stained with the neuron marker β-tubulin III (green) and astrocyte marker GFAP (red); DAPI (blue) was used for nuclear staining. Scale bar, 200 μm. c The dissociated neurospheres were exposed to DMEM/F12/1 % N2 plus media supplement medium (CON) and reactive astrocyte-conditioned medium alone (CM), with LIF or/and CNTF neutralization antibody (CM + a-LIF, CM + a-CNTF, or CM + a-LIF + a-CNTF) to differentiate for 4 days in the absence of bFGF and hEGF, followed by fluorescence staining with β-tubulin III (green) and GFAP (red). Inset: DAPI staining of cell nuclei in the field of view. Scale bar, 100 μm. d The proportions of GFAP-positive or β-tubulin III-positive cells from the experiment shown in c were determined. Values are means ± SEM. Asterisks indicate a statistically significant difference compared with the CON or CM group (*P < 0.05, **P < 0.01, ***P < 0.001, one-way analysis of variance (ANOVA) followed by Tukey’s test, n = 3 independent experiments). e Western blot analysis and quantifications of p-JAK2, p-STAT3, and GFAP protein expressions in NSCs treated with different conditioned media indicated in c for 4 days. JAK2 and STAT3 were used as internal controls. Values are means ± SEM. Asterisks indicate a statistically significant difference compared with the CON or CM group (*P < 0.05, **P < 0.01, ***P < 0.001, one-way ANOVA followed by Tukey’s test, n = 4 independent experiments). For quantitative analysis, the numbers of each type of cell scored in four random fields were averaged, utilizing the Image-Pro Plus image analysis software miR-17-92 cluster targets multiple proteins in JAK2/STAT3 pathway mediated by LIF/CNTF In our previous study, we demonstrated that miR-17 modulates the astrocytogenesis/neurogenesis transition during the mouse corticogenesis by targeting BMP signaling pathway [19]. In addition, miR-17 and miR-20a are reported to target JAK2 and STAT3 [26, 27]. Furthermore, using TargetScan (http://Targetscan.org) and microRNA.org (http://microRNA.org) bioinformatics tools, we found that among those miRNAs in miR-17-92 cluster, miR-19a/b may target CNTFR and miR-17, miR-20a, and miR-19a/b may target glycoprotein 130 (GP130), respectively (Fig. 2a, b). Luciferase assay revealed that miR-17-92 cluster members bind to 3′UTR of CNTFR and GP130 directly (Fig. 2c, d; Additional file 1: Figure S3). When we overexpressed miR-17-92 cluster in NSCs, we found that the levels of miR-17-92 cluster members were all enhanced in various degrees (Fig. 2e), while the levels of targeted proteins including GP130, CNTFR, JAK2, and STAT3 were all declined (Fig. 2f, g). Moreover, we found similar miRNA-binding sites exited in these four genes in human database indicating these miRNA gene regulations may also exist in humans (Additional file 1: Figure S4). Based on these results, it is believed that miR-17-92 cluster can systematically regulate LIF/CNTF and downstream JAK2/STAT3 signaling by targeting multiple proteins (GP130, CNTFR, JAK2, and STAT3) in this pathway (Fig. 2h).Fig. 2 miR-17-92 cluster members target multiple proteins in JAK2/STAT3 pathway. a The schematic genomic organization of the miR-17-92 cluster on the mouse chromosome 14 (Chr.14). b Sequence alignment of mature miR-19a, miR-19b, miR-17, and miR-20a revealed their seed sequences that were reverse complementary to the seed-matched sequence within the 3′ UTR of mouse CNTFR or GP130, respectively. The mutated sequences in the 3′ UTRs of CNTFR (mut CNTFR) and GP130 (mut1 GP130, mut2 GP130) are indicated. c, d Luciferase activity was measured 24 h after transfecting HEK 293T cells. Reporter plasmids with the wild-type (wt CNTFR, wt GP130) or mutated (mut CNTFR, mut1 GP130, or mut2 GP130) 3′UTR of CNTFR or GP130 were transfected either alone (vector) or with the expression vector pcDNA3.1-overexpressing miR-17-92 cluster (miR-17-92) or empty pcDNA 3.1 (con). Values are means ± SEM. Asterisks indicate a statistically significant difference compared with the control (*P < 0.05, one-way ANOVA followed by Tukey’s test, n = 3 independent experiments). e Quantitative real-time PCR was used to determine the efficiency of overexpressing miR-17-92 cluster members using lentivirus (miR-17-92). Values are means ± SEM (n = 3 independent experiments). f Western blot analysis of GP130, CNTFR, JAK2, and STAT3 protein expressions in NSCs 72 h after infecting them with lentivirus overexpressing miR-17-92 cluster (miR-17-92) or control lentivirus (CON). g Quantifications for the protein expression levels from the experiment shown in f. Values are means ± SEM. Asterisks indicate a statistically significant difference compared with the control group (*P < 0.05, unpaired two-tailed t test, n = 3 independent experiments). h Schematic representation of multiple proteins regulated by miR-17-92 cluster in JAK2/STAT3 signaling pathway miR-17-92 cluster represses JAK2/STAT3 signaling pathway and modulates neurogenesis/astrocytogenesis transition of cultured NSCs To investigate the role of miR-17-92 cluster in modulating JAK2/STAT3 pathway mediated by LIF/CNTF, we overexpressed miR-17-92 cluster in cultured NSCs followed with LIF, CNTF, or CM treatment. Afterwards, we examined the protein levels of JAK2, STAT3, phosphorylated JAK2, phosphorylated STAT3, and GFAP (Fig. 3a). Our results demonstrated that miR-17-92 cluster significantly reduced the expression levels of JAK2 and STAT3, as well as their phosphorylation levels and their downstream GFAP protein levels (Fig. 3b). These data suggest miR-17-92 cluster overexpression in NSCs represses JAK2/STAT3 signaling pathway activated by the secreted LIF/CNTF during neuroinflammation.Fig. 3 miR-17-92 cluster modulates JAK2-STAT3 signaling pathway. a Seventy-two hours after infecting them with lentivirus overexpressing miR-17-92 cluster (Lenti-miR-17-92) or control lentivirus (Lenti-CON), NSCs were cultured in the different differential media (DMEM/F12/1 % N2 plus media supplement medium alone (CON) or with 10 ng/ml LIF, CNTF or CM) for 4 days. Expression levels of p-JAK2, JAK2, p-STAT3, STAT3, and GFAP were analyzed by western blot. GAPDH was used as an internal control. b Quantification of the protein expression levels from the experiment shown in a. Values are means ± SEM. Asterisks indicate a statistically significant difference compared with the Lenti-CON group (*P < 0.05, **P < 0.01, unpaired two-tailed t test, n = 3 independent experiments). Pound signs indicate a statistically significant difference compared with the untreated control in the same group (# P < 0.05, ## P < 0.01, ### P < 0.001, one-way ANOVA followed by Tukey’s test, n = 3 independent experiments) To evaluate the effects of miR-17-92 cluster on NSC differentiation, we overexpressed miR-17-92 cluster in NSCs before LIF, CNTF, or CM stimulation, respectively. Afterwards, we assessed the level of astrocytogenesis and neurogenesis of differentiated NSCs. As compared to the control group, overexpression of miR-17-92 cluster decreased the percentage of GFAP-positive cells in the total EGFP-positive cells (41.9 ± 2.8 vs.54.3 ± 2.7 %, p < 0.05; 45.0 ± 2.6 vs. 57.8 ± 2.6 %, p < 0.05; 37.0 ± 3.7 vs. 50.7 ± 2.4 %, p < 0.05; Fig. 4a, c), while increasing the DCX-positive cell ratio in the total EGFP-positive cells (29.5 ± 1.3 vs. 20.4 ± 2.7 %, p < 0.05; 25.6 ± 1.5 vs. 17.5 ± 2.4 %, p < 0.05; 33.9 ± 1.8 vs. 23.9 ± 2.7 %, p < 0.05; Fig. 4b, d) in differentiated NSCs upon these stimulations.Fig. 4 miR-17-92 cluster regulates astrocytogenesis and neurogenesis of NSCs in vitro. a, b Immunofluorescence labeling of NSCs showing the expression of astrocyte marker GFAP or neuron marker doublecortin (DCX) after infecting with lentivirus overexpressing miR-17-92 cluster (miR-17-92) or lentivirus expressing EGFP alone (CON) in the different differential media. Arrow heads indicate double positive cells. Scale bar, 50 μm. c, d Percentage of GFAP-positive or DCX-positive among total EGFP-positive cells from the experiment shown in a or b was determined. Values are means ± SEM. Asterisks indicate a statistically significant difference compared with the control group (*P < 0.05, unpaired two-tailed t test, n = 3 independent experiments) Taken together, these results suggest that miR-17-92 cluster members can effectively repress LIF/CNTF-mediated JAK2/STAT3 signaling pathway by targeting GP130, CNTFR, JAK2, and STAT3, which further facilitates neuronal differentiation at the expense of astrocytogenesis of NSCs in vitro. miR-17-92 cluster represses astrocytogenesis and increases neurogenesis in grafted NSCs in vivo To further investigate the effects of miR-17-92 cluster on the differentiation of grafted NSCs in vivo, we first generated a stab wound injury in the motor cortex of adult C57BL/6 mice. We examined the status of astrocytes activation by staining GFAP around the injured area. We found that 1 day after injury, there were already an amount of reactive astrocytes and the number of reactive astrocytes increased 1 week after the injury (Additional file 1: Figure S5A), indicating a progressive inflammatory condition. Four weeks after injury, the damaged area was basically healed where there were no neuron-specific nuclear protein NeuN-positive cells (Additional file 1: Figure S5B). The temporal course of transplantation, immunohistochemical test, and behavioral analysis were depicted in Fig. 5a. Immediately after injury, NSCs were grafted 0.3 mm ventrally to the injured area just above the subcortical white matter (Fig. 5b). There were no mitotic marker phospho-histone H3 (PH3)-positive signals in NSCs overexpressing miR-17-92 cluster (miR-17-92-NSC) as well as control cells (CON-NSC) 1 day (Fig. 5c) and 7 days (data not shown) after transplantation, excluding the possibility of tumor formation which also indicates that miR-17-92 cluster has little effect on NSC proliferation in the lesioned brain.Fig. 5 PH3 and nestin immunoreactivity in grafted NSCs. a Schematic presentation of temporal course of transplantation, immunohistochemical test, and behavioral analysis. b Diagram illustrating the injury site (red line) in the motor cortex (MC) and NSC grafting (green dot). c Immunofluorescence labeling of coronal sections for the mitotic marker PH3 (red), 1 day after transplantation of either control (CON-NSC) or miR-17-92-overexpressing (miR-17-92-NSC) cells (green). No positive signal was observed in the grafted cells. Scale bar, 50 μm. d, f Immunofluorescence labeling of coronal sections for the NSC marker nestin (red), 1 day and 1 week after transplantation of either CON-NSC or miR-17-92-NSC (green). Both types of grafted cells express nestin at 1 day, while no positive signal was observed at 7 days. Scale bar, 50 μm. e The ratio of nestin-positive to EGFP-positive cells from the experiment shown in d was determined. There was no statistically significant difference between the two groups (N.S. not significant, unpaired two-tailed t test, n = 3 independent experiments) To explore whether grafted NSCs retained their neural stem/precursor cell identity, the expression of NSCs marker nestin was determined. One day after transplantation, both miR-17-92-NSC and CON-NSC groups highly expressed nestin (64.9 ± 4.2 vs. 62.6 ± 3.2 %, p = 0.68), and there were no nestin-positive cells 1 week after transplantation (Fig. 5d–f). These findings reveal that miR-17-92 cluster has little effect on grafted NSCs in retaining their neural stem cell identity in the injured area. To further evaluate the potential role of miR-17-92 cluster on NSC differentiation in injured brain under inflammatory condition, we examined the percentage of GFAP or NeuN-positive cells among grafted cells 12 weeks after transplantation. As compared to the control group, miR-17-92 cluster significantly decreased the proportion of astrocytes (32.1 ± 2.9 vs. 46.8 ± 3.5 %, p < 0.05; Fig. 6a, b), while that of neurons was remarkably increased (13.4 ± 2.0 vs. 4.8 ± 0.9 %, p < 0.05; Fig. 6c, d).Fig. 6 miR-17-92 cluster regulates astrocytogenesis and neurogenesis of grafted NSCs. a, c Expression of GFAP (red) or NeuN (red) in transplanted CON-NSC or miR-17-92-NSC cells (green), 12 weeks after injury and transplantation. Arrows in c indicate double positive cells. Scale bar, 100 μm. The framed areas in c are shown at higher magnification. Scale bar, 10 μm. b, d Percentage of GFAP-positive or NeuN-positive cells among total EGFP-positive cells from the experiment shown in a or c was determined. Values are means ± SEM. Asterisks indicate a statistically significant difference compared with the control group (*P < 0.05, unpaired two-tailed t test, n = 3 independent experiments) miR-17-92 cluster overexpression reduces astrogliosis and improves the motor coordination of brain-injured mice We observed that there are large amount of reactive astrocytes gathering around the lesion site, which normally reflects the recovery situation after brain injury. Therefore, we assessed the potential effect of miR-17-92 cluster overexpression on astrogliosis 12 weeks after NSC transplantation by immunostaining of GFAP. Our results showed that miR-17-92 cluster overexpression in NSCs significantly decreased astrogliosis by 32.2 % at the lesion site (Fig. 7a, b). This is partly because of the reduced astrocytogenesis of grafted NSCs, which directly diminished the number of reactive astrocytes. Meanwhile, the result also points to a potential non-cell autonomous effect of grafted NSCs on host astrocytes that contributes to the reduction of astrogliosis.Fig. 7 miR-17-92 cluster overexpression in grafted NSCs reduces astrogliosis and improves the motor coordination of brain-injured mice. a Immunostaining for GFAP (red), 12 weeks after transplantation of CON-NSC or miR-17-92-NSC (green). The area within white dotted lines corresponds to the astroglial scar. Scale bar, 200 μm. b Quantification of astrogliosis, as measured by relative GFAP fluorescence intensity from all coronal sections containing the injury site along the rostro-caudal axis (*P < 0.05, unpaired two-tailed t test, n = 3 independent experiments). c The mice that received control NSC (CON-NSC), miR-17-92-overexpressing NSC (miR-17-92-NSC), or no cells (Only Injury) were tested 1, 4, 6, and 12 weeks after injury and transplantation. Asterisks indicate a statistically significant difference compared with the only injury group or CON-NSC group (*P < 0.05, one-way ANOVA followed by Tukey’s test, n = 8 mice) Due to the mild injury in the motor cortex, no obvious motor defects related to walking, climbing, or feeding abilities were observed among almost all the operated animals. Therefore, we checked the potential effect of miR-17-92-NSC transplantation using a rotarod test, a more sophisticated task of motor coordination. We showed that mice that received cell grafts of either type appeared to be doing better than the group that received no cells, especially the miR-17-92-NSC group. The mice that received NSC-overexpressing miR-17-92 cluster showed no significant differences in performance compared to control-NSC group until 12 weeks after transplantation (168.8 ± 13.6 vs. 129.8 ± 9.5, p < 0.05; Fig. 7c). These data indicate that miR-17-92 cluster may improve the motor coordination of brain-injured mice via increasing neurogenesis of grafted NSCs. Discussion Transplantation of NSCs into the injured CNS becomes a promising strategy to overcome the regenerative limitations of the lesioned brain [28]. However, the pathological environment in the injured brain strongly affects grafted NSC properties and their lineage selection, which results in low yield of differentiated neurons [10, 29]. It is demonstrated that reactive astrocyte during CNS injury facilitates astrocytic differentiation of NSCs in vitro [25]. Those activated astrocyte under neuroinflammatory condition can secret several cytokines such as LIF and CNTF, which have been proved to regulate NSC differentiation in vitro [30]. In the present study, we have demonstrated that astrocytic CM strongly induces astrocytogenesis from NSCs and this induction is mitigated by the addition of neutralizing antibody of LIF and CNTF. These data indicate that LIF and CNTF are the major factors produced in astrocytic CM which promote glial cell fate of NSCs. In the nervous system, the effects of LIF and CNTF are complicated which depend on the stages of development. In the gliogenic phase, there are evidence that LIF and CNTF enhance generation, maturation, and survival of oligodendrocytes [31]. It is also reported that LIF is secreted by grafted NSCs and provides protective effects in animal model of multiple sclerosis by promoting survival, differentiation, and the remyelination capacity of both endogenous oligodendrocyte precursors and mature oligodendrocytes [32, 33]. However, in the early neurogenic phase when NSCs mainly give rise to neurons, LIF and CNTF can modulate neuron-astrocyte transition by promoting astrocytic differentiation from NSCs [34]. There is a similar, post-transplantation transition of neurons into astrocytes [35]. It is believed that inhibition of LIF and CNTF signals can reduce astrocytogenesis and enhance neuronal generation from NSCs [34]. There are circumstantial evidences suggesting that JAK-STAT pathway, which plays a crucial role in astrocytogenesis, is activated upon LIF, CNTF, and IL-6 stimulation [36, 37]. Furthermore, several studies suggest that inhibition of JAK-STAT pathway can induce fate switch and facilitate neuronal differentiation [38, 39]. We have demonstrated that astrocytic CM stimulation activates JAK-STAT pathway, while blocking LIF and CNTF in astrocytic CM significantly repress such activation leading to the increase of neuronal differentiation. These results together suggest that inhibition of LIF/CNTF signal and JAK-STAT pathway in NSCs under inflammatory conditions may increase neurogenesis at the expense of astrocytogenesis. Our previous work demonstrated that miR-17 modulates neuron-astrocyte transition via inhibiting BMPR2 expression [19]. Meanwhile, Naka-Kaneda et al. also reported the role of miR-17 in controlling neurogenic to gliogenic transition of NSCs [40]. In the present work, we found that miR-17-92 cluster has multiple protein targets in LIF/CNTF signal pathway including CNTFR, GP130, JAK2, and STAT3. Among these four identified targets, JAK2 and STAT3 have been proved to be regulated by miR-17 in several models [26, 41, 42]. In addition, there are similar miRNA-binding sites in these four genes in the human database, indicating that such regulation of miR-17-92 cluster is conserved in humans. Further investigation showed that overexpression of miR-17-92 cluster significantly reduces those protein levels in NSCs leading to strong inhibition of JAK-STAT pathway under LIF/CNTF or astrocytic CM stimulation. As a result, such multi-targeted inhibition promoted neurogenesis while reducing astrocytic differentiation from NSCs subjected to LIF/CNTF or astrocytic CM treatment. These results demonstrated great potential of miR-17-92 cluster in promoting neuronal differentiation from NSCs under inflammatory condition in vitro. Next, we assessed the effect of miR-17-92 cluster in modulating NSC differentiation under inflammatory conditions in vivo. Although it is reported that miR-17-92 cluster promotes cell expansion in vitro and in vivo [43, 44], in our study, overexpression of miR-17-92 cluster showed no significant impact on the proliferation of grafted NSCs since a majority of NSCs stop dividing 1 day after transplantation. Meanwhile, overexpression of miR-17-92 cluster significantly reduced astrocytogenesis and increased neuronal differentiation after NSCs transplantation in our brain injury model. Besides, we demonstrated that overexpression of miR-17-92 cluster in NSCs also reduced astrogliosis in the lesion site after cell transplantation. There are evidences that miRNA-containing microvesicles regulate inflammation in a cell-to-cell manner [45]. Additionally, miR-17-92 cluster is suggested as a key player in control inflammation [46]. Furthermore, several evidences suggest that miRNA-containing microvesicles are involved in cell-to-cell communication in CNS under pathological conditions related to neuroinflammation [23, 47]. Together, these data suggest that grafted NSCs may secrete microvesicles containing miR-17-92 cluster, which further modulate the inflammatory response of resident astrocytes that ultimately contributes to the reduction of astrogliosis. Behavioral test of brain-injured mice showed remarkable improvement of motor coordination when miR-17-92 cluster-overexpressing NSCs were grafted, which is in agreement with our morphological findings. It is well known that increased neuronal differentiation is helpful to the functional recovery after CNS insults [3]. Besides, another potentially beneficial effect of miR-17-92 cluster-overexpressing NSCs is the significant inhibition of astrogliosis since reactive astrocyte as well as glial scar may have detrimental consequences after CNS damage [48, 49]. It is undoubted that during brain lesion, both astrocyte and microglia are activated, which result in a neuroinflammatory condition [50]. Cytokines secreted by both cell types may affect the cell lineage selection of grafted NSCs. In addition to the LIF and CNTF secreted by activated astrocyte, it is demonstrated that activated microglia also facilitates astrocytogenesis via producing IL-6 and LIF [8, 10]. Notably, all these cytokines affect NSCs differentiation through the modulation of JAK-STAT pathway [51, 52]. Furthermore, one of the targeted proteins of miR-17-92 cluster, GP130, controls IL-6, and downstream JAK-STAT pathway that regulates NSC lineage selection [31, 51]. Therefore, we believe that overexpression of miR-17-92 cluster in grafted NSCs can remarkably block the activation of JAK-STAT pathway mediated by LIF/CNTF/IL-6 under neuroinflammatory condition in vivo. Conclusions Taken together, our results suggest that LIF and CNTF produced by reactive astrocyte activate JAK-STAT pathway in NSCs. miR-17-92 cluster effectively inhibits the activation of JAK-STAT pathway under neuroinflammatory condition via targeting multiple proteins, leading to the increase of neuronal differentiation from grafted NSCs after brain injury. These findings not only provide a better understanding of NSC differentiation regulated by neuroinflammation but also potentiate the clinical application of neural cell replacement. Additional file Additional file 1: Figures S1–S5. Figure S1. Characterization of the NSCs. Figure S2. Effect of LIF and CNTF on differentiation of NSCs. Figure S3. miR-17-92 members directly target the 3′UTR of CNTFR or GP130. Figure S4. Bioinformatics analysis of miR-17-92 cluster members binding sites within CNTFR, GP130, JAK2, and STAT3. Figure S5. Traumatic brain injury. (PDF 558 kb) Abbreviations NSCsNeural stem/precursor cells LIFLeukemia inhibitory factor CNTFCiliary neurotrophic factor GP130Glycoprotein 130 CNTFRCiliary neurotrophic factor receptor JAK2Janus kinases 2 STAT3Signal transducer and activator of transcription 3 CNSCentral nervous system BMPBone morphogenetic protein IL-6Interleukin 6 3′UTR3′ Untranslated region bFGFBovine fibroblast growth factor hEGFHuman epidermal growth factor GFAPGlial fibrillary acidic protein OSPOligodendrocyte-specific protein CMConditioned medium NeuNNeuronal nuclei PH3Phospho-histone H3 DMEM/F12Dulbecco’s modified Eagle’s medium: nutrient mixture F-12 DAPI4′,6-Diamidino-2-phenylindole GAPDHGlyceraldehyde 3-phosphate dehydrogenase EGFPEnhanced green fluorescent protein Acknowledgements We thank Qipeng Zhang and Ke Zen for helpful comments. Technical assistance by Hanqin Li and Qun Chen is gratefully acknowledged. Funding This work was supported by grants from the National Natural Science Foundation of China (31471019). Availability of data and materials All raw data used in this manuscript are available on request. Authors’ contributions SM and LL designed the study. SM, XL, JW, and XD performed the experiments. CZ provided space and equipment for the project. SM wrote the draft of the manuscript. LL supervised the study and edited the manuscript. All authors read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. Consent for publication Not applicable. Ethics approval and consent to participate Our study does not include any human participants. All of the animal care and experimental procedures were performed in accordance with the Laboratory Animal Care Guidelines approved by the Model Animal Research Center of Nanjing University. ==== Refs References 1. Kim DY Hwang I Muller FL Paik JH Functional regulation of FoxO1 in neural stem cell differentiation Cell Death Differ 2015 22 2034 2045 10.1038/cdd.2015.123 26470727 2. Kokaia Z Martino G Schwartz M Lindvall O Cross-talk between neural stem cells and immune cells: the key to better brain repair? Nat Neurosci 2012 15 1078 1087 10.1038/nn.3163 22837038 3. Makri G Lavdas AA Katsimpardi L Charneau P Thomaidou D Matsas R Transplantation of embryonic neural stem/precursor cells overexpressing BM88/Cend1 enhances the generation of neuronal cells in the injured mouse cortex Stem Cells 2010 28 127 139 19911428 4. Mosher KI Andres RH Fukuhara T Bieri G Hasegawa-Moriyama M He Y Guzman R Wyss-Coray T Neural progenitor cells regulate microglia functions and activity Nat Neurosci 2012 15 1485 1487 10.1038/nn.3233 23086334 5. Lukovic D Stojkovic M Moreno-Manzano V Jendelova P Sykova E Bhattacharya SS Erceg S Concise review: reactive astrocytes and stem cells in spinal cord injury: good guys or bad guys? Stem Cells 2015 33 1036 1041 10.1002/stem.1959 25728093 6. Bertrand N Castro DS Guillemot F Proneural genes and the specification of neural cell types Nat Rev Neurosci 2002 3 517 530 10.1038/nrn874 12094208 7. Li H Grumet M BMP and LIF signaling coordinately regulate lineage restriction of radial glia in the developing forebrain Glia 2007 55 24 35 10.1002/glia.20434 17001659 8. Covacu R Brundin L Effects of neuroinflammation on neural stem cells Neuroscientist 2015 26659565 9. Liu Q Sanai N Jin WN La Cava A Van Kaer L Shi FD Neural stem cells sustain natural killer cells that dictate recovery from brain inflammation Nat Neurosci 2016 19 243 252 10.1038/nn.4211 26752157 10. Ideguchi M Shinoyama M Gomi M Hayashi H Hashimoto N Takahashi J Immune or inflammatory response by the host brain suppresses neuronal differentiation of transplanted ES cell-derived neural precursor cells J Neurosci Res 2008 86 1936 1943 10.1002/jnr.21652 18335525 11. Wang FW Hao HB Zhao SD Zhang YM Liu Q Liu HJ Liu SM Yuan QH Bing LJ Ling EA Hao AJ Roles of activated astrocyte in neural stem cell proliferation and differentiation Stem Cell Res 2011 7 41 53 10.1016/j.scr.2011.03.004 21530437 12. Kang SS Keasey MP Arnold SA Reid R Geralds J Hagg T Endogenous CNTF mediates stroke-induced adult CNS neurogenesis in mice Neurobiol Dis 2013 49 68 78 10.1016/j.nbd.2012.08.020 22960105 13. Molyneaux BJ Arlotta P Menezes JR Macklis JD Neuronal subtype specification in the cerebral cortex Nat Rev Neurosci 2007 8 427 437 10.1038/nrn2151 17514196 14. Seo H Lee K Epac2 contributes to PACAP-induced astrocytic differentiation through calcium ion influx in neural precursor cells BMB Rep 2016 49 128 133 10.5483/BMBRep.2016.49.2.202 26645637 15. He F Ge W Martinowich K Becker-Catania S Coskun V Zhu W Wu H Castro D Guillemot F Fan G A positive autoregulatory loop of Jak-STAT signaling controls the onset of astrogliogenesis Nat Neurosci 2005 8 616 625 10.1038/nn1440 15852015 16. Le MTN Xie HM Zhou BY Chia PH Rizk P Um M Udolph G Yang H Lim B Lodish HF MicroRNA-125b promotes neuronal differentiation in human cells by repressing multiple targets Mol Cell Biol 2009 29 5290 5305 10.1128/MCB.01694-08 19635812 17. Fan Z Lu M Qiao C Zhou Y Ding JH Hu G MicroRNA-7 enhances subventricular zone neurogenesis by inhibiting NLRP3/caspase-1 axis in adult neural stem cells Mol Neurobiol 2015 18. Zhang JF Shi LL Zhang L Zhao ZH Liang F Xu X Zhao LY Yang PB Zhang JS Tian YF MicroRNA-25 negatively regulates cerebral ischemia/reperfusion injury-induced cell apoptosis through Fas/FasL pathway J Mol Neurosci 2016 58 507 516 10.1007/s12031-016-0712-0 26768135 19. Mao S Li H Sun Q Zen K Zhang CY Li L miR-17 regulates the proliferation and differentiation of the neural precursor cells during mouse corticogenesis FEBS J 2014 281 1144 1158 10.1111/febs.12680 24314167 20. Ma S Liu M Xu Z Li Y Guo H Ge Y Liu Y Zheng D Shi J A double feedback loop mediated by microRNA-23a/27a/24-2 regulates M1 versus M2 macrophage polarization and thus regulates cancer progression Oncotarget 2016 7 13502 13519 26540574 21. Sun Q Mao S Li H Zen K Zhang CY Li L Role of miR-17 family in the negative feedback loop of bone morphogenetic protein signaling in neuron PLoS One 2013 8 e83067 10.1371/journal.pone.0083067 24349434 22. Wu W Takanashi M Borjigin N Ohno SI Fujita K Hoshino S Osaka Y Tsuchida A Kuroda M MicroRNA-18a modulates STAT3 activity through negative regulation of PIAS3 during gastric adenocarcinogenesis Br J Cancer 2013 108 653 661 10.1038/bjc.2012.587 23322197 23. Mao S Sun Q Xiao H Zhang C Li L Secreted miR-34a in astrocytic shedding vesicles enhanced the vulnerability of dopaminergic neurons to neurotoxins by targeting Bcl-2 Protein Cell 2015 6 529 540 10.1007/s13238-015-0168-y 26091620 24. Hamm RJ Pike BR O’Dell DM Lyeth BG Jenkins LW The rotarod test: an evaluation of its effectiveness in assessing motor deficits following traumatic brain injury J Neurotrauma 1994 11 187 196 10.1089/neu.1994.11.187 7932797 25. Faijerson J Tinsley RB Aprico K Thorsell A Nodin C Nilsson M Blomstrand F Eriksson PS Reactive astrogliosis induces astrocytic differentiation of adult neural stem/progenitor cells in vitro J Neurosci Res 2006 84 1415 1424 10.1002/jnr.21044 16998910 26. Li YM Vecchiarelli-Federico LM Li YJ Egan SE Spaner D Hough MR Ben-David Y The miR-17-92 cluster expands multipotent hematopoietic progenitors whereas imbalanced expression of its individual oncogenic miRNAs promotes leukemia in mice Blood 2012 119 4486 4498 10.1182/blood-2011-09-378687 22451425 27. Zhang MM Liu QF Mi SP Liang X Zhang ZQ Su XM Liu JY Chen YY Wang MM Zhang YA Both miR-17-5p and miR-20a alleviate suppressive potential of myeloid-derived suppressor cells by modulating STAT3 expression J Immunol 2011 186 4716 4724 10.4049/jimmunol.1002989 21383238 28. Chau MJ Deveau TC Song M Gu X Chen D Wei L iPSC transplantation increases regeneration and functional recovery after ischemic stroke in neonatal rats Stem Cells 2014 32 3075 3087 10.1002/stem.1802 25132189 29. Lozano D Gonzales-Portillo GS Acosta S de la Pena I Tajiri N Kaneko Y Borlongan CV Neuroinflammatory responses to traumatic brain injury: etiology, clinical consequences, and therapeutic opportunities Neuropsychiatr Dis Treat 2015 11 97 106 25657582 30. Bauer S Kerr BJ Patterson PH The neuropoietic cytokine family in development, plasticity, disease and injury Nat Rev Neurosci 2007 8 221 232 10.1038/nrn2054 17311007 31. Mayer M Bhakoo K Noble M Ciliary neurotrophic factor and leukemia inhibitory factor promote the generation, maturation and survival of oligodendrocytes in vitro Development 1994 120 143 153 8119123 32. Laterza C Merlini A De Feo D Ruffini F Menon R Onorati M Fredrickx E Muzio L Lombardo A Comi G iPSC-derived neural precursors exert a neuroprotective role in immune-mediated demyelination via the secretion of LIF Nat Commun 2013 4 2597 10.1038/ncomms3597 24169527 33. Butzkueven H Emery B Cipriani T Marriott MP Kilpatrick TJ Endogenous leukemia inhibitory factor production limits autoimmune demyelination and oligodendrocyte loss Glia 2006 53 696 703 10.1002/glia.20321 16498619 34. Hirabayashi Y Gotoh Y Stage-dependent fate determination of neural precursor cells in mouse forebrain Neurosci Res 2005 51 331 336 10.1016/j.neures.2005.01.004 15740796 35. Laywell ED Kearns SM Zheng T Chen KA Deng J Chen HX Roper SN Steindler DA Neuron-to-astrocyte transition: phenotypic fluidity and the formation of hybrid asterons in differentiating neurospheres J Comp Neurol 2005 493 321 333 10.1002/cne.20722 16261530 36. Jhaveri K Teplinsky E Silvera D Valeta-Magara A Arju R Giashuddin S Sarfraz Y Alexander M Darvishian F Levine PH Hyperactivated mTOR and JAK2/STAT3 pathways: molecular drivers and potential therapeutic targets of inflammatory and invasive ductal breast cancers after neoadjuvant chemotherapy Clin Breast Cancer 2016 16 113 122 10.1016/j.clbc.2015.11.006 26774497 37. Lee J Son MJ Woolard K Donin NM Li A Cheng CH Kotliarova S Kotliarov Y Walling J Ahn S Epigenetic-mediated dysfunction of the bone morphogenetic protein pathway inhibits differentiation of glioblastoma-initiating cells Cancer Cell 2008 13 69 80 10.1016/j.ccr.2007.12.005 18167341 38. DeVito WJ Stone S Ethanol inhibits prolactin-induced activation of the JAK/STAT pathway in cultured astrocytes J Cell Biochem 1999 74 278 291 10.1002/(SICI)1097-4644(19990801)74:2<278::AID-JCB12>3.0.CO;2-P 10404396 39. Zhang YL Zhou Z Han WW Zhang LL Song WS Huang JH Liu S Oleanolic acid inhibiting the differentiation of neural stem cells into astrocyte by down-regulating JAK/STAT signaling pathway Am J Chin Med 2016 44 103 117 10.1142/S0192415X16500075 26916917 40. Naka-Kaneda H Nakamura S Igarashi M Aoi H Kanki H Tsuyama J Tsutsumi S Aburatani H Shimazaki T Okano H The miR-17/106-p38 axis is a key regulator of the neurogenic-to-gliogenic transition in developing neural stem/progenitor cells Proc Natl Acad Sci U S A 2014 111 1604 1609 10.1073/pnas.1315567111 24474786 41. Carraro G El-Hashash A Guidolin D Tiozzo C Turcatel G Young BM De Langhe SP Bellusci S Shi W Parnigotto PP Warburton D miR-17 family of microRNAs controls FGF10-mediated embryonic lung epithelial branching morphogenesis through MAPK14 and STAT3 regulation of E-cadherin distribution Dev Biol 2009 333 238 250 10.1016/j.ydbio.2009.06.020 19559694 42. Foshay KM Gallicano GI miR-17 family miRNAs are expressed during early mammalian development and regulate stem cell differentiation Dev Biol 2009 326 431 443 10.1016/j.ydbio.2008.11.016 19073166 43. Li J Chen L Qiuqin T Wu W Hao G Lou L Jie W Hua J Hongjuan D Xia Y The role, mechanism and potentially novel biomarker of microRNA-17-92 cluster in macrosomia Sci Rep 2015 5 17212 10.1038/srep17212 26598317 44. Zhou Y Zhang L Ji H Lu X Xia J Li L Chen F Bu H Shi Y MiR-17~92 ablation impairs liver regeneration in an estrogen-dependent manner J Cell Mol Med 2016 20 939 948 10.1111/jcmm.12782 26781774 45. Hulsmans M Holvoet P MicroRNA-containing microvesicles regulating inflammation in association with atherosclerotic disease Cardiovasc Res 2013 100 7 18 10.1093/cvr/cvt161 23774505 46. Philippe L Alsaleh G Bahram S Pfeffer S Georgel P The miR-17 approximately 92 cluster: a key player in the control of inflammation during rheumatoid arthritis Front Immunol 2013 4 70 10.3389/fimmu.2013.00070 23516027 47. Brites D Fernandes A Neuroinflammation and depression: microglia activation, extracellular microvesicles and microRNA dysregulation Front Cell Neurosci 2015 9 476 10.3389/fncel.2015.00476 26733805 48. Fan X Luo G Yang D Ming M Liu H Pu P Le W Critical role of lysosome and its associated protein cathepsin D in manganese-induced toxicity in cultured midbrain astrocyte Neurochem Int 2010 56 291 300 10.1016/j.neuint.2009.11.001 19913584 49. Zhang X Li L Zhang X Xie W Li L Yang D Heng X Du Y Doody RS Le W Prenatal hypoxia may aggravate the cognitive impairment and Alzheimer’s disease neuropathology in APPSwe/PS1A246E transgenic mice Neurobiol Aging 2013 34 663 678 10.1016/j.neurobiolaging.2012.06.012 22795785 50. Pekny M Wilhelmsson U Pekna M The dual role of astrocyte activation and reactive gliosis Neurosci Lett 2014 565 30 38 10.1016/j.neulet.2013.12.071 24406153 51. Coorey NJ Shen W Zhu L Gillies MC Differential expression of IL-6/gp130 cytokines, Jak-STAT signaling and neuroprotection after Muller cell ablation in a transgenic mouse model Invest Ophthalmol Vis Sci 2015 56 2151 2161 10.1167/iovs.14-15695 25722214 52. Nicola NA Babon JJ Leukemia inhibitory factor (LIF) Cytokine Growth Factor Rev 2015 26 533 544 10.1016/j.cytogfr.2015.07.001 26187859

PMC005xxxxxx/PMC5002216.txt

==== Front BMC Infect DisBMC Infect. DisBMC Infectious Diseases1471-2334BioMed Central London 177510.1186/s12879-016-1775-9Research ArticleEstimated burden of group a streptococcal pharyngitis among children in Beijing, China Wu Shuangsheng wushuangsheng0606@126.com 12Peng Xiaomin xminp@tom.com 12Yang Zuyao zyang@cuhk.edu.hk 3Ma Chunna ma_chun_na@126.com 12Zhang Daitao zdt016@163.com 12Wang Quanyi bjcdcxm@126.com 12Yang Peng +86 10 6440 7108yangpengcdc@163.com 121 Institute for Infectious Disease and Endemic Disease Control, Beijing Center for Disease Prevention and Control, No. 16 Hepingli Middle Street, Dongcheng District Beijing, 100013 China 2 School of Public Health, Captial Medical University, Beijing, China 3 Division of Epidemiology, The Jockey Club School of Public Health and Primary Care, The Chinese University of Hong Kong, Hong Kong, SAR China 26 8 2016 26 8 2016 2016 16 1 45225 12 2015 11 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Burden of Group A streptococcus (GAS) pharyngitis is scarce in developing countries, still unknown in China. The objective of this study was to determine the incidence of clinical cases of pharyngitis and GAS culture-positive pharyngitis, and their outpatient visits among children aged 0–14 years in Beijing, the capital of China. Methods Multiplier model was used to estimate the numbers of pharyngitis cases, based on reported numbers of clinical cases and GAS culture-positive rates from GAS surveillances in Beijing, consultation rate, population coverage of GAS surveillances, sampling success rate, and test sensitivity of GAS culture from previous studies, surveys and surveillances. Results An average of 29804.6 (95 % CI: 28333.2–31276.0) clinical cases of pharyngitis per 100,000 person-years occurred among children aged 0–14 years, resulting in correspondingly 19519.0 (95 % CI: 18516.7–20521.2) outpatient visits per 100,000 person-years from 2012 to 2014 in Beijing. On average, there were 2685.1 (95 % CI: 2039.6–3330.6) GAS culture-positive cases of pharyngitis and 1652.7 (95 % CI: 1256.5–2049.0) outpatient visits per 100,000 person-years during the same period. The estimated burden of GAS pharyngitis was significantly higher than that of scarlet fever. Children aged 5–14 years had a higher burden of GAS pharyngitis than those aged 0–4 years. Conclusions The present data suggests that GAS pharyngitis is very common in children in China. Further studies and surveillances are needed to monitor trends and the effectiveness of control measures. Keywords BurdenGroup A streptococcusPharyngitisChinaBeijing Health System High Level Health Technology Talent Cultivation Plan2013-3-098Yang Peng Beijing Talents Fund2014000021223ZK36Yang Peng Beijing Science and Technology Planning ProjectD141100003114001Wang Quanyi issue-copyright-statement© The Author(s) 2016 ==== Body Background Pharyngitis is one of the most common presenting symptom for seeking medications, accounting for an estimated 15 million outpatient visits in 2006 in United States [1]. Group A streptococcus (GAS) is the main bacterial cause of pharyngitis, and responsible for a large number of pharyngitis cases in children [2]. Although GAS pharyngitis may seem relatively benign and unimportant, it causes enormous use of health resources and economic costs [3]. Furthermore, the primary infection can lead to severe GAS diseases (eg, acute rheumatic fever, rheumatic heart disease, post-streptococcal glomerulonephritis, and invasive infection) [4, 5]. A study reviewed recent population-based data and estimated that there were 616 million new GAS pharyngitis cases per year in the world, with an estimated number of more than 1.78 million new cases and 517,000 deaths of severe GAS disease each year [5]. It occurs most commonly among children aged 5–15 years, and its incidence in children varies from region to region [6]. However, epidemiological data on GAS pharyngitis from developing countries is scarce [4–6]. To our knowledge, the disease burden is still unknown in China. In this study, we aimed to determine the incidence of clinical cases of pharyngitis and GAS culture-positive pharyngitis, and their outpatient visits among children aged 0–14 years in Beijing, the capital of China. Methods GAS surveillances in Beijing GAS surveillances were conducted in the pediatric clinics of 36 hospitals within Beijing’s 18 districts since May 2011. Since November 2014, the number of sentinel hospitals has been cut down from 36 to 17. The surveillance system was designed and managed by Beijing Center for Disease Prevention and Control. Under the system, clinicians were required to diagnose scarlet fever or pharyngitis and to record the weekly numbers of outpatient visits by age groups (0–4 years, and 5–14 years) on a fixed form. All children with scarlet fever diagnosed by clinicians were invited to participate in the study. Meantime, each week, 10 children with pharyngitis diagnosed by clinicians were randomly selected from each hospital. If there are less than 10 patients in a hospital in a week, all of them were invited to participate in the study. The pharyngeal swab samples were collected by trained clinicians from study participants after their guardians gave informed consent, and tested by the collaborating laboratories. A detailed description of the surveillance was published in a previous study [7]. National notifiable infectious disease surveillance system in China Scarlet fever is a notifiable disease according to Law of the People Republic of China on the Prevention and Treatment of Infectious Diseases. All clinical cases of scarlet fever should be reported to the National Notifiable Infectious Disease Surveillance System (NNIDSS) by clinicians and hospitals when they sought medical services in China. Case definition In this study, clinical cases of pharyngitis was defined as follows: pharyngitis is the inflammation of the back of the throat including the tonsils, and its common symptoms include fever, sore throat, red tonsils, and enlarged lymph nodes in the neck and so on. Case definition of clinical cases of scarlet fever conformed to the Diagnostic Criteria for Scarlet Fever (WS282–2008) enacted by the Chinese Ministry of Health [8]. Patients with scarlet fever or pharyngitis from whom GAS was isolated were identified as GAS culture-positive patients. Model for cases numbers estimation Previous studies had used the multiplier model to estimate the burden of pandemic (H1N1) 2009 in China and USA [9, 10], and hand, foot and mouth disease in China [11]. In this study, the multiplier model was used to estimate the burden of GAS pharyngitis among children aged 0–14 years in Beijing. To estimate the total number of cases of GAS culture-positive pharyngitis, we built the multiplier model to adjust the count of laboratory-confirmed cases from GAS surveillances in Beijing for the following steps: medical care seeking (B to A), population coverage of GAS surveillances in Beijing (C to B), collection of samples (D to C), and laboratory detection of GAS (E to D) (see Fig. 1). At each step, data for parameter estimations was identified as a range of proportions observed in previous studies, surveys and surveillances.Fig. 1 Model parameters for estimating the burden of GAS pharyngitis in Beijing, 2012–2014. Note: A, Total no. cases of GAS culture-positive pharyngitis in Beijing; B, Cases refer to hospitals in Beijing = A* Consultation rate; C, Cases refer to sentinel hospitals of GAS surveillances in Beijing = B*Population coverage of GAS surveillances in Beijing; D, Cases for which specimens were successfully collected = C*Sampling success rate; E, Predicted positive specimens that were correctly identified by laboratory tests = D*Test sensitivity Because NNIDSS covered all hospitals and all population in Beijing, population coverage of GAS surveillances among scarlet fever cases was calculated by dividing reported number of clinical cases of scarlet fever from GAS surveillances by reported number from NNIDSS in Beijing. In this study, we assumed that population coverage among cases of scarlet fever was equal to that among cases of pharyngitis. The number of cases of GAS culture-positive pharyngitis was calculated using: (1) Na was age-specific reported number of clinical cases from GAS surveillances. (2) Ra was age-specific GAS culture-positive rate by GAS surveillance in Beijing. (3) Ca was age-specific consultation rate. Consultation rate was calculated by the function (1-proportion of patient who do not visit any health institutions), based on literature data from the Fourth National Health Services Survey of China in 2008, which showed that among children aged 0–4 and 5–14 years, 27.3 and 40.3 % of patients do not visit any health institutions when they were ill [12]. Accordingly, consultation rate among children aged 0–4 and 5–14 years was 72.7 and 59.7 % respectively. (4) Pa, age-specific population coverage of GAS surveillances. (5) s was the success rate for sampling swab samples, ranged from 80 to 90 %, and it was obtained from a study from Beijing [9]. (6) t was the test sensitivity of throat culture, ranged from 75.2 to 97.1 %, and it was obtained from a review study of 13 published literatures [13] (see Eq. 1). 1 TotalnumberofcasesofGASculture‐positivepharyngitis=∑Na*RaCa*Pa*s*t The number of clinical cases of pharyngitis was calculated using: (1) Na, (2) Ca, and (3) Pa (see Eq. 2). 2 Totalnumberofclinicalcasesofpharyngitis=∑NaCa*Pa The number of outpatient visits for GAS culture-positive pharyngitis was calculated using: (1) Na, (2) Ra, (3) Pa, (4) s, and (5) t (see Eq. 3). 3 TotalnumberofoutpatientvisitsforGASculture‐positivepharyngitis=∑Na*RaPa*s*t The number of outpatient visits for clinical cases of pharyngitis was a calculation of reported number of clinical cases from GAS surveillances divided by population coverage of GAS surveillances (see Eq. 4). 4 Totalnumberofoutpatientvisitsforclinicalcasesofpharyngitis=∑NaPa The same approach was used to estimate the total number of cases of GAS culture-positive scarlet fever, the total number of clinical cases of scarlet fever, the total number of outpatient visits for GAS culture-positive scarlet fever, and the total number of outpatient visits for clinical cases of scarlet fever. Annual incidence rate was equal to the estimated number of cases divided by population number from the 2010 National Population Census in Beijing, China [14]. Data from the census showed that there were 1,687,437 persons in the age group of 0–14 in Beijing, of which those aged 0–4 years accounted for 40.65 %. Data analysis Ninety-five percent confidence intervals (CIs) for population coverage and GAS culture-positive rate of GAS surveillances by age groups were calculated using the normal approximation. Difference among the subgroups were tested by Pearson’s Chi-square test with a two-sided p value <0.05 as the statistical significance level. Data analyses were carried out using SPSS Version 13.0 (SPSS Inc, Chicago, IL). Ninety-five percent CIs for estimated numbers of cases and annual incidence rate were determined by Monte Carlo simulation, using a multiplier model (Impact 2009, version 1.0) by United States Centers for Disease Control and Prevention. Results Data from NNIDSS and GAS surveillances in Beijing The reported numbers of clinical cases of scarlet fever and pharyngitis from GAS surveillances and NNIDSS by years and age groups are shown in Table 1. From 2012 to 2014, a total of 9078 clinical cases of scarlet fever aged 0–14 years were reported from NNIDSS in Beijing, 26.1 % of whom were children aged 0–4 years, and 73.9 % were between the age of 5 and 14 years.Table 1 Reported numbers of clinical cases of scarlet fever and pharyngitis from GAS surveillances and NNIDSS by age groups, Beijing, 2012–2014 Year Age group Reported number of clinical cases of scarlet fever from GAS surveillances Reported number of clinical cases of pharyngitis from GAS surveillances Reported number of clinical cases of scarlet fever from NNIDSS 2012 0–4 573 79280 998 5–14 1226 70872 2154 Overall(0–14) 1799 150152 3152 2013 0–4 286 80595 526 5–14 712 74356 1503 Overall(0–14) 998 154951 2029 2014 0–4 299 71132 842 5–14 997 70997 3055 Overall(0–14) 1296 142129 3897 2012–2014 0–4 1158 231007 2366 5–14 2935 216225 6712 Overall(0–14) 4093 447232 9078 During the same period, a total of 4093 clinical cases of scarlet fever and 447,232 ones of pharyngitis were reported from GAS surveillances in Beijing. Of the 4093 clinical cases of scarlet fever, 28.3 % were children aged 0–4 years and 71.7 % were between the age of 5 and 14 years. Of the 447,232 clinical cases of pharyngitis, 51.7 % were children aged 0–4 years and 48.3 % were between the age of five and fourteen years. Population coverage of GAS surveillances in Beijing Using the surveillance data from NNIDSS and GAS surveillances, we estimated the population coverage of GAS surveillances among children aged 0–4 years and 5–14 years in Beijing, respectively. Because the number of sentinel hospitals was cut down from 36 to 17, the population coverage of GAS surveillances in 2014 was significantly lower than that in 2012 and 2013 (see Table 2).Table 2 Estimates of population coverage of GAS surveillances by age groups, Beijing, 2012–2014 Year Age group Reported number of clinical cases of scarlet fever from GAS surveillances Reported number of clinical cases of scarlet fever from NNIDSS Population coverage of GAS surveillances Proportion (%) 95 % CI 2012 0–4 573 998 57.4 54.3 60.5 5–14 1226 2154 56.9 54.8 59.0 Overall(0–14) 1799 3152 57.1 55.3 58.8 2013 0–4 286 526 54.4 50.1 58.6 5–14 712 1503 47.4 44.8 49.9 Overall(0–14) 998 2029 49.2 47.0 51.4 2014 0–4 299 842 35.5 32.3 38.7 5–14 997 3055 32.6 31.0 34.3 Overall(0–14) 1296 3897 33.3 31.8 34.7 GAS culture-positive rate of GAS surveillances in Beijing Bacterial surveillance data for GAS showed that the overall positive rate of GAS causing scarlet fever was significantly higher than pharyngits during the 3 years (p < 0.05). The lower rate of GAS causing pharyngits in 2013 compared to 2012 and 2014 was statistically significant (p < 0.05), but no significant difference of positive rate of GAS causing scarlet fever was found between the 3 years (p > 0.05). With respect to the difference between the two age groups, we observed higher positive rate of GAS causing pharyngits among children aged 5–14 years in all the 3 years (p < 0.05). Nonetheless, no significant association between age and positive rate of GAS causing scarlet fever existed in any of 3 years (p > 0.05) (see Table 3).Table 3 GAS culture-positive rate of GAS surveillances in Beijing by age groups, Beijing, 2012–2014 Year Clinical manifestations Age group Samples no. GAS culture-positive rate Proportion (%) 95 % CI 2012 Scarlet fever 0–4 71 45.1 33.5 56.6 5–14 154 53.2 45.4 61.1 Overall(0–14) 225 50.7 44.1 57.2 Pharyngitis 0–4 1012 3.0 1.9 4.0 5–14 906 14.9 12.6 17.2 Overall(0–14) 1918 8.6 7.3 9.9 2013 Scarlet fever 0–4 50 36.0 22.7 49.3 5–14 110 39.1 30.0 48.2 Overall(0–14) 160 38.1 30.6 45.7 Pharyngitis 0–4 1382 1.0 0.5 1.5 5–14 1295 4.6 3.4 5.7 Overall(0–14) 2677 2.7 2.1 3.3 2014 Scarlet fever 0–4 50 42.0 28.3 55.7 5–14 216 43.5 36.9 50.1 Overall(0–14) 266 43.2 37.3 49.2 Pharyngitis 0–4 1479 2.3 1.5 3.1 5–14 1472 11.4 9.8 13.0 Overall(0–14) 2951 6.8 5.9 7.8 Numbers of clinical cases and GAS culture-positive cases Using the multiplier model, we estimated that there were 398,720 clinical cases of pharyngitis in 2012, 467,670 cases in 2013 and 642,410 cases in 2014, with an annual incidence rate of 23628.7, 27714.8 and 38070.2 cases per 100,000 children, respectively. Of these clinical cases, a total of 50,499, 19,258, and 66,172 pharyngitis cases were GAS culture-positive during 2012 to 2014, with an annual incidence rate of 2992.6, 1141.3 and 3921.5 cases per 100,000 children, respectively. The incidence of GAS culture-positive pharyngitis was statistically significantly lower in 2013 than in 2012 and 2014 (p < 0.05), but no significant difference was found between 2012 and 2014 (p > 0.05). With respect to the difference between the two age groups, we observed a higher risk for incidence of GAS culture-positive pharyngitis among children aged 5–14 years in all the 3 years (4273.5 vs. 1122.2 cases per 100,000 children in 2012, 1636.4 vs. 418.2 cases per 100,000 children in 2013, and 5728.5 vs. 1282.8 cases per 100,000 children in 2014, p < 0.05). The annual incidence rate of GAS culture-positive pharyngitis was more than 10 times higher than that of scarlet fever during the 3 years (2992.6 vs. 206.9 cases per 100,000 children in 2012, 1141.3 vs. 101.7 cases per 100,000 children in 2013, and 3921.5 vs. 220.6 cases per 100,000 children in 2014, p < 0.05) (see Table 4).Table 4 Estimated numbers of clinical cases and GAS culture-positive cases, Beijing, 2012–2014 Year Clinical manifestations Age groups Clinical cases GAS culture-positive cases n 95 % CI Annual incidence rate 95 % CI n 95 % CI Annual incidence rate 95 % CI 2012 Scarlet fever 0–4 1374.5 1289.8 1459.2 200.4 188.0 212.8 848.8 558.4 1139.3 123.8 81.4 166.1 5–14 3607.6 3452.8 3762.4 360.2 344.7 375.7 2642.1 2003.3 3280.9 263.8 200.0 327.6 Overall(0–14) 4982.1 4806.0 5158.2 295.2 284.8 305.7 3490.9 2712.2 4269.6 206.9 160.7 253.0 Pharyngitis 0–4 190170.0 178448.6 201891.4 27726.4 26017.5 29435.4 7696.7 4258.0 11135.4 1122.2 620.8 1623.5 5–14 208550.0 199602.5 217497.5 20822.6 19929.2 21715.9 42802.0 32187.9 53416.1 4273.5 3213.8 5333.3 Overall(0–14) 398720.0 384004.8 413435.2 23628.7 22756.7 24500.8 50499.0 38576.9 62421.1 2992.6 2286.1 3699.2 2013 Scarlet fever 0–4 725.1 660.7 789.5 105.7 96.3 115.1 360.6 199.1 522.1 52.6 29.0 76.1 5–14 2521.7 2368.5 2674.9 251.8 236.5 267.1 1355.8 920.0 1791.6 135.4 91.9 178.9 Overall(0–14) 3246.7 3080.6 3412.8 192.4 182.6 202.2 1716.4 1213.6 2219.2 101.7 71.9 131.5 Pharyngitis 0–4 204320.0 186174.5 222465.5 29789.5 27143.9 32435.0 2868.5 1098.3 4638.7 418.2 160.1 676.3 5–14 263340.0 247341.8 279338.2 26293.1 24695.7 27890.4 16389.0 10957.0 21821.0 1636.4 1094.0 2178.7 Overall(0–14) 467670.0 443493.5 491846.5 27714.8 26282.1 29147.6 19258.0 13397.0 25119.0 1141.3 793.9 1488.6 2014 Scarlet fever 0–4 1166.2 1043.0 1289.4 170.0 152.1 188.0 678.0 396.4 959.7 98.9 57.8 139.9 5–14 5125.4 4829.3 5421.5 511.7 482.2 541.3 3044.8 2316.5 3773.1 304.0 231.3 376.7 Overall(0–14) 6291.5 5974.1 6608.9 372.8 354.0 391.7 3722.8 2881.0 4564.6 220.6 170.7 270.5 Pharyngitis 0–4 277430.0 248129.5 306730.5 40448.8 36176.8 44720.7 8798.4 5111.4 12485.4 1282.8 745.2 1820.4 5–14 364980.0 343892.7 386067.3 36441.3 34335.8 38546.7 57374.0 43707.7 71040.3 5728.5 4364.0 7093.0 Overall(0–14) 642410.0 606815.7 678004.3 38070.2 35960.8 40179.5 66172.0 51279.1 81064.9 3921.5 3038.9 4804.0 Average: 2012−2014 Scarlet fever 0−4 1088.6 997.8 1179.3 158.7 145.5 171.9 629.2 384.6 873.7 91.7 56.1 127.4 5−14 3751.6 3550.2 3952.9 374.6 354.5 394.7 2347.6 1746.6 2948.5 234.4 174.4 294.4 Overall(0–14) 4840.1 4620.2 5060.0 286.8 273.8 299.9 2976.7 2268.9 3684.5 176.4 134.5 218.3 Pharyngitis 0−4 223973.3 204250.9 243695.8 32654.9 29779.4 35530.4 6454.5 3489.2 9419.8 941.1 508.7 1373.4 5−14 278956.7 263612.3 294301.0 27852.3 26320.2 29384.4 38855.0 28950.9 48759.1 3879.5 2890.6 4868.3 Overall(0–14) 502933.3 478104.7 527762.0 29804.6 28333.2 31276.0 45309.7 34417.7 56201.7 2685.1 2039.6 3330.6 Outpatient visits for clinical cases and GAS culture-positive cases As shown in Table 5, we estimated that there were 262,760, 305,760, and 419,590 outpatient visits for clinical cases of pharyngitis during 2012 to 2014, with an annual incidence rate of 15571.5, 18119.8 and 24865.5 cases per 100,000 children, respectively. Of these outpatient visits for clinical cases, a total of 31,148, 11,870 and 40,648 cases of pharyngitis were GAS culture-positive during the same period, with an annual incidence rate of 1845.9, 703.4, and 2408.9 cases per 100,000 children, respectively. The incidence of outpatient visits for GAS culture-positive pharyngitis was statistically significantly lower in 2013 than in 2012 and 2014 (p < 0.05), but no significant difference was found between 2012 and 2014 (p > 0.05). Compared to children aged 0–4 years, those aged 5–14 years had a higher risk of outpatient visits for GAS culture-positive pharyngitis in all the 3 years (2551.3 vs. 815.8 cases per 100,000 children in 2012, 976.9 vs. 304 cases per 100,000 children in 2013, and 3419.9 vs. 932.6 cases per 100,000 children in 2014, p < 0.05). The incidence of outpatient visits for GAS culture-positive pharyngitis was significantly higher than that of scarlet fever during the 3 years (1845.9 vs. 130 cases per 100,000 children in 2012, 703.4 vs. 231.6 cases per 100,000 children in 2013, and 2408.9 vs. 136.9 cases per 100,000 children in 2014, p < 0.05) (see Table 5).Table 5 Estimates of outpatient visits for clinical cases and GAS culture-positive cases, Beijing, 2012−2014 Year Clinical manifestations Age groups Clinical cases GAS culture−positive cases n 95 % CI Annual incidence rate 95 % CI n 95 % CI Annual incidence rate 95 % CI 2012 Scarlet fever 0−4 999.3 937.7 1060.8 145.7 136.7 154.7 617.1 405.9 828.3 90.0 59.2 120.8 5−14 2153.7 2061.3 2246.1 215.0 205.8 224.3 1577.3 1195.9 1958.7 157.5 119.4 195.6 Overall(0–14) 3153.0 3042.2 3263.8 186.9 180.3 193.4 2194.4 1705.0 2683.8 130.0 101.0 159.0 Pharyngitis 0−4 138260.0 129738.6 146781.4 20158.0 18915.6 21400.5 5595.5 3095.6 8095.4 815.8 451.3 1180.3 5−14 124500.0 119158.3 129841.7 12430.6 11897.3 12964.0 25553.0 19216.4 31889.6 2551.3 1918.7 3184.0 Overall(0–14) 262760.0 252722.4 272797.6 15571.5 14976.7 16166.4 31148.0 23795.2 38500.8 1845.9 1410.1 2281.6 2013 Scarlet fever 0−4 527.1 480.3 573.9 76.9 70.0 83.7 262.2 144.8 379.5 38.2 21.1 55.3 5−14 1505.4 1413.9 1596.9 150.3 141.2 159.4 809.4 549.2 1069.6 80.8 54.8 106.8 Overall(0–14) 2032.6 1929.9 2135.3 120.5 114.4 126.5 1071.6 759.5 1383.7 63.5 45.0 82.0 Pharyngitis 0−4 148540.0 135348.2 161731.8 21656.8 19733.5 23580.2 2085.4 798.4 3372.4 304.0 116.4 491.7 5−14 157220.0 147669.1 166770.9 15697.6 14744.0 16651.2 9784.3 6541.4 13027.2 976.9 653.1 1300.7 Overall(0–14) 305760.0 289483.0 322037.0 18119.8 17155.2 19084.4 11870.0 8276.2 15463.8 703.4 490.5 916.4 2014 Scarlet fever 0−4 847.8 758.3 937.3 123.6 110.6 136.7 492.9 288.2 697.7 71.9 42.0 101.7 5−14 3059.8 2883.0 3236.6 305.5 287.9 323.2 1817.7 1382.9 2252.5 181.5 138.1 224.9 Overall(0–14) 3907.6 3711.8 4103.4 231.6 220.0 243.2 2310.6 1787.2 2834.0 136.9 105.9 167.9 Pharyngitis 0−4 201690.0 180388.5 222991.5 29406.0 26300.3 32511.7 6396.4 3715.9 9076.9 932.6 541.8 1323.4 5−14 217890.0 205300.9 230479.1 21755.1 20498.2 23012.1 34252.0 26093.2 42410.8 3419.9 2605.3 4234.5 Overall(0–14) 419590.0 395166.5 444013.5 24865.5 23418.1 26312.9 40648.0 31534.6 49761.4 2408.9 1868.8 2948.9 Average: 2012−2014 Scarlet fever 0−4 791.4 725.4 857.4 115.4 105.8 125.0 457.4 279.6 635.2 66.7 40.8 92.6 5−14 2239.6 2119.4 2359.8 223.6 211.6 235.6 1401.5 1042.7 1760.3 139.9 104.1 175.8 Overall(0–14) 3031.1 2894.6 3167.5 179.6 171.5 187.7 1858.9 1417.2 2300.5 110.2 84.0 136.3 Pharyngitis 0−4 162830.0 148491.8 177168.2 23740.3 21649.8 25830.8 4692.4 2536.7 6848.2 684.1 369.8 998.5 5−14 166536.7 157376.1 175697.3 16627.8 15713.1 17542.4 23196.4 17283.7 29109.2 2316.0 1725.7 2906.4 Overall(0–14) 329370.0 312457.3 346282.7 19519.0 18516.7 20521.2 27888.7 21202.0 34575.3 1652.7 1256.5 2049.0 Discussion This is the first study to estimate the incidence of clinical cases of pharyngitis and GAS culture-positive pharyngitis among children in China. It was estimated that an average of 502933.3 clinical cases of pharyngitis occurred in children aged 0–14 years with about 29.8 (95 % CI: 28.3–31.3) cases per 100 child-years in Beijing, from 2012 to 2014. Although pharyngitis may seem relatively benign and unimportant, our study found that it caused as many as 329,370 outpatient visits each year with about 19.5 (95 % CI: 18.5–20.5) cases per 100 child-years in Beijing during 2012 to 2014, which indicated enormous use of health resources and medical costs. Our results are consistent with data from the National Ambulatory Medical Care Survey in United States, which reported that pharyngitis are responsible for 20 visits to a physician per 100 population annually in the United States [15]. These data highlighted that pharyngitis is a very common presenting symptom for seeking medications among children in Beijing, China. In our study, we estimated that an average of 45309.7 pharyngitis cases were laboratory confirmed per year from 2012 to 2014 in Beijing, with an annual incidence rate of 2.7 (95 % CI: 2.0–3.3) per 100 person-years for children, respectively, similar to the rate of GAS pharyngitis reported from a study in Europe (3.9 per 100 persons per year) [16], but lower than recently reported rates in Australia (13 per 100 person-years in 2001/2002 [17], and 14 per 100 person-years in 2001 [18]), and Fiji (14.7 per 100 person-years in 2006 [19]). In a peri-urban population of northern India, the incidence of GAS pharyngitis among 5–15-year-old schoolchildren, with an estimation of 95 per 100 person-years in 1995/1996 [20], was significantly higher than our estimation in Beijing. There were two reasons which may partially explain the lower rates of GAS pharyngitis in Beijing. First, GAS-related diseases are highly transmissible within populations characterized by crowding, limited access to hygiene and inadequate medical care [20]. As one of the most developed regions in China, Beijing has undergone epidemiological transition from communicable to non-communicable diseases as the predominant causes of morbidity and mortality. Scarlet fever is usually used as a proxy of GAS-related diseases because of its distinctive clinical features. During the period from 1949 to 2006, the incidence of scarlet fever has remarkably fallen from 488.3 to 1.86 cases per 100000 persons among all age population in Beijing [21]. Second, although immunity to GAS is emm-type specific [22], most of GAS samples from Beijing GAS surveillances were indentified as emm-1 and 12 types, from 2011 to 2014 [23]. Therefore, a large percentage of Beijing population had become immune to GAS infection from 2012 to 2014 because of the 2011 GAS epidemic in Beijing [7]. GAS causes a broad spectrum of diseases, ranging from mild superficial infections of the throat or skin to severe invasive infections and the post-streptococcal complications of acute rheumatic fever and acute post-streptococcal glomerulonephritis. In China, scarlet fever is the only notifiable disease among these GAS diseases according to the Law, but other GAS related infections are not notifiable and there are very few active surveillance systems. In our study, we found that the incidence of GAS pharyngitis was more than 10 times higher than that of scarlet fever. Accordingly, GAS pharyngitis caused more outpatient visits than GAS scarlet fever among children in Beijing. These results indicated that the disease burden of GAS pharyngitis has been significant, and posing great threats to the health of children in Beijing. Therefore, more epidemiological studies and surveillance of other GAS related infections should be developed in Beijing, and other regions in China. Our result demonstrated that the incidence of GAS pharyngitis among children aged 5–15 years was heavier than in preschool children aged 0–4 years due to the distinct difference of GAS culture-positive rates across the two agegroups. The result confirmed that disease burden of GAS is higher from 5–15 years than with younger children. However, the GAS culture-positive rates of scarlet fever didn’t vary smililarly. Unlike clinical cases of pharyngitis, clinical cases of scarlet fever had more distinctive clinical features and specific case definition in this study. Thus it was easily for clinicians to diagnose these cases without laboratory tests. Accordingly, the GAS culture-positive rate of scarlet fever was much higher than that of pharyngitis in all agegroups, and only a small difference was observed between the two agegroups. In this study, we observed the significant variation in yearly incidence, which may be explained by epidemic period of GAS. As a proxy of GAS-related diseases, epidemiological data showed that the epidemic period of scarlet fever was about 6–8 years in Beijing [21]. The factors including genetic variation, environmental factors, and host immune status might have contributed to the epidemic period [7]. But further studies should be conducted to analyse these factors. Our study has several limitations. First, because pharyngitis was not a notifiable disease in Beijing, some records for pharyngitis outpatient visits might be omitted by clinicians when they were busy to diagnose and treat the patients. Therefore, the estimated incidence for GAS pharyngitis in our study might have been underestimated. Second, there was no survey of health-seeking behavior among patients of GAS pharyngitis, so we have to use the age-specific consultation rate of all diseases from the Fourth National Health Services Survey of China (NHSS) in 2008. Nevertheless, the consultation rate of patient-defined Hand, Foot, and Mouse disease was estimated at 75.1 % among children under 5 years of age in Beijing, which is almost equal to the rate of all diseases at 72.7 % from NHSS [11]. The finding may indicate that the consultation rates for parameter estimations were acceptable and reliable. Third, as a model study, additional data from surveys of health-seeking behavior among patients of GAS diseases and reporting quality of physician practice could help refine the parameter estimates [10]. Although the multiplier model provided a quicker and more representative results than population-based surveys, a prospective, cohort study in China is needed to confirm our estimations since the model has not been used to estimate GAS burden. Conclusions In conclusion, we estimated an average of 29.8 clinical cases of pharyngitis and 2.7 cases of GAS culture-positive pharyngitis per 100 person-years among children aged 0–14 years, resulting in a large number of outpatient visits from 2012 to 2014 in Beijing, China. These estimates suggest that the disease burden of GAS pharyngitis has been significant, and posing great threats to the health of children in China. More epidemiological studies and surveillances should be developed to estimate the burden of GAS related diseases in Beijing, and other regions in China. ᅟ ᅟ Funding This work was supported by Beijing Health System High Level Health Technology Talent Cultivation Plan (2013–3–098), Beijing Talents Fund (2014000021223ZK36) and Beijing Science and Technology Planning Project (D141100003114001). The study sponsors had no role in study design; collection, analysis, and interpretation of data; writing the report; or the decision to submit the report for publication. Availability of data and materials The datasets supporting the conclusions of this article are included within the article. Authors’ contributions WS, WQ, and YP designed the study; WS, PX, MC, and ZD performed the data collection; WQ and YP coordinated and supervised the data collection; YZ and MC participated in the analysis of data; YZ and YP participated in the interpretation of data; WS drafted the initial manuscript. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of work. Competing interests The authors have indicated they have no competing interests. Ethics approval and consent to participate This study was approved by the Institutional Review Board and Human Research Ethics Committee of Beijing Center for Disease Prevention and Control. Written consent forms were obtained from the participants’ guardians before collecting the pharyngeal swab samples, and anonymity of the participants was guaranteed. ==== Refs References 1. Wessels MR Clinical practice. Streptococcal pharyngitis N Engl J Med 2011 364 7 648 655 10.1056/NEJMcp1009126 21323542 2. Bisno AL Acute pharyngitis N Engl J Med 2001 344 3 205 211 10.1056/NEJM200101183440308 11172144 3. Pfoh E Wessels MR Goldmann D Lee GM Burden and economic cost of group A streptococcal pharyngitis Pediatrics 2008 121 2 229 234 10.1542/peds.2007-0484 18245412 4. World Health Organization. The Current Evidence for the Burden of Group A Streptococcal Diseases. Discussion papers on Child Health. Department of Child and Adolescent Health and Development, WHO. http://whqlibdoc.who.int/hq/ 2005/WHO_FCH_CAH_05.07.pdf. Accessed 25 Nov 2015. 5. Carapetis JR Steer AC Mulholland EK Weber M The global burden of group A streptococcal diseases Lancet Infect Dis 2005 5 685 694 10.1016/S1473-3099(05)70267-X 16253886 6. Carapetis JR Currie BJ Kaplan EL Epidemiology and prevention of group A streptococcal infections: acute respiratory tract infections, skin infections, and their sequelae at the close of the twentieth century Clin Infect Dis 1999 28 2 205 210 10.1086/515114 10064227 7. Yang P Peng X Zhang D Wu S Liu Y Cui S Characteristics of group A Streptococcus strains circulating during scarlet fever epidemic, Beijing, China, 2011 Emerg Infect Dis 2013 19 6 909 915 10.3201/eid1906.121020 23735582 8. Ministry of Health WS282-2008, Diagnostic Criteria for Scarlet Fever 2008 9. Wang X Yang P Seale H Zhang Y Deng Y Pang X Estimates of the true number of cases of pandemic (H1N1) 2009, Beijing, China Emerg Infect Dis 2010 16 11 1786 1788 10.3201/eid1611.100323 21029546 10. Reed C Angulo FJ Swerdlow DL Lipsitch M Meltzer MI Jernigan D Estimates of the prevalence of pandemic (H1N1) 2009, United States, April-July 2009 Emerg Infect Dis 2009 15 12 2004 2007 10.3201/eid1512.091413 19961687 11. Wang X Wu X Jia L Li X Li J Li S Estimating the number of hand, foot and mouth disease amongst children aged under-five in Beijing during 2012, based on a telephone survey of healthcare seeking behavior BMC Infect Dis 2014 14 437 10.1186/1471-2334-14-437 25117760 12. Ministry of Health. Analysis report of National Health Services Survey in China, 2008. http://www.moh.gov.cn/cmsresources/mohwsbwstjxxzx/cmsrsdocument/doc9912.pdf. Accessed 25 Nov 2015. 13. Webb KH Does culture confirmation of high-sensitivity rapid streptococcal tests make sense? A medical decision analysis Pediatrics 1998 101 2 E2 10.1542/peds.101.2.e2 9445512 14. Beijing Bureau of Statistics. Release on Major Figures of the 2010 National Population Census in Beijing, China. http://www.bjstats.gov.cn/tjnj/rkpc-2010/indexch.htm. Accessed 25 Nov 2015. 15. Armstrong GL Pinner RW Outpatient visits for infectious diseases in the United States, 1980 through 1996 Arch Intern Med 1999 159 2531 2536 10.1001/archinte.159.21.2531 10573043 16. Duben J Jelínková J Jelínek J Rotta J Prospective study on streptococcal pharyngitis among a town population J Hyg Epidemiol Microbiol Immunol 1979 23 2 159 167 575381 17. Danchin MH Rogers S Kelpie L Selvaraj G Curtis N Carlin JB Burden of acute sore throat and group A streptococcal pharyngitis in school-aged children and their families in Australia Pediatrics 2007 120 5 950 957 10.1542/peds.2006-3368 17974731 18. Danchin MH Rogers S Selvaraj G Kelpie L Rankin P Vorich R The burden of group A streptococcal pharyngitis in Melbourne families Indian J Med Res 2004 119 Suppl 144 147 15232181 19. Steer AC Jenney AW Kado J Good MF Batzloff M Magor G Prospective surveillance of streptococcal sore throat in a tropical country Pediatr Infect Dis J 2009 28 6 477 482 10.1097/INF.0b013e318194b2af 19483515 20. Nandi S Kumar R Ray P Vohra H Ganguly NK Group A streptococcal sore throat in a periurban population of northern India: a 1-year prospective study Bull World Health Organ 2001 79 6 528 533 11436474 21. Shen YH Wang QY Wu J Wu J Gao T Ye H Study on epidemiological characteristics of Scarlet Fever in Beijing, China, 1949–2006 Strait J Prev Med 2008 14 2 30 31 22. Sanderson-Smith M De Oliveira DM Guglielmini J McMillan DJ Vu T Holien JK A systematic and functional classification of Streptococcus pyogenes that serves as a new tool for molecular typing and vaccine development J Infect Dis 2014 210 8 1325 1338 10.1093/infdis/jiu260 24799598 23. Peng X Yang P Wu S Lu G Shi W Zhao J Emm types of mutation in scarlet-fever-related group A streptococcal, among children in Beijing, 2011–2014 Zhonghua Liu Xing Bing Xue Za Zhi 2015 36 12 1397 1400 26850399

PMC005xxxxxx/PMC5002218.txt

==== Front Med AcupunctMed AcupunctacuMedical Acupuncture1933-65861933-6594Mary Ann Liebert, Inc. 140 Huguenot Street, 3rd FloorNew Rochelle, NY 10801USA 2761020910.1089/acu.2016.118810.1089/acu.2016.1188Case ReportsDC Electroacupuncture Effects on Scars and Sutures of a Patient with Postconcussion Pain Chevalier Antoine PhD-ND(cand), MPP1Armstrong Kelly OTR/L, MPP2Norwood-Williams C. PharmD, CPh3Gokal Raman MBChB, MD, FRCP21 University of Natural Health, Sterling, VA.2 Women's Integrative Healing Inc., St. Augustine, FL.3 Necessity Strategies LLC, Melbourne, FL.Address correspondence to:Antoine Chevalier, PhD-ND(cand), MPPUniversity of Natural Health21010 Southbank Street, #5050Sterling, VA 20165E-mail:antoine@healedandhappy.net01 8 2016 01 8 2016 01 8 2016 28 4 223 229 © Antoine Chevalier et al., 2016; Published by Mary Ann Liebert, Inc.2016This Open Access article is distributed under the terms of the Creative Commons License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.Abstract Introduction: This case study offers a detailed comparative analysis of the effects of direct-current electroacupuncture (DC-EA) on the autonomic nervous system (ANS), when DC-EA was applied to the cranial sutures and scars of a patient with a history of ischemic stroke and postconcussion syndrome (PCS) pain. Case: A 56-year-old female suffering from severe tremors and debilitating headaches requested acupuncture after conventional biomedicines failed to relieve her symptoms. Evaluations were performed to check the status of 27 ANS functions. These detailed evaluations were performed to obtain a baseline status of ANS function on this patient, who had a history of ischemic stroke, PCS, and chronic pain. All evaluations were repeated pre–post her DC-EA treatment. Results: This patient experienced significant relief from her symptoms after DC-EA treatment. An analysis of this patient's risk for ANS complications showed improvements in four key homeostatic markers post treatment. Conclusions: The ANS response of a patient with ischemic stroke, PCS, and chronic pain, who received electrical nerve stimulation using DC-EA reflected a measurable improvement in sympathetic tone, along with reductions in pain levels and PCS symptoms. The positive results in this case study could have applications to other pathologies that can be affected by the sympathetic nervous system activation on the body. Keywords:  Heart Rate VariabilityDirect-Current ElectroacupuncturePostconcussion SyndromeHeadache PainAtaxic Tremors ==== Body Introduction Postconcussion syndrome, also known as postconcussive syndrome or PCS, is a set of symptoms that might continue for weeks, months, or a year or more after a concussion—the latter being a minor form of traumatic brain injury.1–3 PCS symptoms include headaches, chronic pain, dizziness, fatigue, memory problems, and insomnia. The literature supports the concept that concussions from common traumas, as well as sports-related and military-related concussions, interfere with optimal levels of heart rate variability (HRV).4–6 According to the Centers for Disease Control and Prevention (CDC) there are up to 3.8 million concussions per year in the United States, with the incidence of sport-related concussions now reaching epidemic levels.7 The current standard for treatment of concussion is rest.8 It is widely accepted in science that imbalances of the parasympathetic (rest/healing/calming) and sympathetic (flight/fight/stress) branches of the autonomic nervous system (ANS) are directly linked to pain and a wide variety of and diseases.9–15 The sympathetic nervous system (SNS) is designed to facilitate short-term survival by creating a cascade of neurophysiologic responses. However, it is the upregulation or persistent tone in this system that is believed to be related to PCS symptomology.6–8 Microcurrent therapies involve applying weak direct currents (80 μA–1 mA) and are now being increasingly recognized as adjuncts for pain relief and ANS regulation.16–20 Microcurrent, applied transcranially, has been documented as being effective for treating anxiety, depression, insomnia, chronic pain,21,22 and, previously, for post-traumatic brain injuries,23–25 Microcurrent modulates cortical function and has been used to facilitate learning, alter behavioral performance, and improve impaired brain function.26–31 Studies of cranial sutures in adults have demonstrated a degree of movement, which contributes to the compliance and elasticity of the skull.32 Trauma to the head physically moves the cranial bones, stressing cranial sutures and affecting the ANS adversely in a similar manner as scars.33 Given that 80% of the fibers of the SNS go to the layers of the skin via the peripheral sympathetic fibers, these fibers' direct relationship to scars and cranial sutures is considered to be a major stressor of the ANS in patients who have PCS-related pain. Scars and trauma have long been recognized in neural therapy as a source of ANS upregulation.34–39 It is theorized that damaged local cells lose their normal membrane potential, transmitting abnormal electric signals throughout the rest of the body via the ANS.40,41 This case study addresses how direct-current electroacupuncture (DC-EA) applied to sutures and scars influences ANS functioning and patient pain levels, when applied in a PCS treatment protocol. This is significant, as there is no consensus in the literature regarding the best practice for electrostimulation applied to patients with postconcussive symptomology. The purpose of this case report is to convey the effect of DC-EA on the ANS of a patient with symptoms of PCS. Case Presentation A female client, G.G., suffering from severe tremors and debilitating headaches, presented for acupuncture treatment. G.G. was a 56-year-old, well-developed, well-nourished female in no apparent distress. Her weight was 72.3 kg, her height 170.2 cm, and her body mass index was 24.9. G.G.'s medical history involved a major car crash, 25 years prior (1991), in which she reported having hit her head on a window and being diagnosed with a concussion. The severe headaches and tremors started soon after her sustained injury and were currently present most of the time but were accentuated with activity; this is also the presumed cause of her tremors. She had a previous surgical scar on her abdominal area from a Cesarean section in 1984, and a chest scar from a biopsy procedure in 1993. The severe headaches started after she suffered a thrombotic right cerebrovascular stroke 15 years prior in 2001. She had rehabilitative treatments for weakness in her left limbs and had no other disabilities. She described having an aching pain at the bilateral frontal and corona capitis areas, with a Numeric Pain Rating Scale (NPRS) pain score of 7/10, and 10/10 on the “top of her head.”39–41 She wore a scarf over the top of her head because of the severity of pain she experienced when anything—even wind—touched the top of her head. Her cerebellar ataxic tremors were so severe that they interfered with writing and fine motor skills, such as pulling items out of her purse or wallet. Recently prior to presentation, she had received a medical suggestion from a physician to have a brain stimulation device implanted, with the hope of decreasing the hand tremors. She also reported having a tight and stiff neck. She was opposed to medication and preferred to manage her diet and exercise regimen. She avoided both alcohol and caffeine and reported a having a consistent routine of standard mealtimes and sleeping habits. She had not been taking any prescription medications, choosing only ibuprofen to manage her headaches. Her physical examination revealed an ataxic tremor but no other abnormalities; the diagnosis she was given was PCS with ataxic tremors that occurred post–ischemic stroke. No magnetic resonance imaging or computed tomography scanning was performed for her case. To date, she has not had any specific treatment for these symptoms. She gave informed consent to receiving the DC-EA treatment. Treatment DC-EA was applied simultaneously and bilaterally to acupuncture-trigger points located beside and along her cranial sutures and surgical scars using two22,23 Dolphin Neurostim (Acumed Medical LTD, Ontario, Canada) devices. These are U.S. Food and Drug Administration (FDA)–approved devices that are used to apply low-frequency, concentrated microcurrent stimulation (at 10K ohms) to relieve chronic pain.22,23 DC-EA application time was 30 seconds per point at approximately ½[[[inch mark]]]-intervals along the length of the scar-suture. As polarity of application is important, on one side of the scar-suture, the device was set to a negative pole (–) and on the other side of the scar-suture, the second device was set to a positive–negative pole (+/–). The intent of this methodology was to push a negatively charged current back and forth through a positively oriented suture or scar tissue. For this case study, all cranial sutures along with her two surgical scars, a transverse C-section scar, and a breast biopsy scar were treated. This treatment was applied once to this patient. ANS assessments were taken before and after electrostimulation with an ANS1 (Biosensor Equipment LLC, Houston TX), an FDA-approved device that is used to measure HRV, sympathetic, parasympathetic, adrenergic, and cardiovagal functions. The device enables a multimodal approach to assessing SNS, parasympathetic nervous system, and galvanic skin response functions through an autonomic nerve assessment, an arterial assessment, and an assessment of cardiometabolic markers.42,43 Measurements of 27 physiologic “markers” are included to place each patient measurement categorically into abnormal, borderline, and optimal goal columns. The Numeric Pain Rating Scale (NPRS) was used to evaluate this patient's pain. The NPRS is an 11-point scale from 0 to 10, with 0 being no pain and 10 being the most intense pain imaginable.44–46 The patient verbally selects a value that is most in line with the intensity of the pain that he or she has experienced in the last 24 hours, or the report will serve to rate pain experienced during a specific movement pattern or functional task. The NPRS has good sensitivity46 and excellent test–retest reliability. The objective data collection was aimed at revealing: (1) Whether or not DC-EA, when applied to sutures and scars, can modulate any variables within the ANS and scoring on the NPRS pain scale for a patient with PCS-related pain. (2) Whether or not DC-EA is a valid option for nonpharmacologic pain management of PCS-related pain and symptoms. Results G.G. reported a significant reduction in pain symptoms after treatment. Her initial report of a 7/10 severe baseline headache pain improved to “0” upon completion of the treatment and was reported during the post-test assessment. She also reported a 0/10 pain level when touching the top of her head and a loose and unrestricted neck. There was resolution of her presenting ataxic tremors. Six months later, at this writing, her headache and tremor symptoms have not returned, and she has not taken any headache medication since the DC-EA treatment (Table 1). Her neurologist commented to her that it appeared, according to her physical symptoms, that she was 93% improved from his baseline stroke assessment. Moreover, she has not taken any headache medication since the DC-EA application. Table 1. G.G.'s NPRS Pain Score Pre- and Post-DC EA Treatment Showing Dramatic Improvement in All Scores NPRS Pain Scale Pre–DC EA Treatment Post–DC EA Treatment Headaches 7/10 0/10 Touch top of head 10/10 0/10 Tight/restricted neck Tight Loose/no restrictions Severe ataxic tremors Severe None NPRS, Numeric Pain Rating Scale, DC EA, direct current electroacupuncture. ANS Pre–Post Chart Results Improvements were evident in all indicators of ANS function. The medical data related to the ANS improved, going from borderline into the optimal goal range. The autonomic nerve assessment was focused upon four key homeostatic markers in the analysis of risk for ANS complications: (1) HRV, for which a straightforward and useful metric is the standard deviation of all normal R–R intervals (those measured between consecutive sinus beats on a Holter electrocardiogram); (2) the Stress Index; (3) Standard Deviation of All Normal to Normal (SDANN) results; and (4) a high-frequency (HF) indicator of parasympathetic vagal nerve activity. ANS dysfunction risk is based upon HRV analysis at rest, and it comprises ANS activity and balance of the SNS and ANS. Details on these markers is provided in the sections below. (1) HRV Total power has been determined to be the main indicator of ANS activity and is reflective of variations in time intervals between heartbeats, known as HRV. Lower than normal HRV values are associated with negative outcomes in heart disease and increased risk for diabetic neuropathy.47 The results of G.G.'s ANS evaluation, post application of DC-EA, showed distinct differences in outcomes (Fig. 1), with marked improvement in the HRV scores after DC stimulation was applied to the noted sutures-scars. FIG. 1. Patient G.G.'s heart rate variability (HRV)—total power. This parameter showed a marked (91%) improvement. HRV normal is ≥780 units. DC-EA, direct current electroacupuncture. Low total power values could indicate a sedentary lifestyle and might indicate a need to increase physical activity. G.G.'s marked improvement in this value would indicate a significantly improved ANS response, compared to baseline and minimal target values, and should reflect an improvement in exercise tolerance and general health. (2) Stress Index This index is used to measure cardiac muscle oxygen demand related to heart work. The Stress Index is correlated with C-reactive protein (cRP) and is a marker of sympathetic failure.48 cRP is produced by the liver and increases with inflammation. HRV reflects the adaptability of the body to daily internal and external stressors that influence ANS function directly, as well as the stress the body is experiencing at the present time. High values indicate a risk for heart disease. Although G.G.'s values were below range to begin with, improvement for this indicator represented a marked improvement of her SNS (i.e., sympathetic downregulation), and subsequently, a reduction in her risk for heart disease if the improvement were to prove sustainable (Fig. 2). FIG. 2. Patient G.G.'s Stress Index showed a 20% reduction post direct-current electroacupuncture (DC-EA), although the pre level was below normal (≤ 180 units) for this index. (3) SDANN The SDANN is a measure of ANS activity and V02 (maximum oxygen consumption in the muscles).49 A low number indicates a sedentary lifestyle, with higher numbers primarily seen in athletes. Improvement in this marker would also imply an increased capacity for endurance, with a lower number reflecting the opposite response to physical activity. G.G.'s improvement in this indicator represents a marked downregulation and improvement of her SNS, as well as a subsequent increase in muscle oxygenation and exercise tolerance (Fig. 3). FIG. 3. Patient G.G.'s Standard Deviation of All Normal to Normal (SDANN) showed a 15.5% improvement from 31 to 37 units. Normal SDANN is ≥40 units. DC-EA, direct current electroacupuncture. (4) HF Indicator of Parasympathetic Vagal Nerve Activity Vagal tone is an internal biologic process referring to the activity of the vagus nerve, which serves as the key component of the parasympathetic branch of the ANS. Research suggests that decreased vagal activity or tone is associated with increased stress vulnerability and poor health.50 A low value (< 220), suggests sympathetic system predominance and possibility of stress or mental anxiety. GG's baseline value for this indicator was below the established target for healthy ANS balance, yet improvement was reflected with DC-EA appearing to have an impact towards the goal of achieving optimal value (Fig. 4). FIG. 4. Patient G.G.'s high-frequency (HF) vagal/parasympathetic activity showed an increase of 20% with a normal value of ≥220 units. DC-EA, direct current electroacupuncture. Discussion Sympathetic upregulation often results in stress and pain, which can make daily life miserable and lead to significantly impaired physical health.51,52 Both can be difficult to understand and, up to now, even harder to measure. Technology, such as advanced autonomic testing, can now provide real-time scientific evidence regarding the inner workings of human nervous systems in ill-health and disease,53–55 permitting collection of quantifiable data for the scientific study and education. The data in this case study clearly showed that application of DC-EA improved the patient's health significantly. She reported severe headache (7/10), head pain (10/10), and ataxic tremors prior the DC-EA treatment, and her pain scores and hand tremors were reduced to nil (0) after DC-EA application to her suture-scars. DC-EA also restored a more normal physiologic state throughout the patient's various nervous systems. There was marked improvement in her HRV Total Power score with DC-EA applied to acupuncture trigger points located beside sutures and scars. The patient's improvements in values would indicate a significantly improved ANS response, compared to baseline and minimal target values, and should reflect an improvement in exercise tolerance and general health. The improvement in the Stress Index represents a marked improvement of her SNS (i.e., downregulation) and, subsequently, a reduction in her risk for heart disease if the improvement were to persist. G.G.'s improvement in the SDANN suggests an improved tolerance to exercise. Her increased HF indicator of parasympathetic vagal nerve activity represents improved vagal–parasympathetic activity. Given that her pain and tremor symptomology have not returned 6 months post treatment, the positive outcomes of this case study suggest ANS regulation of PCS-related pain may help relieve PCS related symptoms. It is suggested that a low-amplitude DC current mimics human biocellular communications, and its application may create a shift or change in cellular membrane configuration, producing a bodywide therapeutic effect. These biochemical processes might provide a plausible explanation for the prolonged autonomic modulation after the DC microcurrent; this is an area where future research is required. Conclusions The current case study showed that DC-EA provided overall improvements in HRV, parasympathetic, and cardiometabolic markers, with subsequent long-term improvements in the patient's ataxic tremors and pain levels, suggesting a possible future role for DC-EA in the management of PCS, stroke, or stress-related diseases. Chronic PCS pain can limit quality of life, and restrict work and social engagement, and is often blamed for the development of drug dependencies of various forms. The changes produced in the ANS functions help validate the potential application of DC-EA to cranial sutures and scars as an option for clinicians treating patients with PCS-related chronic pain. However, further investigation is warranted with a much-larger focus group to confirm these results and to assess their duration. Author Disclosure Statement No competing financial conflicts exist. ==== Refs References 1 Mayoclinic.com Diseases and Conditions: Post Concussion Syndrome . Online document at: www.mayoclinic.com/health/post-concussion-syndrome/DS01020 Accessed 6 27 , 2016 2 McHugh T , Laforce R , Gallagher P , Quinn S Diggle P , Buchanan L Natural history of the long-term cognitive, affective, and physical sequelae of a minor traumatic brain injury . Brain Cogn . 2006 ;60 (2 ):209 –211 16646125 3 Mittenberg W , Strauman S Diagnosis of mild head injury and the postconcussion syndrome . J Head Trauma Rehabil . 2000 ;15 (2 ):783 –791 10739967 4 Barr WB , Prichep LS , Chabot R , Powell MR , McCrea M Measuring brain electrical activity to track recovery from sport-related concussion . Brain Inj . 2012 ;26 (1 ):58 –66 22107157 5 Baugley IJ , Heriseanu RE , Felmingham KL , Camerson ID Dysautonomia and heart rate variability following severe traumatic brain injury . Brain Inj . 2006 ;20 (4 ):437 –444 16716989 6 Blechert J , Michael T , Grossman P , Lajtman M , Wilhelm FH Autonomic and respiratory characteristics of post traumatic stress disorder and panic disorder . Psychosom Med . 2007 ;69 :935 –943 17991823 7 Centers for Disease Control and Prevention . National Center for Injury Prevention and Control, Division of Unataxical Injury Prevention . Online document at: www.cdc.gov/injury/about/ Accessed 6 27 , 2016 8 Moser RS , Glatts C , Schatz P Efficacy of immediate and delayed cognitive and physical rest for treatment of sports-related concussion . J Pediatr . 2012 ;161 (5 ):922 –926 22622050 9 Vinik AI , Maser RE , Ziegler D Autonomic imbalance: Prophet of doom or scope for hope? Diabet Med . 2011 ;28 (6 ):643 –651 21569084 10 Marsland AL , Bachen EA , Cohen S , Rabin B , Manuck SB Stress, immune reactivity and susceptibility to infectious disease . Physiol Behav . 2002 ;77 (4–5 ):711 –716 12527024 11 Hayano J , Yamada A , Mukai S , et al. Severity of coronary atherosclerosis correlates with the respiratory component of heart rate variability . Am Heart J . 1991 ;121 (4[pt1] ):1070 –1079 2008828 12 Bigger JT Jr , La Rovere MT , Steinman RC , Fleiss JL , Rottman JN , Rolnitzky LM , Schwartz PJ Comparison of baroreflex sensitivity and heart period variability after myocardial infarction . J Am Coll Cardiol . 1989 ;14 (6 ):1511 –1158 2809012 13 Carney RM , Saunders RD , Freedland KE , Stein P , Rich MW , Jaffe AS Association of depression with reduced heart rate variability in coronary artery disease . Am J Cardiol . 1995 ;76 (8 ):562 –564 7677077 14 Spallone V , Ziegler O , Freeman R , et al. ; Toronto Consensus Panel on Diabetic Neuropathy . Cardiovascular autonomic neuropathy in diabetes: Clinical impact, assessment, diagnosis, and management . Diabetes Metab Res Rev . 2011 ;27 (7 ):639 –653 21695768 15 Lewis JE , Lantigua L , Atlas SE , et al. A cross-sectional assessment to detect type 2 diabetes with endothelial and autonomic nervous system markers using a novel system . J Diabetes Metab Disord . 2014 ;13 (1 ):118 25530952 16 Poltawski L , Watson T Bioelectricity and microcurrent therapy for tissue healing—a narrative review . Phys Ther Rev . 2009 ;14 (2 ):104 –114 17 Cheng N , Van Hoof H , Bockx E , et al. The effects of electric currents on ATP generation, protein synthesis, and membrane transport in rat skin . Clin Orthop Relat Res . 1982 ;171 :264 –272 18 McMakin C Microcurrent therapy: A novel treatment method for chronic low back myofascial pain . J Bodyw Movement Ther . 2004 ;8 (2 ):143 –153 19 Armstrong K Electro-Therapy Exposed . Rehabpub. 2016 Online document at: www.rehabpub.com/2016/01/electroTherapy-exposed Accessed 6 12 , 2016 20 Department of Health and Human Services . Dolphin DC-EA, FDA 510K. Online document at: www.accessdata.fda.gov/cdrh_docs/pdf13/K133789.pdf Accessed 6 22 , 2016 21 Zaghi S , Acar M , Hultgren B , Boggio PS , Fregni F Noninvasive brain stimulation with low-intensity electrical currents: Putative mechanisms of action for direct and alternating current stimulation . Neuroscientist . 2010 ;16 (3 ):285 –307 20040569 22 Kirsch DL , Nichols F Cranial electrotherapy stimulation for treatment of anxiety, depression, and insomnia . Psychiatr Clin North Am . 2013 ;36 (1 ):169 –176 23538086 23 Boggio PS , Bermpohl F , Vergara AO , et al. Go–no-go task performance improvement after anodal transcranial DC stimulation of the left dorsolateral prefrontal cortex in major depression . J Affect Disord . 2007 ;101 (1–3 ):91 –98 17166593 24 Childs A , Crismon M The use of cranial electrotherapy stimulation in post-traumatic amnesia: A report of two cases . Brain Inj . 1988 ;2 (3 ):243 –247 3262394 25 Michals M , Crismon M , Misko J , Childs A A double-blind, sham controlled evaluation of cranial electrotherapy stimulation in posttraumatic memory impairment . J Head Trauma Rehabil . 1993 ;8 (4 ):77 –86 26 Nitsche MA , Schauenburg A , Lang N , Liebetanz D , Exner C , Paulus W , Tergau F Facilitation of implicit motor learning by weak transcranial direct current stimulation of the primary motor cortex in the human . J Cogn Neurosci . 2003 ;15 (4 ):619 –626 12803972 27 Uy J , Ridding MC Increased cortical excitability induced by transcranial DC and peripheral nerve stimulation . J Neurosci Methods . 2003 ;127 (2 ):193 –197 12906948 28 Kincses TZ , Antal A , Nitsche MA , Bártfai O , Paulus W Facilitation of probabilistic classification learning by transcranial direct current stimulation of the prefrontal cortex in the human . Neuropsychologia . 2004 ;42 (1 ):113 –117 14615081 29 Antal A , Kincses TZ , Nitsche MA , Bártfai O , Paulus W Excitability changes induced in the human primary visual cortex by transcranial direct current stimulation: Direct electrophysiological evidence . Invest Ophthalmol Vis Sci . 2004 ;45 (2 ):702 –707 14744917 30 Boggio PS , Ferrucci R , Rigonatti SP , Covre P , Nitsche M , Pascual-Leone A , Fregni F Effects of transcranial direct current stimulation on working memory in patients with Parkinson's disease . J Neurol Sci . 2006 ;249 (1 ):31 –38 16843494 31 Boggio PS , Khoury LP , Martins DC , Martine OE , de Macedo EC , Fregni F Temporal cortex direct current stimulation enhances performance on a visual recognition memory task in Alzheimer's disease . J Neurol Neurosurg Psychiatr . 2009 ;80 (4 ):444 –447 32 Boundless Anatomy and Physiology . Unique Features of the Skull: Sutures . Online document at: www.boundless.com/physiology/textbooks/boundless-anatomy-and-physiology Accessed 3 12 , 2016 33 Fregni F , Boggio PS , Lima MC , et al. A sham-controlled, phase II trial of transcranial direct current stimulation for the treatment of central pain in traumatic spinal cord injury . Pain . 2006 ;122 (1–2 ):197 –209 16564618 34 Kidd RF Interference fields . In: Neural Therapy . Renfrew, Ontario, Canada : Custom Printers of Renfrew Ltd.; 2005 , pp. 24 –40 35 Williams L. Blocks to healing: Chronic dominant foci . In: Radical Medicine: Cutting-Edge Natural Therapies That Treat the Root Causes of Disease . Rochester, VT : Healing Arts Press ; 2007 :402 –420 36 Brook M. The Importance of Scar Tissue Release Therapy . Massage Today .2009 ;9 (6 ). Online document at: www.massagetoday.com/mpacms/mt/article.php?id=14020 Accessed 4 24 , 2013 37 Baker R , Urso-Baiarda F , Linge C , Grobbelaar A Cutaneous scarring: A clinical review . Dermatol Res Pract . 2009 ;2009 :625376 20585482 38 Defalque RJ Painful trigger points in surgical scars . Anesth Analg . 1982 ;61 (6 ):518 –520 7200742 39 Dosch P Manual of Neural Therapy According to Huneke , 11th ed. Stuttgart, Germany : Haug Publishers ; 1984 40 Hauser R Prolo Your Pain Away: Curing Chronic Pain with Prolotherapy . Oak Park, IL : Beulah Land Press ; 1998 41 Fischer L Neurotherapy after Huneke [in German]. Stuttgart : Hippokrates Verlag ; 1998 42 Gibbons CH , Cheshire PW Jr. , Fife TD Model Coverage Policy: Autonomic Testing . American Academy of Neurology 9 2 , 2014 Online document at: www.aan.com/uploadedFiles/Website_Library_Assets/Documents/3.Practice_Management/1.Reimbursement/1.Billing_and_Coding/5.Coverage_Policies/14%20Autonomic%20Testing%20Policy%20v001.pdf Accessed 6 12 , 2016 43 Pop-Busui R , Evans GW , Gerstein HC , et al. ; Action to Control Cardiovascular Risk in Diabetes Study Group . Effects of cardiac autonomic dysfunction on mortality risk in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial . Diabetes Care . 2010 ;33 (7 ):1578 –1584 20215456 44 Williamson A , Hoggart B Pain: A review of three commonly used pain rating scales . J Clin Nurs . 2005 ;14 (7 ):798 –804 16000093 45 Jensen MP , McFarland CA Increasing the reliability and validity of pain intensity measurement in chronic pain patients . Pain . 1993 ;55 (2 ):195 –203 8309709 46 Hawker GA , Mian S , Kendzerska T , French M Measures of adult pain: Visual Analog Scale for Pain (VAS Pain), Numeric Rating Scale for Pain (NRS Pain), McGill Pain Questionnaire (MPQ), Short-Form McGill Pain Questionnaire (SF-MPQ), Chronic Pain Grade Scale (CPGS), Short Form-36 Bodily Pain Scale (SF-36 BPS), and Measure of Intermittent and Constant Osteoarthritis Pain (ICOAP) . Arthritis Care Res (Hoboken) . 2011 ;63 (suppl11 ):S240 –S252 22588748 47 Eller NH Total power and high frequency components of heart rate variability and risk factors for atherosclerosis . Auton Neurosci . 2007 ;131 (1–2 ):123 –30 16987712 48 Strang F , Schunkert H C-reactive protein and coronary heart disease: All said—is not it? Mediators Inflamm . 2014 ;2014 :757123 24808639 49 Raj SR , Roach DE , Koshman M , Sheldon RS Standard deviation of sequential five-minute R–R interval means (SDANN) is a prognostic marker, but not necessarily an autonomic marker . J Am Coll Cardiol . 2006 ;48 (6 ):1285 –1286 16979024 50 Jarrett ME , Cain KC , Barney PG , et al. Balance of autonomic nervous system predicts who benefits from a self-management intervention program for irritable bowel syndrome . J Neurogastroenterol Motil . 2016 ;22 (1 ):102 –111 26459461 51 Porges SW Vagal tone: A physiologic marker of stress vulnerability . Pediatrics . 1992 ;90 (3[pt2] ):498 –504 1513615 52 Kuo LE , Czarnecka M , Kitlinska JB , Tilan JU , Kvetňanský R , Zukowska Z Chronic stress, combined with a high-fat/high-sugar diet, shifts sympathetic signaling toward neuropeptide Y and leads to obesity and the metabolic syndrome . Ann N Y Acad Sci . 2008 ;1148 :232 –237 19120115 53 Molina PE Neurobiology of the stress response contribution of the sympathetic nervous system to the neuroimmune axis . Traumatic Inj . 2005 ;24 (1 ):3 –10 54 Ravits JM AAEM minimonograph #48: Autonomic nervous system testing . Muscle Nerve . 1997 ;20 (8 ):919 –937 9236782 55 Vinik Al , Ziegler D Diabetic cardiovascular autonomic neuropathy . Circulation . 2007 ;115 (3 ):387 –397 17242296

PMC005xxxxxx/PMC5002269.txt

==== Front CellCellCell0092-86741097-4172Cell Press S0092-8674(16)30926-610.1016/j.cell.2016.07.019ArticleThe Deubiquitinase OTULIN Is an Essential Negative Regulator of Inflammation and Autoimmunity Damgaard Rune Busk 1Walker Jennifer A. 1Marco-Casanova Paola 1Morgan Neil V. 2Titheradge Hannah L. 34Elliott Paul R. 1McHale Duncan 5Maher Eamonn R. erm1000@medschl.cam.ac.uk6∗McKenzie Andrew N.J. anm@mrc-lmb.cam.ac.uk1∗∗Komander David dk@mrc-lmb.cam.ac.uk17∗∗∗1 Medical Research Council Laboratory of Molecular Biology, Francis Crick Avenue, Cambridge Biomedical Campus, Cambridge CB2 0QH, UK2 Institute of Cardiovascular Sciences, College of Medical and Dental Sciences, University of Birmingham, Birmingham B15 2TT, UK3 Department of Clinical Genetics, Birmingham Women’s NHS Foundation Trust, Birmingham B15 2TG, UK4 Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham B15 2TT, UK5 New Medicines, UCB Pharma, Slough SL1 3WE, UK6 Department of Medical Genetics, University of Cambridge and Cambridge NIHR Biomedical Research Centre, Cambridge Biomedical Campus, Cambridge CB2 0QQ, UK∗ Corresponding author erm1000@medschl.cam.ac.uk∗∗ Corresponding author anm@mrc-lmb.cam.ac.uk∗∗∗ Corresponding author dk@mrc-lmb.cam.ac.uk7 Lead Contact 25 8 2016 25 8 2016 166 5 1215 1230.e20 26 10 2015 20 5 2016 13 7 2016 © 2016 The Authors2016This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).Summary Methionine-1 (M1)-linked ubiquitin chains regulate the activity of NF-κB, immune homeostasis, and responses to infection. The importance of negative regulators of M1-linked chains in vivo remains poorly understood. Here, we show that the M1-specific deubiquitinase OTULIN is essential for preventing TNF-associated systemic inflammation in humans and mice. A homozygous hypomorphic mutation in human OTULIN causes a potentially fatal autoinflammatory condition termed OTULIN-related autoinflammatory syndrome (ORAS). Four independent OTULIN mouse models reveal that OTULIN deficiency in immune cells results in cell-type-specific effects, ranging from over-production of inflammatory cytokines and autoimmunity due to accumulation of M1-linked polyubiquitin and spontaneous NF-κB activation in myeloid cells to downregulation of M1-polyubiquitin signaling by degradation of LUBAC in B and T cells. Remarkably, treatment with anti-TNF neutralizing antibodies ameliorates inflammation in ORAS patients and rescues mouse phenotypes. Hence, OTULIN is critical for restraining life-threatening spontaneous inflammation and maintaining immune homeostasis. Graphical Abstract Highlights • Mutation of OTULIN causes OTULIN-related autoinflammatory syndrome (ORAS) in humans • Anti-TNF treatment reverses inflammation in ORAS patient and OTULIN-deficient mice • OTULIN deficiency deregulates M1-polyUb signaling and causes sterile inflammation • Loss of OTULIN has cell-type-specific effects on LUBAC abundance and signaling A homozygous hypomorphic mutation in the deubiquitinase OTULIN causes a potentially fatal autoinflammatory disorder, which is reconciled in mouse models. Published: August 11, 2016 ==== Body Introduction Protein ubiquitination regulates virtually every aspect of cellular homeostasis, in large part through structurally and functionally distinct polyubiquitin (polyUb) signals (Komander and Rape, 2012). PolyUb chains can be linked via one of seven Ub Lys (K) residues (e.g., K63-linked chains) or via Ub Met1 (M1), forming M1-linked (also known as linear) chains. The latter have important roles in regulating the immune system, in which they contribute to regulating nuclear factor-κB (NF-κB) transcription factors that orchestrate immune responses (Bonizzi and Karin, 2004). Ub chains regulate canonical NF-κB activation by mediating timed degradation of the inhibitor of κB (IκB) proteins but also serve as a scaffolding, recruitment, and activation platform in receptor signaling complexes. Non-degradative K63- and M1-linked chains mediate the key upstream event of recruiting the TGFβ-activated kinase (TAK1) and the IκB kinase (IKK) complexes, respectively (Jiang and Chen, 2012). K63 and M1 linkages occur in the same Ub polymers (Emmerich et al., 2013), facilitating TAK1 and IKK co-localization and cross-activation. M1-linked chains are generated by the linear ubiquitin chain assembly complex (LUBAC) consisting of HOIP, HOIL-1, and SHARPIN (Fiil and Gyrd-Hansen, 2014, Iwai et al., 2014). LUBAC is recruited to many immune receptors, including TNF-R1, IL-1R, CD40, TLRs, and NOD2, in a Ub-dependent manner. At the receptors, LUBAC ubiquitinates a host of targets, including RIPK1, RIPK2, MyD88, IRAKs, and NEMO, directly or on pre-existing Ub chains (Fiil and Gyrd-Hansen, 2014, Iwai et al., 2014). Genetic loss of LUBAC components leads to immunodeficiency (MacDuff et al., 2015) and inflammatory phenotypes in mice (Gerlach et al., 2011, Ikeda et al., 2011, Tokunaga et al., 2011, Tokunaga et al., 2009), which can be rescued by co-deletion of TNF-R1 (Gerlach et al., 2011, Kumari et al., 2014, Peltzer et al., 2014, Rickard et al., 2014). Mutations in LUBAC components also cause inflammatory conditions in humans (Boisson et al., 2015, Boisson et al., 2012). Hence, loss of M1-linked chains imbalances immune signaling. Several deubiquitinating enzymes (DUBs), including A20, CYLD, and Cezanne, act as negative regulators of NF-κB signaling and are essential for resolving inflammation and the return to homeostasis (Harhaj and Dixit, 2012). OTULIN (also known as FAM105B or Gumby) is the only DUB known to specifically cleave M1 linkages (Keusekotten et al., 2013, Rivkin et al., 2013). OTULIN directly binds the LUBAC component HOIP, and knockdown of OTULIN in human cell lines increases M1-linked chains on LUBAC and its substrates (Elliott et al., 2014, Fiil et al., 2013, Keusekotten et al., 2013, Rivkin et al., 2013, Schaeffer et al., 2014). Strikingly, while LUBAC translocates to receptor signaling complexes (RSCs), OTULIN is not stably associated with TNF or NOD2 RSCs (Draber et al., 2015), and how it regulates signaling complexes, e.g., TNF signaling, is unclear (Hrdinka et al., 2016). Indeed, the physiological role of OTULIN in the immune system has remained unstudied, since OTULIN loss-of-function mutations lead to early embryonic lethality (E12.5–E14) in mice due to defective Wnt signaling and angiogenesis (Rivkin et al., 2013). Here, we describe that a homozygous hypomorphic OTULIN mutation in a consanguineous family causes a potentially fatal autoinflammatory disorder termed OTULIN-related autoinflammatory syndrome (ORAS), which can be managed by Infliximab (anti-TNF neutralizing antibody). We recapitulate key features of ORAS in mouse models of OTULIN deficiency. In an acute model, induced loss of OTULIN in immune cells leads to multi-organ inflammation and deterioration of animals within a few days; this can be ameliorated by anti-TNF, but not by neutralization of other upregulated cytokines. In addition, loss of OTULIN in myeloid cells generates a chronic model in which mice display increased serum levels of inflammation-associated cytokines and chemokines and develop splenomegaly and autoimmunity. In bone-marrow-derived macrophages (BMDMs), loss of OTULIN leads to overproduction of M1-linked Ub chains and dysregulated NF-κB activation and cytokine secretion. Strikingly, while mice lacking OTULIN in B or T cells do not display overt inflammatory phenotypes, further analysis indicates that these OTULIN-deficient cells have downregulated LUBAC components HOIP and SHARPIN, but not HOIL-1. Together, the data from mouse models and human patients clearly establish OTULIN and M1-linked polyUb chains as key regulators of immune homeostasis, inflammation, and autoimmunity and reveal cell-type-specific effects of OTULIN in immune cells. Results Hypomorphic OTULIN Germline Mutation in Patients with Idiopathic Inflammatory Disease From the late 1990s, three premature newborns (born at week 34, week 36, and week 28+6, respectively) from a consanguineous family (Figure 1A) displayed severe idiopathic inflammatory symptoms. Within days to weeks after birth, they had repeated episodes of systemic inflammation with diarrhea and elevated serum C-reactive protein (CRP; Figure 1B) and white blood cell (WBC) and neutrophil count (Figure 1C) without evidence of infection. All three eventually developed relapsing nodular panniculitis with neutrophil infiltrate and recurrent fevers. The patients showed reduced growth parameters and also exhibited painful swollen joints as well as elevated immunoglobulin levels and autoantibodies in serum (Table S1 and Experimental Models and Subject Details). Genetic testing excluded Mediterranean fever, chronic infantile neurological cutaneous articular syndrome, tumor necrosis factor-receptor associated periodic syndrome, and hyper IgD syndrome as the cause of the symptoms (Experimental Models and Subject Details). SNP array analysis of the three affected individuals identified three regions of extended homozygosity at chromosomes 5 and 15 shared by all three individuals (Figure S1A and Tables S2 and S3). The largest region of shared homozygosity was at chromosome 5p15 from 13,802,063 to 16,722,976 bp between rs795541 and rs4702171. Genotyping of all available family members with microsatellite markers D5S817, D5S1991, D5S1992, D5S1963, and D5S416 confirmed linkage to this candidate region. Within the ∼3 Mb candidate region of chromosome 5p15, sequencing of FAM105A, FBXL7, MARCH11, ZNF622, and FAM34B revealed no candidate pathogenic variants. However, we detected a homozygous missense substitution in OTULIN/FAM105B in all affected individuals (c.815T>C; p.Leu272Pro) (Figure 1D). The parents of patient V:2 (IV:1 and IV:2) were both heterozygous for the substitution (Figure 1D), and none of the healthy siblings were homozygous for the c.815T>C variant (data not shown). Moreover, the variant was not present in the ∼12,000 and ∼120,000 alleles represented in the Exome Variant Server and EXaC datasets, respectively, either. Consistently, whole-exome sequencing of patient V:2 revealed no other homozygous or previously annotated pathogenic variants likely to cause the disease phenotype (Tables S2 and S3). Thus, we termed the described syndrome OTULIN-related autoinflammatory syndrome (ORAS) (Figure 1E). Leu272 is located in a helix of the catalytic OTU domain that forms part of the binding pocket for M1-linked diUb, and mutation to Pro would be predicted to disrupt OTULIN Ub binding (Figure 1F). Recombinant OTULINL272P displayed small differences in overall fold and reduced thermal stability (Figures S1B and S1C), and expression of OTULINL272P in HEK293 cells consistently led to significantly reduced levels of OTULINL272P as compared to wild-type protein (Figure S1D). OTULINL272P, although less stable, maintained its interaction with HOIP (Figure S1D). In addition to reduced stability, OTULINL272P was 1,000–10,000 times less active toward M1-linked di- and tetraUb (Figures 1G and S1E). M1-diUb binding to catalytically inactive OTULINC129A/L272P was significantly diminished yet not completely abrogated (Figure S1F). OTULINL272P was still M1-linkage specific (Figure S1G), as the affected Ub binding site does not dictate OTULIN specificity (Keusekotten et al., 2013). Reduced stability and activity of OTULIN suggested an impact on M1-linked polyUb in ORAS patients. Patient blood samples confirmed the presence of OTULIN, albeit at slightly reduced levels as compared to control samples, as well as the presence of SHARPIN (Figure 1H). Strikingly, while samples from age-matched controls showed barely detectable levels of M1-linked chains, this chain type was strongly increased in the ORAS patient sample. We conclude that loss of OTULIN function and increased levels of M1-linked polyUb in human cells may indeed lead to the severe inflammation and autoimmunity observed in ORAS. Neutralization of TNF Ameliorates Systemic Inflammation in ORAS Importantly, ORAS can be managed by treatment with the anti-TNF antibody Infliximab. Treatment of affected patients with prednisolone (general corticosteroidal immunosuppressant), azathioprine, and methotrexate (anti-proliferative immunosuppressive drugs), as well as Anakinra (recombinant IL-1R-antagonist), had little or no effect on the symptoms (Figure 1E and Experimental Models and Subject Details). Patient IV:3 died at 16 months of age of pneumococcal septicemia, and patient IV:4 died from an episode of systemic inflammation leading to acute renal failure and pulmonary edema at age 5. Of note, neither patient IV:3 nor IV:4 were treated with Infliximab. By contrast, patient V:2 responded well to treatment with Infliximab (Figures 1B and 1C) and is currently alive (age 11) with well-controlled disease. Infliximab treatment drastically reduced the symptoms, and serum levels of CRP as well as WBC and neutrophil counts in blood, returned to normal ranges after treatment (Figures 1B, 1C), suggesting a TNF-mediated pathogenesis in the patients. Deletion of Otulin in Mouse Immune Cells Drives Acute Systemic Inflammation To understand how OTULIN deficiency caused autoinflammation and autoimmunity, we attempted to recapitulate ORAS in mouse models. Due to embryonic lethality (Rivkin et al., 2013), heterozygous Otulin+/LacZ mice (Figures S2A and S2B) were bred with appropriate inducible and/or tissue-specific Cre expressing mouse strains. OTULIN was expressed in several tissues in Otulin+/LacZ embryos and adult wild-type mice, including spleen and thymus (Figures S2C and S2D). In immune cells, OTULIN was expressed in T cells, B cells, and natural killer (NK) cells and prominently in dendritic cells and macrophages (Figure 2A). We generated CreERT2-OtulinLacZ/flox mice, in which Otulin can be ablated in all cells with tamoxifen administration. Tamoxifen administration led to mice becoming moribund within a day (data not shown), suggesting strong phenotypes also in adult mice in addition to reported developmental defects (Rivkin et al., 2013). This prevented further study, and to investigate the role of OTULIN in the immune system, we generated CreERT2-Otulinflox mixed bone marrow chimeras (Figure 2B), which were healthy (data not shown) and showed similar levels of CD45.1+ and CD45.2+ cells in vivo (Figure S2E). Six to eight weeks after reconstitution, chimeric mice were treated with tamoxifen to induce Otulin ablation (Figure S2F). Strikingly, this resulted in rapid weight loss in CreERT2-OtulinLacZ/flox chimeras as compared to controls (Figures 2C and S2G). Weight loss was accompanied by a pronounced increase in the number of circulating neutrophils in CreERT2-OtulinLacZ/flox chimeras (Figure 2D), while the number of lymphocytes and monocytes and the overall number of blood cells or splenocytes did not change (Figures 2D, S2E, and S2H). We also observed a marked increase in the pro-inflammatory cytokines TNF and IL-6, the neutrophil cytokines G-CSF and KC, and the monocyte/macrophage chemokine MCP-1 in serum of CreERT2-OtulinLacZ/flox chimeras compared with CreERT2-Otulin+/flox chimeras and vehicle-treated controls (Figures 2E and 2F and Table S4). These cellular and molecular mediators are indicative of an inappropriate and damaging inflammatory response mediated by OTULIN-deficient myeloid cells. As expected in the short timeframe, cytokines associated with adaptive immunity were not elevated (Figure 2E and Table S4). Necropsy of the chimeras did not reveal any gross anatomical changes, but flow cytometry analysis showed substantial infiltration of CD11b+Gr-1+ neutrophils in the peritoneal lavage (PL), spleen, and liver and, to some extent, lung and kidney in CreERT2-OtulinLacZ/flox chimeras compared with CreERT2-Otulin+/flox or vehicle-treated controls (Figures 2G, 2H, and S2I–S2L). Histological analysis confirmed immune cell infiltration in the livers of CreERT2-OtulinLacZ/flox chimeras with inflammatory foci scattered in the parenchyma (Figure 2I, arrowheads). With the exception of PL, where neutrophils were recovered in lower numbers after tamoxifen treatment, tissue-infiltrating neutrophils consisted of overall similar numbers of CD45.1+ wild-type cells and CD45.2+ OTULIN-deficient cells (Figure S2M). This suggests that loss of OTULIN does not cause aberrant cell death in immune cells (Figure S2E) and that the high levels of cytokines and chemokines present activate the entire immune cell population. Collectively, this shows that deletion of Otulin in immune cells leads to spontaneous and severe acute systemic inflammation characterized by rapid weight loss, increased levels of pro-inflammatory cytokines (in particular, TNF, G-CSF, and IL-6) in serum, neutrophilia with all the hallmarks of emergency granulopoiesis (see Figures S3A–S3C), and infiltration of neutrophils into multiple tissues. This highly pro-inflammatory phenotype of the OTULIN-deficient adult mice is unlikely to arise from defects in Wnt signaling during development (Rivkin et al., 2013) but indeed shows several of the cardinal symptoms of ORAS (Figure 1E). Rescue of Systemic Inflammation in OTULIN-Deficient Mice We next tested whether the strong ORAS-like phenotype can be reversed by antibody treatments, as observed in the human disease. It was unclear whether any of the upregulated cytokines were independently regulated by OTULIN/M1-linked chains. Hence, we tested whether neutralization of TNF, G-CSF, or IL-6 impacted on the weight loss and emergency granulopoiesis observed in CreERT2-OtulinLacZ/flox chimeric mice. In accordance with the effective treatment of ORAS patients with Infliximab, we found that anti-TNF neutralizing antibodies, administered in parallel with tamoxifen-induced Otulin deletion, completely ameliorated weight loss (Figure 3A). While the numbers of blood neutrophils were still elevated (Figure 3B), infiltrating neutrophils in spleen and PL of CreERT2-OtulinLacZ/flox were reduced to a level comparable to isotype-treated CreERT2-Otulin+/flox controls (Figure 3C). In contrast, anti-G-CSF or anti-IL-6 were unable to rescue the phenotype fully. While anti-G-CSF treatment strongly reduced blood neutrophil count and infiltration, it failed to rescue weight loss and mice remained cachexic (Figures 3D–3F). Anti-IL-6 treatment had no impact on neutrophilia, but weight loss was partially rescued (Figure 3G–3I). This differential amelioration extended to neutrophil production in the bone marrow, where anti-G-CSF, but not anti-IL-6, reduced the number of uncommitted blood cells (“LSK” cells, for Lin−Sca1+c-Kit+ cells) and immature neutrophils to levels observed in isotype-treated controls (Figures S3D–S3F). This confirms that G-CSF drives emergency granulopoiesis in CreERT2-OtulinLacZ/flox chimeras and may contribute to the neutrophilia observed in ORAS patients (Figure 1D). Strikingly, anti-TNF treatment of CreERT2-OtulinLacZ/flox chimeras resulted in a global reduction of cytokines, including G-CSF, KC, and TNF itself. IL-6 was reduced to background levels (Figures 3J and S3G). Anti-G-CSF did not affect TNF or IL-6 levels and, in fact, caused elevated production of KC and G-CSF itself (although its functional blockade ameliorated neutrophilia) (Figures 3J and S3H). Anti-IL-6 increased MCP-1 levels, possibly indicating a pathway for compensation in the absence of IL-6 (Figures 3J and S3I). Together, these data implicate TNF as the pre-eminent factor in driving inflammation in OTULIN-deficient mice, although other cytokines and chemokines clearly contribute to the composite phenotype (Figure 3K). This finding correlates with the fact that anti-TNF treatment reverses the inflammatory symptoms in the ORAS patient and therefore suggests that OTULIN-deficient mice provide a good model to understand the mechanisms underlying ORAS. Differential Effects of OTULIN Deletion in Cells of the Immune System To dissect immune mechanisms contributing to the strong OTULIN deficiency phenotype, we employed mouse models in which Otulin can be constitutively ablated in specific immune cell lineages, namely T cells (“CD4Cre” mice, expressing Cre recombinase under control of the Cd4 promoter), B cells (“MB1Cre” mice, expressing Cre under control of the Mb1 promoter), and myeloid cells (“LysMCre” mice, expressing Cre under control of the LysM promoter, leading to knockout in macrophages, neutrophils, and some dendritic cells). Surprisingly, loss of OTULIN in T cells or B cells generated healthy mice with no overt inflammatory phenotypes, as assessed by phenotypic, blood cell, cytokine, and serum IgG analysis (Figures 4 and S4A–S4C and Tables S5 and S6). Closer inspection of the B cell subsets in the MB1Cre-OtulinLacZ/flox mice indicated a trend toward lower B1 cell numbers but did not reach statistical significance (Figure S4C). In stark contrast, deletion of OTULIN in myeloid cells resulted in a strong inflammatory phenotype. LysMCre-OtulinLacZ/flox mice were born at Mendelian ratios, appeared healthy at weaning, and showed similar survival as compared to controls but were slightly smaller in terms of body weight (Figures S4D and S4E). Indeed, closer inspection of 3- to 9-month-old LysMCre-OtulinLacZ/flox mice revealed a marked increase in the size of lymphoid organs and liver (Figures 4E and S4F) and prominent leukocytosis with increased numbers of circulating neutrophils, lymphocytes, and monocytes (Figures 4F) as compared to controls. Deletion of Otulin in Myeloid Cells Leads to Chronic Inflammation and Autoimmunity Strikingly, serum analysis showed that 16 out of 25 tested cytokines and chemokines were markedly elevated in LysMCre-OtulinLacZ/flox mice (Figures 4G and 4H and Table S7), including the pro-inflammatory cytokines TNF, IL-6, and IL-1β; the neutrophil and monocyte attractants/activators MCP-1 and MIP-1α; and G-CSF. Interestingly, cytokines associated with T cell activation and adaptive immunity, such as IL-2, IL-4, IFNγ, and RANTES, were also elevated (Figures 4G and 4H and Table S7), as were serum IgG levels (Figure 4I). Collectively, these findings indicate ongoing, systemic inflammation in LysMCre-OtulinLacZ/flox mice involving both innate and adaptive immune cells. Histological analysis revealed substantial immune cell infiltration in livers of LysMCre-OtulinLacZ/flox mice, particularly around veins but also in the parenchyma, and a distorted splenic architecture (Figures 5A and 5B). Collagen deposits in liver (Figure 5A, right) and spleen (Figure 5B, right) were visible and consistent with chronic inflammation. Moreover, germinal center activation was evident in LysMCre-OtulinLacZ/flox spleens (Figure 5B, left), suggesting B cell hyperactivation and potential for immunoglobulin-mediated pathology. Flow cytometric analysis showed that LysMCre-OtulinLacZ/flox mice had greater numbers of both CD11b+Gr-1+ neutrophils and CD11b+Gr-1− macrophages in PL, spleen, liver, lungs, and kidney compared with LysMCre-Otulin+/flox mice (Figures 5C, 5D, S5A, and S5B). LysMCre-OtulinLacZ/flox mice also had increased numbers of CD8+ T cells in liver and kidney (Figures 5D, S5C, and S5D), consistent with systemic chronic inflammation and involvement of the adaptive immune system. The number of CD4+ T cells remained normal (data not shown). High-serum IgG and the appearance of active germinal centers in LysMCre-OtulinLacZ/flox mice suggested B cell hyperactivation, leading to higher levels of antibodies in the serum, a hallmark of autoimmunity. Indeed, the amounts of serum immunoglobulin isotypes IgG1, IgG2a and b, IgA, and IgM were elevated in LysMCre-OtulinLacZ/flox serum, while IgG3 levels were unchanged when compared to LysMCre-Otulin+/flox (Figure 5E). This indicated polyclonal B cell activation and suggested that LysMCre-OtulinLacZ/flox mice could be autoimmune. Consistent with this, ELISA analysis revealed higher antibody reactivity against extractable nuclear antigens (ENA), double-stranded DNA (dsDNA), and Smith antigen in LysMCre-OtulinLacZ/flox serum (Figure 5F). B cell activating factor (BAFF) is the primary cytokine that governs peripheral B cell tolerance, and increased BAFF levels can lead to T-cell-independent B cell activation and immunoglobulin production (Groom et al., 2007). Indeed, we found that LysMCre-OtulinLacZ/flox mice had elevated levels of BAFF in serum (Figure 5G), suggesting that secretion of BAFF resulting from OTULIN deficiency leads to breakdown of peripheral tolerance and activation of autoreactive B cells. Collectively, these results show that OTULIN is essential in myeloid cells to prevent unwarranted secretion of cytokines leading to inflammation, as well as autoimmunity (Figure 3K). OTULIN Deficiency Leads to Sterile Autoactivation of Inflammatory Pathways We next investigated the molecular mechanisms resulting in the strong inflammatory phenotypes, using BMDMs from LysMCre-Otulin+/flox, LysMCre-OtulinLacZ/flox, or LysMCre-Otulindel/flox mice, which were cultured and studied for pathway activation by western blotting in the absence of any exogenous stimulation. Interestingly, in agreement with the patient samples, the levels of M1-linked Ub chains were markedly increased (∼8-fold on average) in OTULIN-deficient BMDMs, in particular in the high-molecular-weight range, while K63-linked, K48-linked, or total Ub levels were unchanged (Figure 6A). Strikingly, this resulted in NF-κB activation evidenced by degradation of IκBα and increased phosphorylation of p65/RelA in the absence of exogenous stimuli (Figure 6B). Cycloheximide chase experiments confirmed the increased turnover of IκBα (Figure 6C), a hallmark of NF-κB pathway activation. This correlated with increased transcription of NF-κB target genes such as Tnf, Nfkbia (IκBα), Il6, and Tnfaip3 (A20) (Figure 6D) and concomitant secretion of TNF and IL-6 (Figure 6E). Also, LysMCre-Otulin+/flox and LysMCre-OtulinLacZ/flox BMDMs were similarly viable when left untreated or when treated with LPS, TNF, or Staurosporine (Figure 6F), suggesting no apparent sensitization to cell death in these cells. Importantly, while SHARPIN levels were slightly reduced, the levels of HOIP and HOIL-1 were unchanged in LysMCre-OtulinLacZ/flox BMDMs (Figure 6G), consistent with strong production of M1-linked polyUb in these cells (Figure 6A). Protein levels of NEMO/IKKγ, A20, and CYLD were similar to controls (Figure 6G). LUBAC signaling may impact alternative inflammatory pathways, yet in LysMCre-Otulindel/flox BMDMs, NF-κB-inducing kinase (NIK) was not stabilized (indicating no activation of non-canonical NF-κB), and IRF-3 or p38 phosphorylation was not increased (suggesting that IRF or MAPK signaling pathways were not active) (Figure 6H). Finally, we examined the response of BMDMs to TNF. Since these cells secrete TNF (Figure 6E), the observed cellular response may, in part, originate from autocrine signaling. Indeed, when MEFs were treated with anti-TNF antibodies, exogenous TNF was unable to induce IκBα degradation and p65 phosphorylation (Figure S6A). Importantly, when TNF was neutralized in the culture medium of untreated BMDMs, IκBα levels decreased with similar kinetics as compared to isotype-treated BMDMs (Figure 6C, quantified in Figure 6I). This provided further strong evidence for a cell-autonomous activation of NF-κB in OTULIN-deficient BMDMs. Autocrine signaling would be further enhanced in a positive feedback mechanism if loss of OTULIN would strengthen the response to TNF. Indeed, stimulation of BMDMs with TNF led to stronger transcriptional upregulation of Tnf in cells from LysMCre-Otulindel/flox (Figure 6J), consistent with a role of OTULIN as a negative regulator in this pathway. Together, this shows that OTULIN-deficient macrophages are unable to control LUBAC-mediated production of M1-linked polyUb chains and that this signal leads to stimulus-independent basal NF-κB activation and “sterile” inflammatory signaling, possibly enhanced by autocrine feedback. The idea of sterile inflammation is further supported in CreERT2-OtulinLacZ/flox chimeras that have been treated with broad-spectrum antibiotics to reduce the microbial load; these mice display an identical inflammatory phenotype as compared to untreated CreERT2-OtulinLacZ/flox chimeras (Figures S6B–S6G, compare Figure 2). Cell-Type-Specific Differences in Response to OTULIN Deficiency While loss of OTULIN from myeloid cells and the concomitant dysregulation of M1-linked polyUb resulted in a profound inflammatory phenotype, this was not evident in mice lacking OTULIN in B or T cells (Figure 4). This is surprising since M1 signaling has recently been implicated in T cell signaling (Park et al., 2016, Redecke et al., 2016) and B cell signaling (Sasaki et al., 2013, Satpathy et al., 2015, Yang et al., 2016). The DUB-regulating M1-linked polyUb in B and T cells has not been elucidated. To understand this, the M1-Ub signaling cascade was characterized in purified T and B cells from respective knockout mice. To our great surprise, this revealed an almost complete loss of LUBAC components HOIP and SHARPIN, but not HOIL-1, in either of the two independent models (Figures 7A and 7B). As discussed above, LysMCre-OtulinLacZ/flox BMDMs showed mildly reduced SHARPIN levels but no difference in HOIP levels. Further evaluation of myeloid cells purified from various tissues likewise showed no loss of HOIP or HOIL-1 and only marginally reduced SHARPIN levels (Figure 7C). Such tissue- and subunit-specific regulation of LUBAC components has not been observed previously. We next investigated the mechanism of OTULIN-dependent LUBAC regulation. Transcription of HOIP (Rnf31), SHARPIN (Sharpin), and HOIL-1 (Rbck1) genes was unaltered in CD4Cre-Otulindel/flox T cells and MB1Cre-OtulinLacZ/flox B cells (Figures 7D and 7E). In contrast, SHARPIN (but not HOIP) levels were partially rescued in purified T and B cells after MG132 treatment, indicating that loss of SHARPIN is due to proteasomal degradation (Figures 7F and 7G). This reveals a role for OTULIN in the stabilization of LUBAC components that warrants further investigation. Discussion We provide genetic evidence that OTULIN is a crucial in vivo regulator of inflammation in mice and humans, identify TNF as a key driver of the phenotypes caused by OTULIN deficiency, and show that OTULIN deficiency and ORAS can be treated with anti-TNF/Infliximab in mice and humans. The discovery of an OTULIN mutation in human patients with a severe inflammatory syndrome, ORAS, highlights how deregulation of a single Ub chain signal, M1-linked polyUb, causes human disease. Mutations of LUBAC components are also associated with inflammatory syndromes (Boisson et al., 2012, Boisson et al., 2015), yet these constitute loss-of-function mutations with regards to M1-polyUb signaling. In contrast, loss of function of OTULIN leads to an amplification of the M1-polyUb signal (Figure 1H). Strikingly, the net result, inflammation, is the same, highlighting how the essential M1-polyUb signal is delicately balanced to determine a cellular output. Many DUBs have been identified to play key roles in regulating signaling pathways; however, we are unaware of another enzyme that, when removed from cells, has such global effects on a single chain type. The marked upregulation of M1-linked chains in OTULIN-deficient BMDMs and ORAS patient samples suggests that there is little redundancy for regulation of this chain type. Two recent reports implicate CYLD in the regulation of M1-polyUb signaling at receptor complexes (Draber et al., 2015, Hrdinka et al., 2016). Quantitative mass spectrometry analysis showed that OTULIN is not present at TNF and NOD2 RSCs, but while one group suggests that OTULIN’s role is restricted to LUBAC homeostasis (Draber et al., 2015), our work suggests that, while OTULIN might not stably associate with the receptor complexes, it plays active roles in receptor signaling (Hrdinka et al., 2016) (see above, Figure 6J). Physiologically, our data support a model whereby the unrestricted accumulation of M1-linked Ub chains leads to sterile inflammation due to stimulation-independent activation of NF-κB (Figure 7H). Importantly, the role of OTULIN in NF-κB inhibition is conceptually distinct from other NF-κB-induced negative feedback regulators, such as A20 (Harhaj and Dixit, 2012). OTULIN is not under transcriptional control by NF-κB (Fiil et al., 2013, Keusekotten et al., 2013) but appears to constitutively and efficiently remove M1-linked polyUb signals. In the absence of OTULIN, LUBAC activity is unrestricted, and such signals accumulate in a deregulated fashion to initiate uncoordinated NF-κB activation and sterile inflammation. The downstream effects of this include secretion of pro-inflammatory cytokines, immune cell activation, and infiltration, culminating in severe inflammatory phenotypes (Figures 3K and 7H). Strikingly, this is not always the case. Mice with OTULIN deficiency in B or T cells do not show overt phenotypes, and this is due to downregulation of LUBAC components HOIP and SHARPIN, but not HOIL-1, at the protein level, in OTULIN-deficient cells (Figures 7A, 7B, 7H). Genetic loss of SHARPIN in cpdm mice destabilizes both HOIP and HOIL-1 (Gerlach et al., 2011, Ikeda et al., 2011, Tokunaga et al., 2011), contrasting the stability of HOIL-1 in our models. This may point to differentiated use of LUBAC components in immune cells. Second, OTULIN only cleaves non-degradative M1-linked polyUb. Knockdown of OTULIN leads to M1 polyubiquitination of all three LUBAC components (Fiil et al., 2013, Hrdinka et al., 2016). It is possible that these chains, at least on HOIP and SHARPIN, are extended or edited in T and B cells by an unidentified E3 ligase to turn a non-degradative into a degradative signal. The mechanism of OTULIN-regulated LUBAC degradation requires further study, but the findings clearly suggest that whole-body OTULIN deficiency generates a complex composite phenotype, signified by cell-type-specific regulation of M1 signaling. There are also subtle but important differences in the observed inflammatory phenotypes comparing CreERT2-OtulinLacZ/flox chimeras and LysMCre-OtulinLacZ/flox mice, which present differently in terms of severity of symptoms of disease, e.g., cachexia/weight loss. The nearly identical absolute levels of serum TNF, the key mediator of the inflammatory phenotype (Figures 1 and 3) indicates that myeloid cells are responsible for TNF production, but this does not explain the phenotypic difference. Rescue experiments suggest that G-CSF regulates emergency granulopoiesis, but not cachexia. Instead, we uncover a prominent role of IL-6, a key mediator of TNF- and inflammation-induced cachexia (Morley et al., 2006). This cytokine is higher in CreERT2-OtulinLacZ/flox chimeras as compared to LysMCre-OtulinLacZ/flox mice (Figures 2F and 4H), explaining, in part, the difference in the severity of phenotypes. Further differences include the temporal profile, with the chimera model being an induced, acute model and the myeloid-specific knockout being a constitutive, lifetime model. Finally, the numerous other upregulated cytokines in LysMCre-OtulinLacZ/flox mice, which includes anti-inflammatory IL-10, will generate a delicate balance of responses, enabling these mice to deal with inflammation induced by loss of OTULIN. We show that excess M1-linked polyUb in cells of the immune system is harmful, but it most likely also affects other cell types and organs. For example, ORAS patients suffer from panniculitis, and it will be informative to study the role of OTULIN in skin homeostasis and inflammation. Moreover, the roles of OTULIN in embryonic development, potentially due to effects in the Wnt signaling pathway (Rivkin et al., 2013), require further investigation. Our immune-specific mouse models are not suitable to investigate this in detail. Significantly, the fact that the here-described OTULIN-related autoinflammatory syndrome/ORAS can be treated with TNF-neutralizing antibodies suggests a potential therapeutic strategy to treat conditions caused by excessive M1-linked polyUb signaling. STAR★Methods Key Resources Table REAGENT or RESOURCE SOURCE IDENTIFIER Antibodies InVivoMAb anti-mouse TNF (neutralizing) (clone XT3.11) BioXcell Cat#BE0058; RRID: AB_1107764 InVivoMAb anti-mouse IL-6 (neutralizing) (clone MP5-20F3) BioXcell Cat#BE0046; RRID: AB_1107709 anti-mouse G-CSF (neutralizing) (clone 67604) R&D Systems Cat#MAB414; RRID: AB_2085954 InVivoMAb Rat IgG1 isotype control (clone HRPN) BioXcell Cat#BE0088; RRID: AB_1107775 anti-B220 conjugated to FITC (clone RA3-6B2) eBioscience Cat#11-0452-81; RRID: AB_465053 anti-CD3ε conjugated to FITC (clone 145-2C11) BioLegend Cat#17-0031-81; RRID: AB_312670 anti-CD3ε conjugated to PE-Cy7 (clone 145-2C11) BioLegend Cat#100319; RRID: AB_312684 anti-CD3ε conjugated to APC (clone 145-2C11) eBioscience Cat#17-0031; RRID: AB_469314 anti-CD4 conjugated to Alexafluor700 (clone GK1.5) eBioscience Cat#56-0041-82; RRID: AB_493999 anti-CD4 conjugated to Alexafluor647 (clone GK1.5) eBioscience Cat#50-0041-82; RRID: AB_469773 anti-CD4 conjugated to biotin (clone GK1.5) BioLegend Cat#100403; RRID: AB_312688 anti-CD4 conjugated to FITC (clone H129.19) BD Biosciences Cat#561831; RRID: AB_10892800 anti-CD4 conjugated to BrilliantViolet785 (clone RM4-5) BioLegend Cat#100551; RRID: AB_11218992 anti-CD5 conjugated to BrilliantViolet-510 (clone 53-7.3) BioLegend Cat#100627; RRID: AB_2563930 anti-CD5 conjugated to FITC (clone 53-7.3) BD Biosciences Cat#553020; RRID: AB_394558 anti-CD8α conjugated to FITC (clone 53-6.7) eBioscience Cat#11-0081-81; RRID: AB_464914 anti-CD8α conjugated to APC (clone 53-6.7) eBioscience Cat#17-0081-81; RRID: AB_469334 anti-CD8α conjugated to PE-Cy7 (clone 53-6.7) eBioscience Cat#25-0081-81; RRID: AB_469583 anti-CD8α conjugated to PE (clone 53-6.7) eBioscience Cat#12-0081-81; RRID: AB_465529 anti-CD8α conjugated to biotin (clone 53-6.7) eBioscience Cat#13-0081-81; RRID: AB_466345 anti-CD11b conjugated to PE-Cy7 (clone M1/70) eBioscience Cat#25-0112-81; RRID: AB_469587 anti-CD11b conjugated to eFluor-450 (clone M1/70) eBioscience Cat#48-0112-82; RRID: AB_1582236 anti-CD11b conjugated to BrilliantViolet-421 (clone M1/70) BioLegend Cat#101235; RRID: AB_10897942 anti-CD11b conjugated to FITC (clone M1/70) BioLegend Cat#101205; RRID: AB_312788 anti-CD11c conjugated to PE (clone N418) BioLegend Cat#117307; RRID: AB_313776 anti-CD11c conjugated to FITC (clone N418) eBioscience Cat#11-0114-81; RRID: AB_464939 anti-CD19 conjugated to FITC (clone eBio1D3) eBioscience Cat#11-0193-81; RRID: AB_657667 anti-CD19 conjugated to PerCP-Cy5.5 (clone eBio1D3) eBioscience Cat#45-0193-82; RRID: AB_1106999 anti-CD19 conjugated to Alexafluor700 (clone 6D5) BioLegend Cat#115528; RRID: AB_493735 anti-CD21/CD35 conjugated to APC (clone 7E9) BioLegend Cat#123411; RRID: AB_940395 anti-CD23 conjugated to PE-Cy7 (clone B3B4) eBioscience Cat#25-0232-81; RRID: AB_469603 anti-CD25 conjugated to BrilliantViolet-421 (clone PC61) BioLegend Cat#102033; RRID: AB_10895908 anti-CD43 conjugated to PE (clone S7) BD Biosciences Cat#553271; RRID: AB_394748 anti-CD44 conjugated to PE (clone IM7) eBioscience Cat#12-0441-81; RRID: AB_465663 anti-CD45.1 conjugated to eFluor-450 (clone A20) eBioscience Cat#48-0453-82; RRID: AB_1272189 anti-CD45.1 conjugated to BrilliantViolet-510 (clone A20) BioLegend Cat#110741; RRID: AB_2563378 anti-CD45.2 conjugated to Alexafluor700 (clone 104) eBioscience Cat#56-0454-81; RRID: AB_657753 and anti-CD62L conjugated to BrilliantViolet-421 (clone MEL-14) BioLegend Cat#104435; RRID: AB_10900082 anti-cKit/CD117 conjugated to PerCP-Cy5.5 (clone 2B8) BioLegend Cat#105823; RRID: AB_2131598 anti-FcεRIα conjugated to FITC (clone MAR-1) BioLegend Cat#134305; RRID: AB_1626102 anti-IgD conjugated to BrilliantViolet-650 (clone 11-26c.2a) BioLegend Cat#405721; RRID: AB_2562731 anti-IgM conjugated to PerCP-Cy5.5 (clone RMM-1) BioLegend Cat#406511; RRID: AB_2075944 anti-Ly6G/Gr-1 conjugated to FITC (clone RB6-8C5) eBioscience Cat#11-5931-81; RRID: AB_465313 anti-Ly6G/Gr-1 conjugated to PE (clone RB6-8C5) eBioscience Cat#12-5931-81; RRID: AB_466044 anti-Ly6G/Gr-1 conjugated to APC (clone RB6-8C5) eBioscience Cat#17-5931-81; RRID: AB_469475 anti-NK1.1 conjugated to BrilliantViolet-421 (clone PK136) BioLegend Cat#108731; RRID: AB_10895916 anti-NK1.1 conjugated to FITC (clone PK136) BioLegend Cat#108705; RRID: AB_313392 anti-Sca1 conjugated to PE-Cy7 (clone D7) eBioscience Cat#25-5981-81; RRID: AB_469668 anti-Ter-119 conjugated to FITC (clone TER-119) eBioscience Cat#11-5921; RRID: AB_2206887 anti-A20 Cell Signaling Technology Cat#4625; RRID: AB_2204524 anti-Actin (clone C4) Millipore Cat#MAB1501R; RRID: AB_94235 anti-CYLD Santa Cruz Biotechnology Cat#sc-74435; RRID: AB_1122022 anti-GAPDH Ambion Cat#AM4300; RRID: AB_437392 anti-HOIL-1/RBCK1 Novus Biologicals Cat#NBP2-27105; RRID: AB_2576210 anti-HOIL-1/RBCK1 Santa Cruz Biotechnology Cat#sc-49718; RRID: AB_2175281 anti-HOIP/RNF31 Abcam Cat#46322; RRID: AB_945269 anti-mouse HOIP/RNF31 Laboratory of Kazuhiro Iwai Tokunaga et al., 2011; RRID: N/A anti-IκBα Cell Signaling Technology Cat#9242; RRID: AB_10694550 anti-IRF-3 Santa Cruz Biotechnology Cat#sc-9082; RRID: AB_2264929 anti-phospho-IRF-3 (pS396) Cell Signaling Technology Cat#4947; RRID: AB_823547 anti-Lys63-linked ubiquitin (clone Apu3) Millipore Cat#05-1308; RRID: AB_1587580 anti-Lys48-linked ubiquitin (clone Apu2) Millipore Cat#05-1307; RRID: AB_1587578 anti-Met1-linked/linear ubiquitin (clone LUB9) LifeSensors Cat#AB130; RRID: AB_2576211 anti-Met1-linked/linear ubiquitin (clone 1E3) Millipore Cat#MABS199; RRID: AB_2576212 anti-NIK Santa Cruz Biotechnology Cat#sc-8417; RRID: AB_628021 anti-NEMO/IKKγ Santa Cruz Biotechnology Cat#sc-8330; RRID: AB_2124846 anti-OTULIN Cell Signaling Technology Cat#14127; RRID: AB_2576213 anti-p38 (clone M138) Abcam Cat#ab31828; RRID: AB_881839 anti-phospho-p38 (pT180/pY182) (clone ERP18120) Abcam Cat#ab195049; RRID: AB_2576214 anti-p65/RelA Cell Signaling Technology Cat#8242; RRID: AB_10860244 anti-phospho-p65/RelA (pS563) (clone 93H1) Cell Signaling Technology Cat#3033; RRID: AB_331284 anti-SHARPIN Proteintech Cat#14626-1-AP; RRID: AB_2187734 anti-ubiquitin (clone UBI-1) Novus Biologicals Cat#NB300-130; RRID: AB_2238517 anti-mouse CD19-coupled MACS MicroBeads Miltenyi Biotec Cat#130-052-201; RRID: N/A anti-mouse CD11b-couple MACS MicroBeads Miltenyi Biotec Cat#130-049-601; RRID: N/A anti-Biotin-coupled MACS MicroBeads Miltenyi Biotec Cat#130-090-485; RRID: N/A Chemicals, Peptides, and Recombinant Proteins Recombinant mouse M-CSF R&D Systems Cat#416-ML-050 Recombinant mouse TNF GIBCO Cat#PMC3014 Ultrapure LPS from E. coli K12 InvivoGen Cat#tlrl-peklps Staurosporine from Streptomyces sp. Sigma-Aldrich Cat#S6942; CAS 62996-74-1 MG132 (Z-Leu-Leu-Leu-al) Sigma-Aldrich Cat#C2211; CAS 133407-82-6 Tamoxifen Sigma Cat#T5648-1G; CAS 10540-29-1 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliym bromide (MTT) Sigma Cat#M2128-250MG; CAS 298-93-1 Collagenase D Roche Cat#11088866001 Liberase Roche Cat#5401020001 DNase I Sigma-Aldrich Cat#D5025 Percoll PLUS GE Healthcare Cat#17-5445-02 Recombinant OTULIN-WT Laboratory of David Komander Keusekotten et al., 2013 Recombinant OTULIN-C129A/L272P This paper N/A Critical Commercial Assays MILLIPLEX MAP Cytokine/Chemokune Magnetic Bead Panel – Premixed 25-plex Merck Millipore Cat#MCYTOMAG-70K-PMX Mouse BAFF/BLyS/TNFSF13B Quantikine ELISA kit R&D Systems Cat#MBLYS0 Mouse G-CSF Quantikine ELISA kit R&D Systems Cat#MCS00 Mouse anti-nuclear antigen/ENA IgA+G+M ELISA kit Alpha Diagnostics Cat#5210 Mouse anti-Smith antigen IgA+G+M ELISA kit Alpha Diagnostics Cat#5405 Mouse anti-dsDNA IgA+G+M ELISA kit Alpha Diagnostics Cat#5110 Mouse Total IgG Ready-Set-Go ELISA kit eBioscience Cat#88-50400-22 IgG Mouse ELISA kit Abcam Cat#ab151276 Mouse Isotyping Panel 1 multiplex sandwich electrochemiluminescence immunoassay kit Meso Scale Discovery Cat#K15183-1 MACS Dead Cell Removal Kit Miltenyi Biotec Cat#130-090-101 Experimental Models: Organisms/Strains Mouse: ACTB-FLPe Laboratory of Susan M. Dymeki Rodríguez et al., 2000 Mouse: ROSA26-CreERT2 Laboratory of Ernesto Bockamp Hameyer et al., 2007 Mouse: LysM-Cre Laboratory of Irmgard Förster Clausen et al., 1999 Mouse: Mb1-Cre Laboratory of Michael Reth Hobeika et al., 2006 Mouse: Cd4-Cre Laboratory of Christopher B. Wilson Lee et al., 2001 Experimental Models: Cell Lines Otulin/Fam105b-targeted JM8A3.N1 ES cells EuMMCR Fam105btm1a(EUCOMM)Hmgu HEK293 N/A N/A Recombinant DNA pEGFP-N1-OTULIN-WT Laboratory of David Komander Keusekotten et al., 2013. pEGFP-N1-OTULIN-L272P This paper N/A Sequence-Based Reagents For primer sequences, please see Table S8. Software and Algorithms HomozygosityMapper Seelow et al., 2009 http://www.homozygositymapper.org/ Annovar / wAnnovar Yang and Wang, 2015 http://wannovar.usc.edu/index.php SAMtools Li et al., 2009 http://samtools.sourceforge.net/ Other Exome Variant Server NHLBI GO Exome Sequencing Project (ESP) http://evs.gs.washington.edu/EVS/ EXaC Browser Exome Aggregation Consortium http://exac.broadinstitute.org/gene/ENSG00000154124 Contact for Reagent and Resource Sharing Further information and requests for reagents may be directed to, and will be fulfilled by the corresponding author David Komander (dk@mrc-lmb.cam.ac.uk). Experimental Models and Subject Details ORAS Patients Consent Information Written informed consent was obtained for all subjects and family members (n = 7). The study was approved by the South Birmingham Research Ethics Committee and performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. Clinical Description of ORAS Patients, Related to Figure 1 and Tables S1, S2, and S3 Patient IV:3 The consultand is the first child born to consanguineous parents (III:3 and III:4) (Figure 1A) at 34 weeks gestation via normal vaginal delivery and was small for gestational age. She presented at 3 weeks of age with protracted diarrhea and failure to thrive. She required total parenteral nutrition (TPN) and later nasogastric tube feeds. During her lengthy admission she was noted to have recurrent episodes of a widespread nodular erythematous rash associated with fever, elevated white blood cell count, neutrophilia, raised C-reactive protein (CRP) (Figures 1B-D), and an exacerbation of her diarrhea. An initial skin biopsy, taken as the rash was resolving, showed non-specific changes. Duodenal biopsies showed microvillous dystrophy. She also had mild hepatomegaly. A liver biopsy showed TPN-associated liver disease with micronodular cirrhosis and macrovisicular steatosis. Liver function later returned to normal, although hepatomegaly and the permanent damage from cirrhosis persisted. There are no records of splenomegaly in this patient. Echocardiogram identified an atrial septal defect. The patient was extensively investigated to identify the cause of her protracted diarrhea and episodic skin rash including a number of immunological investigations. Of note she had normal alpha 1 anti-trypsin level, normal functional antibodies, normal lymphocyte subsets although with a modest increase in CD19+ B cells (2120 cells/mm3; normal 500-1500), normal lymphocyte function tests, neutrophil function test, and normal NBT tests. She did have elevated IgG, IgM, and IgA levels (Table S1), raised CRP (Figure 1C), elevated C1q (171 mg/L; normal 80-150 mg/L) and C3 (2.75 g/L; normal 0.75-1.75 g/L), and she was positive for anti-neutrophil cytoplasmic antibody (ANCA) and anti-smooth muscle antibody (SMA), both 1:100 titers. When she was treated with intravenous methylprednisolone, her fever, diarrhea, and rash would resolve, but these clinical features would return when attempts were made to change to oral prednisolone. Azathioprine and methotrexate were also trialed. In a course of 3 months she had three cases of severe bilateral pneumonia progressing to sepsis and renal tubular necrosis, hematuria, proteinuria, and raised urea and creatinine requiring admittance to the pediatric intensive care unit. She survived two episodes, but sadly died at 16 months during the third episode of pneumococcal septicemia, which led to respiratory collapse. Patient IV:4 The consultand is the second child born to consanguineous parents (III:3 and III:4) (Figure 1A) at 36 weeks gestation via normal vaginal delivery with a birth weight of 2 kg. She was diagnosed with relapsing nodular panniculitis at 3 days of age. A skin biopsy showed no evidence of vasculitis. This rash, similarly to her sister (patient IV:3) and cousin (patient V:2), was episodic and associated with diarrhea (bloody at presentation), vomiting, fever, and painful swollen joints and difficulty sleeping. She had elevated white blood cell counts with pronounced neutrophilia as well as raised CRP levels (Figures 1B-D), as well as elevated C3 (2.77 g/L; normal 0.75-1.75 g/L), but was negative for ANCA autoantibodies (she was not tested for other autoantibodies). She had normal lymphocyte subsets. There are no records of hepatosplenomegaly in this patient. She required nasogastric tube feeds for failure to thrive. All growth parameters were below the 0.4th centile. She also had bilateral cataracts diagnosed at the age of 5 months and developmental delay. The cataracts were not present in the neonatal period on formal ophthalmological examination and were not secondary to steroid use. She was treated with prednisolone from the age of 1 month, and with azathioprine and Anakinra (recombinant IL-1R-antagonist) unsuccessfully. She had recurrent urinary tract infections and viral illnesses, which coincided with immunosuppressive therapy. During a severe episode of inflammation at the age of five she demonstrated features consistent with a high cell turnover including elevated potassium, phosphate, uric acid, LDH, elevated white blood cell count and a metabolic acidosis. She was admitted to pediatric intensive care with acute renal failure, pulmonary edema and an ileus, but sadly passed away just before turning 5 years old. Treatment had included intubation and ventilation, intravenous methylprednisolone, dexamethasone, and methotrexate. In this patient, Familial Mediterranean Fever, Chronic Infantile Neurological Cutaneous Articular syndrome, Tumor necrosis factor Receptor-Associated Periodic Syndrome, and Hyper IgD syndrome had all been excluded by genetic analysis. Patient V:2 The consultand is the second child born prematurely at 28+6 weeks gestation to consanguineous parents (IV:1 and IV:2) (Figure 1A) with a birth weight of 1.23 kg. He developed relapsing nodular panniculitis at 8 weeks of age while he was in the neonatal intensive care unit. Prior to the appearance of his skin rash, he had repeated episodes of possible infection with raised CRP levels and elevated white blood cell count and neutrophilia (Figures 1B-D), but no focus of infection could be identified. The rash was biopsied and this confirmed inflammation in the dermis extending into the subcutaneous layer with a mixed inflammatory cell infiltrate. No granulomas or vasculitis was seen. He has had frequent flare-ups involving widespread painful lumps in the skin lasting 2 days to 2 weeks. During these episodes he was systemically unwell with fever, vomiting, diarrhea (sometimes bloody), inflamed painful joints, swollen feet, and weight loss associated with elevated CRP levels and white blood cell count (Figures 1C and 1D), with a neutrophilia showing toxic granulation and a left shift. He would lose his appetite, lose weight, and become dehydrated, which required admission to the high dependency unit on some occasions. These episodes appeared to resolve with an increase in the dose of prednisolone. In addition, he seemed susceptible to frequent infections especially viral illnesses including varicella zoster virus (VZV), influenza A, respiratory syncytial virus (RSV), adenovirus, and cytomegalovirus (CMV), although these coincided with immunosuppressive therapy. He had poor weight gain, slow linear growth, developmental delay, mild learning difficulties, congenital hydroceles, dental caries, a pathological osteoporotic tibial fracture and he developed juvenile cortical cataracts at 2-3 years of age, which were treated with a bilateral lensectomy and vitrectomy. There are no records of hepatosplenomegaly in this patient. On examination, his growth parameters were all less than the 0.4th centile. He had a prominent nodular rash in keeping with a flare of panniculitis. He had coarse hair with bushy eyebrows, slight hyper-telorism, broad nasal bridge, prominent nose, protruding normally formed ears and a prominent chin. He had comprehensive immunological investigations which were normal including lymphocyte subset panel and proliferation analysis, a normal neutrophil oxidative burst, and normal expression of CD11a,b,c, CD18, CD55, CD95, MHCI, MHCII, and CD25. He had increased IgA and IgM levels (Table S1) in serum and a single strongly positive anti-smooth muscle antibody (SMA; 1:320 titer), but no other autoantibodies. He does not have alpha-1 antitrypsin deficiency or pancreatitis. The most recent skin biopsy demonstrated inflammatory infiltrates composed of lymphocytes and neutrophils within the subcutis, associated with foci of fat necrosis where neutrophils are particularly prominent. The inflammation seemed predominantly septal, but with extension into the lobules. Some of the septal blood vessels showed edema of their walls, but no frank vasculitis was identified. The features were very similar to those seen in the biopsy from his cousin (patient IV:4) and suggested a neutrophil-rich panniculitis with fat necrosis favoring septal distribution. He has been treated both with systemic steroids and Anakinra (recombinant IL-1R-antagonist). Neither medication successfully prevented the exacerbations or additional symptoms. Infliximab (TNF neutralizing antibody) was introduced eight years ago (at age ∼3) and has successfully controlled the disease. He has had a couple of minor exacerbations when the frequency of the Infliximab treatments was decreased, or the dose per kg fell to nearly 5 mg/kg. In addition to this, he takes prophylactic methotrexate, azithromycin, and aciclovir. Mice All animal experiments were undertaken with the approval of the UK Home Office. All mice were on a C57BL/6 or B6.SJL background and maintained under specific pathogen-free conditions in individually ventilated cages (Techniplast GM500, Techniplast) on Lignocel FS14 spruce bedding (IPS, Ltd.) with environmental enrichment (fun tunnel, chew stick, and Enviro-Dri nesting material (LBS)) at 19-23°C with light from 7.00 a.m. to 7.00 p.m. and fed Dietex CRM pellets (Special Diet Services) ad libitum. Otulin/Fam105b-targeted JM8A3.N1 (C57BL/6 background strain) ES cells (Fam105btm1a(EUCOMM)Hmgu) were obtained from EUCOMM and used to generate mice bearing neomycin selection and LacZ cassettes and a loxP-flanked exon 3 of Otulin, termed the targeted “LacZ“ or “LZ” (figures) allele (see also Figure S1). Heterozygous Otulin+/LacZ mice were normal and asymptomatic (data not shown), but intercrossing of Otulin+/LacZ mice failed to generate viable OtulinLacZ/LacZ offspring, consistent with previous work (Bonizzi and Karin, 2004, Rivkin et al., 2013). The neomycin selection and LacZ cassettes were removed from the targeted allele by intercrossing with FLPe-recombinase expressing mice (Rodríguez et al., 2000) to generate the conditional, floxed allele, which can be conditionally ablated by the Cre recombinase leading to a frameshift and a premature stop codon (p.Glu77Glyfs∗Ter3) (Figures S2A and S2B), and the Otulinflox strain (Otulin+/flox and OtulinLacZ/flox mice). Otulinflox mice were bred with Rosa26-Cre-ERT2 (Hameyer et al., 2007) (CreERT2-Otulinflox), LysM-Cre (Clausen et al., 1999) (LysMCre-Otulinflox), CD4-Cre (Lee et al., 2001) (CD4Cre-Otulinflox), and Mb1-Cre (Hobeika et al., 2006) (MB1Cre-Otulinflox) mice to facilitate conditional deletion of Otulin. Pilot experiments showed that CreERT2-OtulinLacZ/flox mice reacted adversely to tamoxifen administration, becoming moribund within a day of Otulin deletion (data not shown). In individual experiments, mice were matched for age and background strain. All experiments, except for embryo staining (see below) were performed on adult mice. For ERT2Cre-Otulinflox bone marrow chimeric mice, recipient mice were 2-3 months old and in each experiment were either all male or all female. No differences in results were observed between experiments performed on male and female mice. For LysMCre-Otulinflox, CD4Cre-Otulinflox, MB1Cre-Otulinflox mice, experiments were performed with a mix of male and female mice in experimental groups. LysMCre-Otulinflox, CD4Cre-Otulinflox, MB1Cre-Otulinflox mice used in experiments were 2-9 months old. Mice were allocated to experimental groups based on genotype. Grouping of mice into cages was determined at weaning, but where possible animals of equivalent age and gender were allocated to each experimental group. Where multiple groups of the same genotype were required, these were allocated randomly to the particular treatment conditions. Sample sizes were estimated using pilot experiments. Bone Marrow-Derived Macrophages Bone marrow-derived macrophages (BMDMs) were generated from bone marrow cells derived from tibias, femurs, and pelvic bones from LysMCre-Otulin+/flox, LysMCre-OtulinLacZ/flox or LysMCre-Otulindel/flox mice. Bones were flushed in PBS supplemented with 3% (v/v) FCS and cells were cultured and differentiated in RPMI 1640 + GlutaMAX supplemented with 10% (v/v) FCS, Penicillin/Streptomycin, 5 μM β-mercapto ethanol, nonessential amino acids (GIBCO), and 20 ng/mL recombinant mouse M-CSF (R&D Systems, Minneapolis, MN) as described previously (Damgaard et al., 2012). BMDMs denoted ‘untreated’ in experiments indicates no exogenous stimulation of these cells after differentiation. Cell Lines HEK293 cells and mouse embryonic fibroblasts (MEFs) were cultured in DMEM + GlutaMAX supplemented with 10% (v/v) FCS and Penicillin/Streptomycin at 37°C in a humidified atmosphere at 5% CO2 unless otherwise indicated in figure legends or method details. Method Details Molecular Genetic Analysis Blood genomic DNA was isolated using the DNeasy kit (QIAGEN). A genome-wide linkage scan was carried out using the Affymetrix 250K SNP microarray with DNA from the three affected patients (V:2, IV:3 and IV:4). Homozygous regions were identified in affected patients using the program HomozygosityMapper (http://www.homozygositymapper.org/) (Seelow et al., 2009) and further analyzed to confirm or refute linkage by typing microsatellite markers in all family members from whom DNA was available. Direct sequencing of the genes within the identified region was prioritised according to putative function and position. The genomic DNA sequence of candidate genes was taken from Ensembl (http://www.ensembl.org/index.html) and primer pairs for the translated exons were designed using ExonPrimer software (https://ihg.gsf.de/ihg/ExonPrimer.html). Individual exons and flanking sequences were amplified using standard polymerase chain reaction (PCR) (primer details and conditions on request). PCR products were directly sequenced by the Big Dye Terminator Cycle Sequencing kit and run on an ABI PRISM 3730 DNA Analyzer (Applied Biosystem). Sanger sequencing was repeated on at least two independent PCR products to confirm sequence variants. DNA sequences were analyzed using the Chromas software (Technelysium). Whole Exome Sequencing was performed as described previously (Clark et al., 2014). Exon capture was performed with the SureSelect All Exon 50Mb Target Enrichment System (Agilent) and massively parallel DNA sequencing was undertaken on an Illumina AnalyserIIx with 76bp paired end reads. Single nucleotide substitutions and small insertion deletions were detected and quality filtered within the SamTools software package (http://samtools.sourceforge.net/) (Li et al., 2009) and in-house software tools. Variants were annotated with the Annovar tool (http://wannovar.usc.edu/index.php) (Yang and Wang, 2015). Filtering of variants for novelty was performed by comparison to dbSNP132 and 1000 Genomes SNP calls (June 2011) and patient variants were compared to variants identified in 250 control exomes sequenced and analyzed in a similar manner. The Exome Variant Server (http://evs.gs.washington.edu/EVS/) and the EXaC Browser (http://exac.broadinstitute.org/gene/ENSG00000154124) datasets were accessed October 2015. β-Galactosidase Staining of Mouse Embryos E13.5 Otulin+/LacZ and Otulin+/+ embryos were dissected from the uterine tracts and immediately transferred to ice cold 4% PFA in PBS and left to fix for 1 hr. Fixed embryos were washed in rinse buffer (5 mM EGTA, 0.01% (w/v) deoxycholate, 0.02% NP-40 (v/v), 2 mM MgCl2 in PBS) and transferred to staining buffer (5 mM K3[Fe(CN)6], 5 mM K4[Fe(CN)6], 5 mM EGTA, 0.01% deoxycholate (w/v), 0.02% NP-40 (v/v), 2 mM MgCl2, 1 mg/mL 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal) in PBS) and left to incubate 24 hr in the dark at 37°C. After staining, embryos were washed in rinse buffer and post-fixed in 4% PFA in PBS for 48 hr at room temperature. Embryos were then washed in PBS and dehydrated by sequential washes in increasing concentrations of ethanol ending with 100% ethanol for 30 min and then cleared in methyl salicylate (Sigma Aldrich) for 30 min. Embryos were allocated to experimental groups based on genotype. This experiment was performed on embryos from two independent litters, and a total of five Otulin+/+ and five Otulin+/LacZ embryos were analyzed. Mixed Bone Marrow Chimeras One week before transplantation, the drinking water of recipient mice was supplemented with 0.1 mg/mL enrofloxacin (Baytril®, Bayer). Bone marrow cells (2 × 106) from wild-type CD45.1+ B6/SJL mice and CD45.2+ CreERT2-Otulin+/flox or CreERT2-OtulinLacZ/flox mice were resuspended in 100 μL PBS at a 1:1 ratio and then injected intravenously (i.v.) into 2-3 months old sex- and age-matched γ-irradiated C57BL/6 recipients (given two doses of 4.5 Gy). At 6-8 weeks after reconstitution, mice were given three doses of tamoxifen (Sigma-Aldrich, St Louis, MO; 1 mg in sunflower oil with 10% ethanol per dose) intraperitoneally (i.p.). Mice were closely monitored and weighed daily. At the onset of weight loss (3-6 days after initial tamoxifen dose) mice were culled and samples taken for analyses. Mixed bone marrow chimeric mice were allocated to experimental groups based on genotype of the transplanted bone marrow. Where multiple groups of the same genotype were required, these were allocated randomly to the particular treatment conditions. In Vivo Cytokine Neutralization For in vivo neutralization of cytokines, mixed bone marrow chimeric mice (described above) were injected i.v. with antibodies two to four hours before tamoxifen injection. Chimeric mice were injected with 1 mg/mouse InVivoMab anti-mouse TNF (clone XT3.11, BE0058, BioXcell, West Lebanon, NH), 1 mg/mouse InVivoMab anti-mouse IL-6 (clone MP5-20F3, BE0046, BioXCell), 250 μg/mouse anti-mouse G-CSF (clone 67604, MAB414, R&D Systems), or the equivalent amount of InVivoMab Rat IgG1 isotype control (clone HRPN, BE0088, BioXCell). Mixed bone marrow chimeric mice were allocated to the experimental groups based on genotype of the transplanted bone marrow. Where multiple groups of the same genotype were required, these were allocated randomly to the particular treatment conditions. Cytokine/Chemokine, Autoantibody, and Immunoglobulin Analysis Cytokine/chemokine multiplex analysis was carried out using Luminex xMAP technology. Serum and medium samples were analyzed using magnetic MILLIPLEX MAP antibody-conjugated beads (Merck-Millipore, Bedford, MA) according to the manufacturer’s instructions on a Luminex MAGPIX instrument with the xPONENT 4.2 software. For some G-CSF measurements, samples were diluted 1/10. For cytokine measurements from BMDMs, cells were split and equal numbers were reseeded in fresh medium 24 hr prior to sample collection. BAFF levels, and G-CSF levels in some experiments, were determined using the Mouse BAFF/BLyS/TNFSF13B Quantikine ELISA or the Mouse G-CSF Quantikine ELISA kits (R&D Systems), respectively, with a 1/20 dilution of samples for BAFF and 1/10 for G-CSF according to the manufacturer’s instructions. Autoantibodies were detected using mouse anti-nuclear antigen/ENA IgA+G+M, mouse anti-Smith antigen IgA+G+M, and mouse anti-dsDNA IgA+G+M ELISA kits (Alpha Diagnostics, San Antonio, TX) with 1/100 dilution of samples according to the manufacturers instructions. All samples were run in duplicate. Serum total IgG levels were determined using the Mouse Total IgG Ready-Set-Go ELISA kit (eBioscience, San Diego, CA) or the IgG Mouse ELISA kit (Abcam, Cambridge, UK) according to the manufacturer’s instructions. Serum immunoglobulin isotyping was performed using the Mouse Isotyping Panel 1 multiplex sandwich electrochemiluminescence immunoassay kit (Meso Scale Discovery, Gaithersburgh, MD) with 1/100,000 dilution of samples according to the manufacturer’s instructions. All samples were run in duplicate. Patient C-reactive protein (CRP) were analyzed on an Abbott Archetect C1600. Histology Tissue samples were harvested in 3% (v/v) FCS in PBS on ice, transferred to 10% neutral buffered formalin, and fixed for 48 hr at room temperature. Fixed tissues were sent to AML Laboratories, Inc., Baltimore, MD, for paraffin embedding, sectioning, and Hematoxylin & Eosin or Masson’s Trichrome staining. Micrographs were taken on a Carl Zeiss Axioplan microscope with an Axiocam camera and processed using the Fiji software (Schindelin et al., 2012). Scale bars represent 200 μm. Flow Cytometry and Cell Sorting Single cells in a solution of 2% (v/v) FCS in PBS at 4°C were then blocked with anti-CD32 antibodies (clone 2.4G2; BioXCell, West Lebanon, NH), washed, and then incubated with fluorophore-conjugated antibodies. Fixable Viability Dye eFluor780 (eBioscience, San Diego, CA) was included in all analyses to exclude dead cells. Cells were analyzed on a BD Fortessa (BD Bioscience, Oxford, United Kingdom) or were sorted with a MoFlo Synergy cell sorter (Beckman Coulter, Inc., Fullerton, CA) according to the manufacturers’ standard operating procedures. Data were analyzed with FlowJo software version X.07. For quantification of cells numbers of infiltrating cells, the absolute number of cells in a gate was multiplied by the dilution factor of the sample and corrected for the percentage of dead cells. Fluorophore-coupled antibodies were anti-CD4 conjugated to Alexafluor700 or Alexafluor647 (clone GK1.5, eBioscience), anti-CD4 conjugated to FITC (clone H129.19, BD Pharmigen), anti-CD4 conjugated to BrilliantViolet785 (clone RM4-5, BioLegend), anti-CD8α conjugated to FITC, APC, PE-Cy7 or PE (clone 53-6.7, eBioscience), anti-CD11b conjugated to PE-Cy7 or eFluor-450 (clone M1/70, eBioscience), anti-CD11b conjugated to BrilliantViolet-421 or FITC (clone M1/70, BioLegend), anti-CD11c conjugated to PE (clone N418, BioLegend), anti-CD11c conjugated to FITC (clone N418, eBioscience), anti-CD19 conjugated to FITC or PerCP-Cy5.5 (clone 1D3, eBioscience), anti-CD19 conjugated to Alexa Fluor-700 (clone 6D5, BioLegend), anti-CD45.1 conjugated to eFluor-450 (clone A20, eBioscience), anti-CD45.1 conjugated to BrilliantViolet-510 (clone A20, BioLegend), anti-CD45.2 conjugated to Alexa Fluor-700 (clone 104, eBioscience), anti-Ly6G/Gr-1 conjugated to FITC, PE, or APC (clone RB6-8C5, eBioscience), anti-NK1.1 conjugated to BrilliantViolet-421 or FITC (clone PK136, BioLegend), anti-Ter-119 conjugated to FITC (clone TER-119, eBioscience), anti-FcεRI conjugated to FITC (clone MAR-1, BioLegend), anti-CD43 conjugated to PE (clone S7, BD Pharmigen), anti-CD25 conjugated to BrilliantViolet-421 (clone PC61, BioLegend), anti-CD23 conjugated to PE-Cy7 (clone B3B4, eBioscience), anti-CD21 conjugated to APC (clone 7E9, BioLegend), anti-IgM conjugated to PerCP-Cy5.5 (clone RMM-1, BioLegend), anti-IgD conjugated to BrilliantViolet-650 (clone 11-26c.2a, BioLegend), anti-CD5 conjugated to BrilliantViolet-510 (clone 53-7.3, BioLegend), anti-Sca1 conjugated to PE-Cy7 (clone D7, eBioscience), anti-cKit/CD117 conjugated to PerCP-Cy5.5 (clone 2B8, BioLegend), anti-B220 conjugated to FITC (clone RA3-6B2, eBioscience), anti-CD3 conjugated to FITC or PE-Cy7 (clone 145-2C11, BioLegend), anti-CD3 conjugated to APC (clone 145-2C11, eBioscience), anti-CD5 conjugated to FITC (clone 53-7.3, BD Pharmigen), anti-CD44 conjugated to PE (clone IM7, eBioscience), and anti-CD62L conjugated to BrilliantViolet-421 (clone MEL-14, BioLegend). Lineage (Lin) stain panel consisted of anti-B220, anti-CD19, anti-CD3, anti-CD4, anti-CD5, anti-CD8, anti-CD11b, anti-CD11c, anti-NK1.1, anti-FcεRI, and anti-Ter119 antibodies conjugated to FITC. Bone Marrow-Derived Macrophages Bone marrow-derived macrophages (BMDMs) were considered “untreated” when no exogenous stimulation was given after completed differentiation (day 7). For assessment of IκBα stability, BMDMs were washed thoroughly in PBS and then incubated with 10 μg/mL anti-TNF neutralizing antibodies (clone XT3.11, BE0058, BioXCell) or isotype control (clone HPRN, BE0088, BioXcell) in complete RPMI for 30 min before they were treated with 50 μg/mL cycloheximide in DMSO (CHX; Santa Cruz Biotechnology) for the indicated times. To verify TNF neutralization, immortalized mouse embryonic fibroblasts (MEFs) were cultured in medium containing the TNF neutralizing antibody or the isotype control and stimulated with 0.25 ng/mL recombinant mouse TNF (GIBCO) for the indicated times. For signaling experiments, BMDMs were stimulated with 1 ng/mL recombinant mouse TNF (GIBCO) for the indicated times before they were lysed for quantitative RT-PCR analysis. Tissue Preparation and MACS Cell Separation Single cell suspensions were obtained by gentle mechanical disruption of tissues using a syringe plunger and a 70 μm cell strainer. Lungs were digested for 30 min with collagenase D (Roche; 0.7 mg/mL in 2% (v/v) FCS in PBS) before disruption. For CD11b+ cell isolation from livers for immunoblotting (Figure 7C), mashed livers were resuspended in 44% Percoll PLUS (GE Healthcare) in PBS and centrifuged at 2000 rpm for 20 min with low deceleration. The top layer of hepatocytes and Percoll were then aspirated, and the pelleted leukocytes were resuspended in 4.5 mL ice-cold PBS, vortexed for 15 s, and then added 500 μL 10x PBS and vortexed again. The remaining leukocytes were pelleted for centrifugation and subjected to isolation using MACS MicroBeads (see below). For lamina propria leukocyte preparations from small intestine and colon (for Figure 7C), the mesentery was removed, the intestinal content gently squeezed out, and the intestines were washed in PBS/HEPES (PBS + 10 mM HEPES) and then cut into 2-3 cm pieces. Digestion and cell isolation was then performed using 0.06 mg/mL Liberase (Roche) and 60 μg/mL DNase I (Sigma Aldrich) as previously described (Morrison et al., 2013). Blood leukocytes were prepared from whole EDTA-blood by diluting 500 μL blood in 8.5 mL ice-cold distilled water and vortexing for 15 s. Immediately thereafter 1 mL 10x PBS was added the preparation was vortexed for 10 s and the blood leukocytes were pelleted by centrifugation. The remaining leukocytes were pelleted for centrifugation and subjected to isolation using MACS MicroBeads (see below). All cell preparations and suspensions were passed through 70 μm nylon cell strainers again before cell isolation. CD19+ B cells were isolated from spleen preparations using anti-mouse CD19-coated magnetic MACS MicroBeads (Miltenyi Biotec, Bergisch Gladbach, Germany), CD4+ and CD8+ T cells were isolated from spleen preparations using anti-biotin-coated magnetic MACS MicroBeads (Miltenyi Biotec) incubated with anti-CD4 conjugated to biotin (clone GK1.5, BioLegend) and anti-CD8α conjugated to biotin (clone 53-6.7, eBioscience), and CD11b+ myeloid cells were isolated from various tissue preparations using anti-mouse CD11b-coated magnetic MACS® MicroBeads (Miltenyi Biotec). All MACS beads-based isolations were performed according to the manufacturer’s instructions using MACS MS Cell Separation Columns (Miltenyi Biotec). Isolated cells were either lysed in sample buffer (50 mM Tris pH 6.8, 10% glycerol (v/v), 100 mM DTT, 2% SDS (w/v), bromophenol blue) immediately after isolation or cultured in RPMI-1640 + GlutaMAX supplemented with 10% (v/v) FCS and Penicillin/Streptomycin and treated with 10 μM MG132 in DMSO (Sigma Aldrich) as indicated, washed in PBS, and then lysed in sample buffer. Quantitative Real-Time PCR Equal numbers of BMDMs were seeded in 12-well plates 24 hr prior to experiments. Total RNA was extracted from BMDMs using RNeasy Mini Kit (QIAGEN, Hilden, Germany), and DNase digestion was performed on column with the RNase-Free DNase Set (QIAGEN, Hilden, Germany) according to the manufacturer’s protocol. Total RNA was reverse transcribed using Quantitect Reverse Transcription Kit (QIAGEN). RT-PCR was performed using QuantiFast SYBR Green RT-PCR Kit (QIAGEN) on a Viia7 Real-Time PCR Instrument (Applied Biosystems, Foster City, CA) with the primers indicated below. Each sample was run in duplicate. Results were normalized to those of 18S rRNA as internal housekeeping control using the 2∧(-ΔCt)-method. Primers were: 18S rRNA 5′-GTAACCCGTTGA-ACCCCATT-3′ and 5′-CCATCCAATCGGTAGTAGCG-3′; Tnf 5′-GGTCTGGGCCATAGAACTGA-3′ and 5′-CAGCCTCTTCTCATTCCTGC-3′; Il6 5′-TCTGAAGGACTCTGGCTTTG-3′ and 5′-GATGGATGCTACCAAACTGGA-3′; Nfkbia 5′-CCAAGTGCAGGAACGAGTCT-3′ and 5′-AAGGACGAGGAGTACGAGCA-3′; Tnfaip3 5′-TTCCTCAGGACCAGGTCAGT-3′ and 5′-AAGCTCGTGGCTCTGAAAAC-3′; Hoip/Rnf31 5′-TACGGTTGTATGGCTATA-3′ and 5′-GTATTCATCTGGTTCCTC-3′; Hoil-1/Rbck1 5′-GCACTTTCATCAACAAAC-3′ and 5′-AGGTATCTGGTAGGTCTC-3′; Sharpin 5′- GAACTGGTATTGTCTTGTGTA-3′ and 5′-AGAAGGCAAGGATGAACT-3′. Immunoblotting For BMDMs, equal numbers of cells were lysed directly in sample buffer (50 mM Tris pH 6.8, 10% glycerol (v/v), 100 mM DTT, 2% SDS (w/v), bromophenol blue), sonicated for 5 s (microtip) and boiled for 2 min. Mouse tissues were lysed in RIPA buffer (50 mM Tris pH 7.4, 1% NP-40 (v/v), 0.5% deoxycholate (w/v), 0.1% SDS (w/v), 150 mM NaCl, 2 mM EDTA, 5 mM MgCl2) for 15 min at 20 Hz on a TisssueLyser II (QIAGEN). Samples were then treated with Benzonase (Novagen, Madison, WI) for 30 min at 4°C for and then sonicated for 5 min in 10 s pulses. Samples were cleared by centrifugation and protein concentration was measured using BCA assay (Thermo Scientific). Proteins were resolved on 4%–12% Bis-Tris gels (Invitrogen, Carlsbad, CA) and transferred to nitrocellulose membranes using the iBlot system (Invitrogen), a HEP-1 Owl Panther semidry electroblotter (Thermo Scientific), or the Trans-Blot Turbo transfer system (Bio-Rad, Hercules, CA). Membranes were blocked in 5% milk in PBS-T (PBS + 0.1% (v/v) Tween-20) for 30 min and incubated with primary antibodies in PBS-T + 3% BSA at 4°C overnight, washed in PBS-T, incubated at room temperature for 1 hr with anti-rabbit IgG-HRP or anti-mouse IgG-HRP, washed, and visualized using Amersham Western Blotting Detection Reagent (GE Healthcare) or SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Scientiric). Primary antibodies recognized OTULIN (14127, Cell Signaling Technology), IκBα (9242, Cell Signaling Technology), phospho-p65/RelA (pS563) (3033, Cell Signaling Technology), p65/RelA (8242, Cell Signaling Technology), A20 (4625, Cell Signaling Technology), NEMO/IKKγ (8330, Santa Cruz Biotechnology), GAPDH (AM4300, Ambion), Met1-linked/linear ubiquitin (LUB9, AB130, LifeSensors), Met1-linked/linear ubiquitin (1E3, MABS199, Millipore), Lys63-linked ubiquitin (Apu3, 05-1308, Millipore), Lys48-linked ubiquitin (Apu2, 05-1307, Millipore), Ubiquitin (NB300-130, Novus Biologicals), HOIP/RNF31 (46322, Abcam), mouse HOIP (kind gift from Professor Kazuhiro Iwai, Kyoto University (Tokunaga et al., 2011)), HOIL-1/RBCK1 (NBP2-27105, Novus Biologicals), HOIL-1/RBCK1 (sc-49718, Santa Cruz Biotechnology), SHARPIN (14626-1-AP, Proteintech), NIK (sc-8417, Santa Cryz Biotechnology), phospho-IRF-3 (pS396) (4947, Cell Signaling Technology), IRF-3 (sc-9082, Santa Cruz Biotechnology), CYLD (sc-74435, Santa Cruz Biotechnology), p38 (M138, ab31828, Abcam), phospho-p38 (pT180/pY182) (ERP18120, ab195049, Abcam), and Actin (C4, MAB1501R, Millipore). Secondary HRP-coupled antibodies were from GE Healthcare (NA931 and NA934). Densitometry analyses of immunoblots were performed using the Fiji software (Schindelin et al., 2012). To analyze IκBα stability, the ratio between the absolute IκBα signal intensity for each sample and the absolute Actin signal intensity for each samples was calculated. This ratio was then normalized to the “0” sample within each group. Immunoblot data are representative of at least two independent experiments. MTT Reduction Assay for Cell Viability MTT reduction assay was performed as previously described (Damgaard et al., 2013). Briefly, equal numbers of LysMCre-Otulin+/flox and LysMCre-OtulinLacZ/flox BMDMs were seeded (as two biological replicas) in 96-well plates 24 hr before treatment. Cells were treated as indicated with recombinant human TNF (100 ng/mL; R&D Systems), ultrapure LPS from E. coli K12 (100 ng/mL; Invivogen, San Diego, CA), or Staurosporine (1 μM; Sigma-Aldrich) for 10 hr. Medium was aspirated and 100 μL fresh medium was added together with 25 μL 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliym bromide (MTT; 5 mg/mL dissolved in PBS (Sigma)) and cells were left to incubate for 2 hr at 37°C. Afterward, 100 μL solubilization buffer (20% SDS (w/v) dissolved in 50% N,N-dimethylformamide) was added and samples were left to incubate overnight. Absorbance at 590 nm was read with a reference filter of 620 nm. Individual experiments were performed in duplicate. Data were normalized to untreated LysMCre-Otulin+/flox samples. Kaplan-Meier Curve The Kaplan-Meier curve for the LysMCre-Otulin mice were produced using Graphpad Prism 6. Cell Culture and Pull-Downs HEK293 cells were transfected with pEGFP-N1-OTULIN-WT or pEGFP-N1-OTULIN-L272P, encoding human OTULIN and its variant, using FuGENE HD (Promega, Madison, WI). Medium was changed 16 hr after transfection, and cells were then kept at 32°C for another 24 hr. Cell were lysed in 25 mM Tris-HCl pH 7.4, 150 mM NaCl, 1 mM DTT, 0.5% NP40 and protease inhibitor cocktail (Roche) on ice for 20 min. Lysate was cleared by centrifugation at 16,000 x g for 20 min, and lysate was incubated with agarose-coupled GFP-Trap A beads (ChromoTek GmbH, Martinsried, Germany) for 90 min at 4°C on rotation to precipitate GFP-tagged OTULIN. Bound GFP-OTULIN was eluted by boiling beads in 2x sample buffer, and the precipitated proteins were resolved and analyzed by SDS-PAGE and immunoblotting. Patient Cell Samples Buffy coat cells from blood samples from patient V:2 and anonymized healthy, age-matched controls were depleted of dead cells using the Dead Cell Removal kit (Miltenyi Biotec) according to the manufacturer’s instructions. Live cells were washed in PBS, pelleted by centrifugation, and lysed in sample buffer (50 mM Tris pH 6.8, 10% glycerol (v/v), 100 mM DTT, 2% SDS (w/v), bromophenol blue) before they were subjected to analysis by immunoblot. This experiment was repeated with two independent samples from patient V:2. Expression and Purification of OTULIN Human OTULIN constructs were expressed in E. coli strain Rosetta2 (DE3) pLacI. Cells were grown at 37°C in 2 xTY medium containing 30 mg/mL kanamycin and 34 mg/mL chloramphenicol to an OD600 of 0.8. The cultures were cooled down to 18°C before induction with 400 μM IPTG and harvested 20 hr post induction. Cells were resuspended and lysed by sonication in lysis buffer (20 mM Tris pH 7.4, 300 nM NaCl, 2 mM β-mercaptoethanol, 40 mM imidazole, DNase I, lysozyme, protease inhibitor cocktail (Roche)). OTULIN was purified by immobilized metal affinity chromatography using a HisTrap HP column (GE Healthcare Life Sciences). The 6-His tag was cleaved by overnight incubation with 3C protease, in a dialysis buffer (20 mM Tris pH 8.0, 4 mM DTT). The protein was further purified by anion exchange chromatography (ResourceQ, GE Healthcare Life Sciences) and the eluted OTULIN was subjected to size exclusion chromatography (HiLoad 16/60 Superdex 75, GE Healthcare Life Sciences) in buffer containing 20 mM Tris pH 8.0, 175 mM NaCl, 4 mM DTT. Qualitative DUB Linkage Specificity Assay Qualitative deubiquitination assays were performed as previously described (Keusekotten et al., 2013). Briefly, OTULINWT and OTULINL272P at several concentrations were diluted in 25 mM Tris pH 7.4, 150 mM NaCl, and 10 mM DTT and incubated with 1 μM di- or tetraUb in DUB buffer (50 mM Tris pH 7.4, 50 mM NaCl, 5 mM DTT) at 37°C. Samples were taken at different time points and mixed with 4 x SDS sample buffer to stop the reaction. They were resolved by SDS-PAGE and visualized by silver staining (BioRad SilverStain Plus kit). Qualitative DUB assay on diUb were repeated three times, and the assay were performed once on tetraUb. Binding Studies To measure binding affinities of OTULINC129A and OTULINC129A/L272P to Met1-diUb, fluorescence anisotropy experiments were performed as previously described (Keusekotten et al., 2013). In brief, 10 μL of 100 nM FlAsH-tagged Met1-diUb were dispensed in a 384-well Corning plate. Serial dilutions of OTULINC129A and OTULINC129/L272P were prepared in FlAsH buffer (20 mM Tris pH 7.4, 150 mM NaCl, 2 mM β-mercaptoethanol, 0.1 mg/mL bovine serum albumin) and 10 μL were added to FlAsH-tagged Met1-diUb containing wells. Fluorescence polarization was recorded on a PheraStar plate reader (BMG Labtech) using an optics module with λex = 485 nm and λem = 520 nm, and values were fitted to a one-site total binding model using Graphpad Prism 6 to derive binding constants (KD). Circular Dichroism Experiments Circular dichroism (CD) measurements were performed on a Jasco-J815 spectropolarimeter (Jasco International Co. Ltd., Tokyo, Japan), using a spectral band width of 1 nm and a cell path-length of 0.1 cm. OTULINWT and OTULINL272P were re-buffered in PBS at 0.38 mg/mL prior to the experiment. The far UV CD spectra were measured from 260-195 nm and were corrected for buffer contribution by baseline subtraction. Nano-DSF Thermal Unfolding Experiments Nano-DSF (differential scanning fluorimetry) measurements were performed using a Prometheus NT.48 instrument (NanoTemper Technologies GmbH, Germany). Experiments on OTULINWT and OTULINL272P were carried out at 0.11 mg/mL and 0.38 mg/mL (not shown). Samples were dialysed into PBS before measurement and 10 μL of each sample were loaded in UV capillaries (NanoTemper Technologies). Temperature gradient was set at 2.5°C/min in a range from 20 to 90°C. Protein unfolding was measured by detection of change in tryptophan fluorescence at emission wavelengths of 330 and 350 nm, dependent on temperature gradient. Tm’s were calculated according to the manufacturer’s instructions. Results were confirmed by differential scanning calorimetry and CD thermal melt experiments (data not shown). Quantification and Statistical Analysis Data are presented as mean ± SEM unless otherwise indicated in figure legends. Sample number (n) indicates the number of independent biological samples in each experiment. Sample numbers and experimental repeats are indicated in figures and figure legends or methods section above. Data were analyzed using the two-sided nonparametric Mann-Whitney U test of the null hypothesis of continuous data unless otherwise indicated in figure legends or method details. Data analysis was not blinded. Differences in means were considered statistically significant at p < 0.05. Significance levels are: ∗ p < 0.05; ∗∗ p < 0.01; ∗∗∗ p < 0.001; ∗∗∗∗ p < 0.0001; n.s., non-significant. Analyses were performed using the Graphpad Prism 6.0 software. Data and Software Availability Data Resources Due to restrictions from the patient consent approved by the research ethics committee it is not be possible to deposit complete exome sequencing data in a public repository, but the data could be made available to interested researchers by contacting the authors. Author Contributions D.K., A.N.J.M., and E.R.M. conceived the project. R.B.D., J.W., A.N.J.M and D.K. designed in vivo experiments, and R.B.D. performed and analyzed the experiments assisted by J.W. R.B.D. designed and performed the in vitro experiments and analyzed the data. E.R.M., D.M., N.V.M., and H.L.T. were responsible for human clinical and molecular genetic studies. P.M.-C. performed and analyzed the biochemical and biophysical experiments, to which also P.R.E. contributed. R.B.D. and D.K. wrote the paper with input from all authors. Supplemental Information Document S1. Tables S1 and S3–S8 Table S2. Rare Homozygous Variants in ORAS Patients, Related to Figure 1 Details of rare homozygous variants (allele frequency < 0.01) identified by exome sequencing in an affected individual. The three variants within the linked autozygous region on chromosome 5p15 are shaded. Acknowledgments We would like to thank The Wellcome Trust Sanger Institute, the EUCOMM Consortium, and the Mutant Mouse and Resource & Research Centers (MMRRC) for Otulin-targeted ES cells. We would also like to thank Helen Jolin for help with mouse colony management and genotyping; Richard Pannell, Angela Middleton, Richard Berks, Xinping Gong, and Ares and Biomed staff for assistance with animal procedures and experiments; Kirstin Keusekotten for help with initial stages of the project; Flow Cytometry Core staff for technical assistance; Louise Tee for assistance with human genetic studies; the late Louise Brueton for clinical genetic assessments; Chris Johnson for assistance with biophysics techniques; Juan Garaycoechea for providing MEFs; Neil Grant for help with photographic images; and Menna Clatworthy for helpful discussions. This work was supported by the Medical Research Council (U105192732 and U105178805), the European Research Council (309756), the Lister Institute for Preventive Medicine, and the EMBO Young Investigator Program (to D.K.); a Marie-Sklodowska Curie Individual Fellowship from the European Commission (MC-IF-654019) and a Research Fellowship from Corpus Christi College Cambridge (to R.B.D.); and Wellcome Trust (to N.V.M., E.R.M., and A.N.J.M [100963/Z/13/Z]), WellChild (to N.V.M. and E.R.M.), and UCB (to H.L.T., D.M. and E.R.M.). Figures 1E, 2B, 3K, and 7H were produced using Servier Medical Art (www.servier.com). D.K. is part of the DUB Alliance that includes Cancer Research Technology and FORMA Therapeutics. D.M. is an employee of UCB. Supplemental Information includes six figures and eight tables and can be found with this article online at http://dx.doi.org/10.1016/j.cell.2016.07.019. Figure 1 Mutations in OTULIN in Patients with a Systemic Autoinflammatory Syndrome (A) Segregation of the inflammatory symptoms (filled symbols) and the c.815T>C substitution in OTULIN in the affected kindred. ○, females; □, males; double lines, consanguineous relationship; crossed symbols, deceased individuals; Δ, miscarriage; ◊, stillbirths. Roman numerals indicate generations. (B and C) Lifetime measurements of (B) C-reactive protein (CRP) serum concentrations and (C) white blood cell (WBC, black line) and neutrophil numbers (cyan line) in blood from patients IV:3, IV:4, and V:2. Reference ranges (dotted lines) are indicated on the graphs. Patient V:2 was treated with Infliximab as indicated (orange shade). (D) OTULIN DNA sequence chromatograms identifying the homozygous single-base substitution (c.815 T>C, p.Leu272Pro, arrowhead). (E) Schematic of the cardinal symptoms of OTULIN-related autoinflammatory syndrome (ORAS). The efficacy of trialed treatments are indicated in the table. (F) Superimposed structures of OTULIN’s catalytic domain (blue) without substrate and bound to Met1 diUb (green [only distal Ub shown]; PDB: 3znv and PDB: 3znz [Keusekotten et al., 2013]), showing the position of Leu272 in the distal Ub binding site. (G) Met1-linked diUb hydrolysis by OTULINWT and OTULINL272P. (H) Immunoblot showing the levels of Met1-linked polyUb, total Ub, OTULIN, and LUBAC in buffy coat cells from patient V:2. See also Figure S1 and Document S1. Tables S1 and S3–S8, Table S2. Rare Homozygous Variants in ORAS Patients, Related to Figure 1. Figure 2 Deletion of Otulin in Immune Cells Causes Acute Systemic Inflammation in Mice (A) OTULIN immunoblot on immune cells from wild-type mice. NK cell, natural killer cell; DC, dendritic cell; MΦ, macrophage. (B) Schematic representation of mixed bone marrow chimera generation. Wild-type (WT) B6.SJL cells are CD45.1+, and CreERT2-Otulinflox cells are CD45.2+. (C) Body weight following i.p. administration of tamoxifen (tx; arrows) to CreERT2-Otulinflox chimeric mice. (D) Neutrophil and lymphocyte counts from blood of CreERT2-Otulinflox chimeras and vehicle-treated controls at day 5 following tamoxifen administration. (E and F) Luminex multiplex analysis of serum cytokines and chemokines from terminal bleeds on day 5 presented as (E) a heatmap of relative changes in concentration of all analytes between CreERT2-Otulin+/flox and CreERT2-OtulinLacZ/flox chimeras and (F) serum concentrations of cytokines and chemokines increased in CreERT2-OtulinLacZ/flox chimeras. Data were pooled from two independent experiments. (G and H) Flow cytometry analysis of CD11b+Gr-1+ neutrophils in total cellular infiltrate (CD45.1+ and CD45.2+) in peritoneal lavage (PL), spleen, and liver from CreERT2-Otulinflox chimeras presented as (G) representative dot plots with percentage of cells in gate indicated and (H) total cell number. (I) Micrographs of hematoxylin and eosin (H&E) stained sections reveal inflammatory foci (arrowheads) in liver parenchyma. Micrographs are representative of 13 CreERT2-Otulin+/flox and 14 CreERT2-OtulinLacZ/flox chimeras from two independent experiments. Scale bars, 200 μm. (C, D, F, and H) Data are presented as mean ± SEM, and n represents number of mice. See also Figure S2 and Table S4. Figure 3 Neutralization of TNF Ameliorates Inflammation Caused by OTULIN Deficiency (A–J) Measurements from tamoxifen (tx)-treated (arrows) CreERT2-Otulinflox bone marrow chimeras injected with (A–C) anti-TNF neutralizing antibodies (αTNF) (data were pooled from two independent experiments), (D–F) anti-G-CSF-neutralizing antibodies (αG-CSF), (G–I) anti-IL-6-neutralizing antibodies (αIL-6), or isotype control as indicated. (A, D, and G) Body weight of CreERT2-Otulinflox chimeric mice treated with neutralizing antibodies as indicated. (B, E, and H) Blood neutrophil counts from CreERT2-Otulinflox chimeric mice treated as indicated. (C, F, and I) Total number of infiltrating CD11b+Gr-1+ neutrophils in spleen and peritoneal lavage (PL) measured by flow cytometry from CreERT2-Otulinflox chimeric mice treated as indicated. (J) Heatmap of Luminex multiplex analysis of cytokines and chemokines in serum from terminal bleeds on days 6 or 7 from chimeric mice treated as indicated. Numbers indicate relative change compared to isotype-treated del/flox mice within each experiment. G-CSF levels for αG-CSF and αIL-6 were measured by ELISA. Asterisks (∗) indicate the level of statistical significance. (K) Model of TNF-driven systemic inflammation and the contributions from different cytokines in OTULIN-deficient mice. (A–J) Data are presented as mean ± SEM, and n represents number of mice. See also Figure S3. Figure 4 Specific Deletion of Otulin in Myeloid Cells, but Not T or B Cells, Causes Systemic Inflammation (A, C, and E) Spleens and spleen weights, thymuses, and inguinal lymph nodes from (A) 2- to 3-months-old CD4Cre-Otulinflox mice (n = 6), (C) 3- to 4-months-old MB1Cre-Otulinflox mice (n = 4), (E) 3- to 9-months-old LysMCre-Otulinflox mice (n = 8). (B, D, and F) Blood cell counts from (B) 2- to 3-months-old CD4Cre-Otulinflox mice (n = 6); (D) 3- to 4-months-old MB1Cre-Otulinflox mice (n = 7); and (F) 3- to 9-months-old LysMCre-Otulinflox mice (n = 11). (G and H) Luminex multiplex analysis of serum cytokine and chemokine concentrations from terminal bleeds of 2- to 3-months-old CD4Cre-Otulinflox mice, 3- to 4-months-old MB1Cre-Otulinflox mice, and 4- to 9-months-old LysMCre-Otulinflox mice presented as (G) a heatmap of relative changes of analytes between OTULIN-deficient mice and their respective +/flox controls and (H) serum concentrations of selected cytokines and chemokines increased in LysMCre-OtulinLacZ/flox mice. (I) ELISA measurements of total IgG concentrations in serum from CD4Cre-Otulinflox (n = 6), MB1Cre-Otulinflox (n = 7), and LysMCre-Otulinflox (n = 14) mice. Data are presented as mean ± SEM, and n represents number of mice. See also Figure S4 and Tables S5–S7. Figure 5 OTULIN Deficiency in Myeloid Cells Leads to Neutrophil Infiltration, Multi-organ Inflammation, Hyper-Immunoglobulinemia, and Autoimmunity (A and B) Micrographs of H&E and Masson’s trichome stained sections of livers and spleens from 4- to 9-months-old LysMCre-Otulinflox mice reveal (A) infiltration in and fibrosis (right, arrowheads) of liver parenchyma and (B) distorted spleen architecture with germinal center activation (left, arrowheads) and fibrosis (right, arrowheads) in LysMCre-OtulinLacZ/flox mice. Scale bars, 200 μm. Micrographs are representative of five mice of each genotype from two independent experiments. (C and D) Flow cytometry analysis of total cellular infiltrate (CD45.1+ and CD45.2+) in peritoneal lavage (PL), spleen, liver, and kidney of LysMCre-Otulinflox mice presented as (C) representative dot plots of neutrophils with percent of cells in gate indicated and (D) total cell number of neutrophils, macrophages, and CD8+ T cells. (E) Concentrations of immunoglobulins in serum from 4- to 9-months-old LysMCre-Otulinflox mice. (F) ELISA analysis of serum autoantibody reactivity to ENA, dsDNA, and Smith antigen from 3- to 9-months-old LysMCre-Otulinflox mice. (G) ELISA analysis of serum B cell activating factor (BAFF) concentration from 3- to 9-months-old LysMCre-Otulinflox mice. (E–G) Data are mean of two technical replicas. (D–F) Data are presented as mean ± SEM, and n represents number of mice. See also Figure S5. Figure 6 OTULIN Deficiency Leads to Autoactivation of Macrophages (A) Immunoblots of different polyUb chains in whole-cell lysate from untreated (i.e., no exogenous stimulation after differentiation) LysMCre-Otulinflox BMDMs. (Right) Densitometry analysis of the Met1-Ub signal from above the 51 kDa marker in immunoblot experiments. (B) Immunoblots of NF-κB signaling proteins from untreated LysMCre-Otulinflox BMDMs. (C) Immunoblot analysis of IκBα stability in LysMCre-Otulinflox BMDMs treated with anti-TNF neutralizing antibodies or isotype control and cycloheximide (CHX) as indicated. (D) Relative mRNA levels of Tnf, Nfkbia, Il6, and Tnfaip3 from untreated LysMCre-Otulinflox BMDMs measured by quantitative RT-PCR. Each data point is mean of two technical replicas. Statistical significance was determined using two-tailed Student’s t test. (E) Luminex analysis of TNF and IL-6 from cell culture supernatants of untreated LysMCre-Otulinflox BMDMs. Cells were split, washed in PBS, and reseeded in fresh cell culture medium 24 hr prior to analysis. Results were pooled from two independent experiments. (F) Viability of LysMCre-Otulinflox BMDMs 10 hr after treatment. Each experiment was performed as biological duplicates. Results were normalized to LysMCre-Otulin+/flox. (G and H) Immunoblots of signaling proteins from untreated LysMCre-Otulinflox BMDMs. (I) Densitometry analysis of IκBα stability from experiments performed as in (C). (J) Relative mRNA levels of Tnf measured by quantitative RT-PCR in LysMCre-Otulinflox BMDMs treated with 1 ng/mL TNF as indicated (n = 3). Data are presented as mean ± SEM, and n represents number of biological replicas. See also Figure S6. Figure 7 OTULIN Deficiency in T and B Cells Leads to Repression of the Linear Ubiquitin Chain Assembly Complex (A–C) Immunoblot of OTULIN, CYLD, and linear ubiquitin chain assembly (LUBAC) components in (A) splenic CD4+ and CD8+ T cells from CD4Cre-Otulinflox mice, (B) splenic CD19+ B cells from MB1Cre-Otulinflox mice, and (C) CD11b+ myeloid cells from multiple tissues from LysMCre-Otulinflox mice. (D and E) Relative mRNA expression of LUBAC components from (D) splenic CD4+ and CD8+ T cells from CD4Cre-Otulinflox mice (n = 3) and (E) splenic CD19+ B cells from MB1Cre-Otulinflox mice (n = 3). Data are presented as mean ± SEM, and n represents number of biological replicas. (F and G) Immunoblot of LUBAC components in (F) splenic CD4+ and CD8+ T cells from CD4Cre-Otulinflox mice and (G) splenic CD19+ B cells from MB1Cre-Otulinflox mice treated with 10 μM MG132 proteasomal inhibitor for 6 hr as indicated. (H) Schematic showing a model of the cellular effect of OTULIN deficiency in myeloid cells and T and B cells. Figure S1 Genetic Linkage Analysis of Affected Patients and Biochemical and Biophysical Analysis of OTULINL272P, Related to Figure 1 (A) Candidate regions of linkage following genome-wide linkage scan using Affymetrix 250K SNP arrays in three affected patients. HomozygosityMapper output shows common regions of genome-wide homozygosity in the three patients. The graph shows the genome-wide homozygosity scores produced by HomozygosityMapper plotted as bar chart with red bars indicating the most promising genomic linkage regions (red). (B) Circular dichroism spectroscopy in the ultraviolet wavelength region of OTULINWT and OTULINL272P. This spectrum is representative of two independent experiments carried out at a protein concentration of 0.38 mg/mL. (C) Tryptophan fluorescence upon thermal unfolding measured by nanoDSF. Destabilization of OTULINWT and OTULINL272P dependent on temperature shows that both proteins are stable at 37°C. Apparent melting temperatures (Tm) were determined as 56.7°C and 53°C, respectively (dashed lines). Data are representative of two independent experiments. (D) Immunoblot of endogenous HOIP co-precipitating with C-terminally GFP-tagged OTULINWT or OTULINL272P ectopically expressed in HEK293 cells. Data are representative of two independent experiments. (E) Met1-linked tetraUb hydrolysis by OTULINWT and OTULINL272P were assayed with 1 μM tetraUb over time with the indicated OTULIN concentrations and visualized on silver-stained 4%–12% gradient SDS-PAGE gels. (F) Affinity measurements by fluorescence anisotropy with FlAsH (Fluorescein Arsenical Helix-binder)-labeled Met1 diUb and catalytically inactive OTULINC129A or OTULINC129A/L272P. Data are means ± SD of one experiment performed in triplicate. Results are representative of three independent experiments. FP, fluorescence polarization. Values were fitted to a one-site total binding model to derive binding constants (KD), which could not be calculated for OTULINC129A/L272P. (G) Hydrolysis of diUb of all possible linkages by OTULINL272P in a time course with the indicated OTULINL272P concentration and visualized on silver-stained 4%–12% gradient SDS-PAGE gels. o/n, overnight treatment. Figure S2 Generation of OTULIN-Targeted Mice, OTULIN Expression, Genotyping, and Reconstitution of Bone Marrow Chimeric Mice, Related to Figure 2 (A) Schematic showing the strategy to generate conditional and cell-type-specific knockouts of Otulin. SA, splice acceptor; neo, neomycin-resistance cassette; pA, polyA signal; PGK, murine PGK-1 promoter; DTA, diphtheria toxin A selection cassette; KanR, Kanamycin-resistance cassette. (B) Genotyping of mouse strains. PCR reactions showing the expected products from each genotype. (C) E13.5 embryos stained with X-gal for β-galactosidase activity and cleared by methyl salicylate shows Otulin promoter activity in multiple tissues. Pictures are representative of five embryos of each genotype from two independent experiments. (D) Immunoblot analysis showing OTULIN expression in multiple tissues from adult wild-type C57BL/6 mice. Blots are representative of three independent experiments. (E) Ratio of CD45.1+ (wild-type B6.SJL) and CD45.2+ (CreERT2-Otulin+/flox or CreERT2-OtulinLacZ/flox C57BL/6) expressing splenocytes determined by flow cytometry at the termination of chimera experiments. Data were pooled from two independent experiments. (F) Genotyping of bone marrow cells or blood leukocytes from CreERT2-Otulinflox chimeras treated with tamoxifen or vehicle shows complete or near-complete conversion of flox alleles to del alleles upon tamoxifen treatment. Note that WT(+) products are present in all reactions as BJ6.SJL WT cells are present in all samples from the chimeras. (G) Body weight following i.p. administration of tamoxifen (tx) or vehicle (vehi) to CreERT2-Otulinflox chimeric mice. Data were pooled from two independent experiments. (H) Blood cell counts from CreERT2-Otulinflox chimeras and vehicle-treated controls at day 5. (I-J) Flow cytometry analysis of CD11b+Gr-1+ neutrophils in total cellular infiltrate (CD45.1+ and CD45.2+) in lung and kidney presented as (I) representative dot plots with percentage of cells in gate indicated and (J) total cell number or percentage of cells in gate quantified. (K) Percentage of neutrophils in infiltrate from CreERT2-Otulinflox chimeras (related to Figure 2H). Data shown in (J) and (K) are repeated in (L). (L) Flow cytometry analysis of CD11b+Gr-1+ neutrophils in liver, lung, peritoneal lavage (PL), spleen, and kidney in tamoxifen- or vehicle-treated CreERT2-Otulinflox chimeras. (M) Ratio of CD45.1+ (wild-type B6.SJL) and CD45.2+ (CreERT2-Otulin+/flox or CreERT2-OtulinLacZ/flox C57BL/6) expressing CD11b+Gr-1+ neutrophils in each tissue determined by flow cytometry at the termination of chimera experiments. Data are presented as mean ± SEM, and n represents number of mice. Figure S3 Emergency Granulopoiesis and Neutrophilia Is Controlled by G-CSF in OTULIN-Deficient Mice, Related to Figure 3 (A) Flow cytometry analysis of lineage negative (Lin-) c-Kit+Sca1+ LSK cells in bone marrow of tamoxifen-treated CreERT2-Otulinflox chimeric mice shows increased numbers of LSK cells in bone marrow consistent with emergency granulopoiesis. (B-C) Flow cytometry analysis of mature and immature neutrophils in (B) bone marrow and (C) blood of tamoxifen-treated CreERT2-Otulinflox chimeric mice shows increased numbers of these cells in both tissues, consistent with emergency granulopoiesis. (D-F) Quantification of flow cytometry analysis as in (A-C) of LSK cells (D) and mature and immature neutrophils in (E) bone marrow and (F) blood of tamoxifen-treated CreERT2-Otulinflox chimeric mice injected with anti-G-CSF neutralizing antibodies, anti-IL-6 neutralizing antibodies, or isotype control as indicated. (G-I) Serum concentrations of TNF, G-CSF, IL-6, KC, and MCP-1 from tamoxifen-treated CreERT2-Otulinflox chimeric mice injected with (G) anti-TNF neutralizing antibodies, (H) anti-G-CSF neutralizing antibodies, (I) anti-IL-6 neutralizing antibodies, or isotype control as indicated measured by Luminex multiplex analysis. These data are represented as a heat map in Figure 3J. Data are presented as mean ± SEM, and n represents number of mice. Figure S4 Characterization of CD4Cre-Otulinflox, MB1Cre-Otulinflox, and LysMCre-Otulinflox mice, Related to Figure 4 (A) Tabulated results from flow cytometry analysis of cellular subsets from thymus, inguinal lymph node, and spleen from CD4Cre-Otulinflox mice. (B) Tabulated results from flow cytometry analysis of cellular subsets from bone marrow, peritoneal lavage (PL), spleen, liver, inguinal lymph node, and lung from MB1Cre-Otulinflox mice. (C) Results of flow cytometry analysis of peritoneal B cell subsets from MB1Cre-Otulinflox mice (n = 3). (D) Kaplan-Meier plot of survival in LysMCre-Otulin mice. Censored deaths are indicated as black ticks. (E) Body weight plotted against age in LysMCre-Otulinflox mice. Data are presented for either both genders (top panel) or stratified for gender (female (♀), middle panel; male (♂), bottom panel). Dashed lines show the linear regression of the data. (F) Liver weight to body weight ratio (left panel) and thymus weight (right panel) in LysMCre-Otulinflox mice. (C and F) Data are presented as mean ± SEM, and n represents number of mice. Figure S5 Flow Cytometry Plots and Quantification of Cell Populations in Mice Lacking Otulin in Myeloid Cells, Related to Figure 5 (A-D) Flow cytometry analysis of (A-B) CD11b+Gr-1+ neutrophils and CD11b+Gr-1- macrophages, and (C-D) CD8+ T cells in liver, peritoneal lavage (PL), spleen, lung, and kidney from from 4-9 month old LysMCre-Otulinflox mice presented as (A and C) representative dot plots or (B and D) total cell number or frequency of infiltrating cells. Missing graphs of total number of infiltrating cells in (B) and (D) are shown in main Figure 5D. (A-D) Data were pooled from two independent experiments. Data are presented as mean ± SEM, and n represents number of mice. Figure S6 Autoactivation of BMDMs and the Effect of Antibiotics on Systemic Inflammation in OTULIN-Deficient Bone Marrow Chimeric Mice, Related to Figure 6 (A) Immunoblot showing NF-κB activation in MEFs treated with 0.25 ng/mL TNF and anti-TNF neutralizing antibodies or isotype control as indicated. (B) Schematic of the experiment indicating timing of enrofloxacin and tamoxifen treatment. (C) Enrofloxacin treatment does not rescue weight loss, indicating sterile inflammation. Tamoxifen (tx) was administered i.p. (arrows) to CreERT2-Otulinflox chimeras treated with enrofloxacin or not. (D) Blood cell counts taken at day 4 from CreERT2-Otulinflox chimeras treated with enrofloxacin or not showing that enrofloxacin does not reduce neutrophilia in CreERT2-OtulinLacZ/flox chimeras, indicating sterile inflammation. (E) Luminex multiplex analysis of cytokine and chemokine concentrations in serum from terminal bleeds on day 4 of CreERT2-Otulinflox chimeras treated with enrofloxacin or not showing that enrofloxacin does not reduce secretion of inflammatory cytokines in CreERT2-OtulinLacZ/flox chimeras, indicating sterile inflammation. (B-E) Data were pooled from two independent experiments (except for G-CSF concentrations in E). (F) Micrographs of H&E stained sections of livers reveal inflammatory foci (arrowheads) in liver parenchyma showing that enrofloxacin does not reduce infiltration and inflammatory foci in CreERT2-OtulinLacZ/flox chimeras, indicating sterile inflammation. Micrographs are representative of 10 CreERT2-Otulin+/flox and 12 CreERT2-OtulinLacZ/flox chimeras from two independent experiments. Scale bars: 200 μm. (G) Frequency of infiltrating CD11b+Gr-1+ neutrophils. (C-E and G) Data are presented as mean ± SEM, and n represents number of mice. ==== Refs References Boisson B. Laplantine E. Prando C. Giliani S. Israelsson E. Xu Z. Abhyankar A. Israël L. Trevejo-Nunez G. Bogunovic D. Immunodeficiency, autoinflammation and amylopectinosis in humans with inherited HOIL-1 and LUBAC deficiency Nat. Immunol. 13 2012 1178 1186 23104095 Boisson B. Laplantine E. Dobbs K. Cobat A. Tarantino N. Hazen M. Lidov H.G.W. Hopkins G. Du L. Belkadi A. Human HOIP and LUBAC deficiency underlies autoinflammation, immunodeficiency, amylopectinosis, and lymphangiectasia J. Exp. Med. 212 2015 939 951 26008899 Bonizzi G. Karin M. The two NF-kappaB activation pathways and their role in innate and adaptive immunity Trends Immunol. 25 2004 280 288 15145317 Clark G.R. Sciacovelli M. Gaude E. Walsh D.M. Kirby G. Simpson M.A. Trembath R.C. Berg J.N. Woodward E.R. Kinning E. Germline FH mutations presenting with pheochromocytoma J. Clin. Endocrinol. Metab. 99 2014 E2046 E2050 25004247 Clausen B.E. Burkhardt C. Reith W. Renkawitz R. Förster I. Conditional gene targeting in macrophages and granulocytes using LysMcre mice Transgenic Res. 8 1999 265 277 10621974 Damgaard R.B. Nachbur U. Yabal M. Wong W.W.-L. Fiil B.K. Kastirr M. Rieser E. Rickard J.A. Bankovacki A. Peschel C. The ubiquitin ligase XIAP recruits LUBAC for NOD2 signaling in inflammation and innate immunity Mol. Cell 46 2012 746 758 22607974 Damgaard R.B. Fiil B.K. Speckmann C. Yabal M. zur Stadt U. Bekker-Jensen S. Jost P.J. Ehl S. Mailand N. Gyrd-Hansen M. Disease-causing mutations in the XIAP BIR2 domain impair NOD2-dependent immune signalling EMBO Mol. Med. 5 2013 1278 1295 23818254 Draber P. Kupka S. Reichert M. Draberova H. Lafont E. de Miguel D. Spilgies L. Surinova S. Taraborrelli L. Hartwig T. LUBAC-Recruited CYLD and A20 Regulate Gene Activation and Cell Death by Exerting Opposing Effects on Linear Ubiquitin in Signaling Complexes Cell Rep. 13 2015 2258 2272 26670046 Elliott P.R. Nielsen S.V. Marco-Casanova P. Fiil B.K. Keusekotten K. Mailand N. Freund S.M.V. Gyrd-Hansen M. Komander D. Molecular basis and regulation of OTULIN-LUBAC interaction Mol. Cell 54 2014 335 348 24726323 Emmerich C.H. Ordureau A. Strickson S. Arthur J.S.C. Pedrioli P.G.A. Komander D. Cohen P. Activation of the canonical IKK complex by K63/M1-linked hybrid ubiquitin chains Proc. Natl. Acad. Sci. USA 110 2013 15247 15252 23986494 Fiil B.K. Gyrd-Hansen M. Met1-linked ubiquitination in immune signalling FEBS J. 281 2014 4337 4350 25060092 Fiil B.K. Damgaard R.B. Wagner S.A. Keusekotten K. Fritsch M. Bekker-Jensen S. Mailand N. Choudhary C. Komander D. Gyrd-Hansen M. OTULIN restricts Met1-linked ubiquitination to control innate immune signaling Mol. Cell 50 2013 818 830 23806334 Gerlach B. Cordier S.M. Schmukle A.C. Emmerich C.H. Rieser E. Haas T.L. Webb A.I. Rickard J.A. Anderton H. Wong W.W.-L. Linear ubiquitination prevents inflammation and regulates immune signalling Nature 471 2011 591 596 21455173 Groom J.R. Fletcher C.A. Walters S.N. Grey S.T. Watt S.V. Sweet M.J. Smyth M.J. Mackay C.R. Mackay F. BAFF and MyD88 signals promote a lupuslike disease independent of T cells J. Exp. Med. 204 2007 1959 1971 17664289 Hameyer D. Loonstra A. Eshkind L. Schmitt S. Antunes C. Groen A. Bindels E. Jonkers J. Krimpenfort P. Meuwissen R. Toxicity of ligand-dependent Cre recombinases and generation of a conditional Cre deleter mouse allowing mosaic recombination in peripheral tissues Physiol. Genomics 31 2007 32 41 17456738 Harhaj E.W. Dixit V.M. Regulation of NF-κB by deubiquitinases Immunol. Rev. 246 2012 107 124 22435550 Hobeika E. Thiemann S. Storch B. Jumaa H. Nielsen P.J. Pelanda R. Reth M. Testing gene function early in the B cell lineage in mb1-cre mice Proc. Natl. Acad. Sci. USA 103 2006 13789 13794 16940357 Hrdinka M. Fiil B.K. Zucca M. Leske D. Bagola K. Yabal M. Elliott P.R. Damgaard R.B. Komander D. Jost P.J. Gyrd-Hansen M. CYLD Limits Lys63- and Met1-Linked Ubiquitin at Receptor Complexes to Regulate Innate Immune Signaling Cell Rep. 14 2016 2846 2858 26997266 Ikeda F. Deribe Y.L. Skånland S.S. Stieglitz B. Grabbe C. Franz-Wachtel M. van Wijk S.J.L. Goswami P. Nagy V. Terzic J. SHARPIN forms a linear ubiquitin ligase complex regulating NF-κB activity and apoptosis Nature 471 2011 637 641 21455181 Iwai K. Fujita H. Sasaki Y. Linear ubiquitin chains: NF-κB signalling, cell death and beyond Nat. Rev. Mol. Cell Biol. 15 2014 503 508 25027653 Jiang X. Chen Z.J. The role of ubiquitylation in immune defence and pathogen evasion Nat. Rev. Immunol. 12 2012 35 48 22158412 Keusekotten K. Elliott P.R. Glockner L. Fiil B.K. Damgaard R.B. Kulathu Y. Wauer T. Hospenthal M.K. Gyrd-Hansen M. Krappmann D. OTULIN antagonizes LUBAC signaling by specifically hydrolyzing Met1-linked polyubiquitin Cell 153 2013 1312 1326 23746843 Komander D. Rape M. The ubiquitin code Annu. Rev. Biochem. 81 2012 203 229 22524316 Kumari S. Redouane Y. López-Mosqueda J. Shiraishi R. Romanowska M. Lutzmayer S. Kuiper J. Martinez C. Dikic I. Pasparakis M. Ikeda F. Sharpin prevents skin inflammation by inhibiting TNFR1-induced keratinocyte apoptosis eLife 3 2014 3 Lee P.P. Fitzpatrick D.R. Beard C. Jessup H.K. Lehar S. Makar K.W. Pérez-Melgosa M. Sweetser M.T. Schlissel M.S. Nguyen S. A critical role for Dnmt1 and DNA methylation in T cell development, function, and survival Immunity 15 2001 763 774 11728338 Li H. Handsaker B. Wysoker A. Fennell T. Ruan J. Homer N. Marth G. Abecasis G. Durbin R. 1000 Genome Project Data Processing Subgroup The Sequence Alignment/Map format and SAMtools Bioinformatics 25 2009 2078 2079 19505943 MacDuff D.A. Reese T.A. Kimmey J.M. Weiss L.A. Song C. Zhang X. Kambal A. Duan E. Carrero J.A. Boisson B. Phenotypic complementation of genetic immunodeficiency by chronic herpesvirus infection eLife 4 2015 4 Morley J.E. Thomas D.R. Wilson M.-M.G. Cachexia: pathophysiology and clinical relevance Am. J. Clin. Nutr. 83 2006 735 743 16600922 Morrison P.J. Bending D. Fouser L.A. Wright J.F. Stockinger B. Cooke A. Kullberg M.C. Th17-cell plasticity in Helicobacter hepaticus-induced intestinal inflammation Mucosal Immunol. 6 2013 1143 1156 23462910 Park Y. Jin H.-S. Lopez J. Lee J. Liao L. Elly C. Liu Y.-C. SHARPIN controls regulatory T cells by negatively modulating the T cell antigen receptor complex Nat. Immunol. 17 2016 286 296 26829767 Peltzer N. Rieser E. Taraborrelli L. Draber P. Darding M. Pernaute B. Shimizu Y. Sarr A. Draberova H. Montinaro A. HOIP deficiency causes embryonic lethality by aberrant TNFR1-mediated endothelial cell death Cell Rep. 9 2014 153 165 25284787 Redecke V. Chaturvedi V. Kuriakose J. Häcker H. SHARPIN controls the development of regulatory T cells Immunology 148 2016 216 226 26931177 Rickard J.A. Anderton H. Etemadi N. Nachbur U. Darding M. Peltzer N. Lalaoui N. Lawlor K.E. Vanyai H. Hall C. TNFR1-dependent cell death drives inflammation in Sharpin-deficient mice eLife 3 2014 3 Rivkin E. Almeida S.M. Ceccarelli D.F. Juang Y.-C. MacLean T.A. Srikumar T. Huang H. Dunham W.H. Fukumura R. Xie G. The linear ubiquitin-specific deubiquitinase gumby regulates angiogenesis Nature 498 2013 318 324 23708998 Rodríguez C.I. Buchholz F. Galloway J. Sequerra R. Kasper J. Ayala R. Stewart A.F. Dymecki S.M. High-efficiency deleter mice show that FLPe is an alternative to Cre-loxP Nat. Genet. 25 2000 139 140 10835623 Sasaki Y. Sano S. Nakahara M. Murata S. Kometani K. Aiba Y. Sakamoto S. Watanabe Y. Tanaka K. Kurosaki T. Iwai K. Defective immune responses in mice lacking LUBAC-mediated linear ubiquitination in B cells EMBO J. 32 2013 2463 2476 23942237 Satpathy S. Wagner S.A. Beli P. Gupta R. Kristiansen T.A. Malinova D. Francavilla C. Tolar P. Bishop G.A. Hostager B.S. Choudhary C. Systems-wide analysis of BCR signalosomes and downstream phosphorylation and ubiquitylation Mol. Syst. Biol. 11 2015 810 26038114 Schaeffer V. Akutsu M. Olma M.H. Gomes L.C. Kawasaki M. Dikic I. Binding of OTULIN to the PUB domain of HOIP controls NF-κB signaling Mol. Cell 54 2014 349 361 24726327 Schindelin J. Arganda-Carreras I. Frise E. Kaynig V. Longair M. Pietzsch T. Preibisch S. Rueden C. Saalfeld S. Schmid B. Fiji: an open-source platform for biological-image analysis Nat. Methods 9 2012 676 682 22743772 Seelow D. Schuelke M. Hildebrandt F. Nürnberg P. HomozygosityMapper--an interactive approach to homozygosity mapping Nucleic Acids Res. 37 2009 W593–W539 Tokunaga F. Sakata S. Saeki Y. Satomi Y. Kirisako T. Kamei K. Nakagawa T. Kato M. Murata S. Yamaoka S. Involvement of linear polyubiquitylation of NEMO in NF-kappaB activation Nat. Cell Biol. 11 2009 123 132 19136968 Tokunaga F. Nakagawa T. Nakahara M. Saeki Y. Taniguchi M. Sakata S. Tanaka K. Nakano H. Iwai K. SHARPIN is a component of the NF-κB-activating linear ubiquitin chain assembly complex Nature 471 2011 633 636 21455180 Yang H. Wang K. Genomic variant annotation and prioritization with ANNOVAR and wANNOVAR Nat. Protoc. 10 2015 1556 1566 26379229 Yang Y. Kelly P. Shaffer A.L. 3rd Schmitz R. Yoo H.M. Liu X. Huang W. Webster D. Young R.M. Nakagawa M. Targeting Non-proteolytic Protein Ubiquitination for the Treatment of Diffuse Large B Cell Lymphoma Cancer Cell 29 2016 494 507 27070702

PMC005xxxxxx/PMC5002270.txt

==== Front Surg Case RepSurg Case RepSurgical Case Reports2198-7793Springer Berlin Heidelberg Berlin/Heidelberg 21110.1186/s40792-016-0211-0Case ReportSuccessful treatment of rectovaginal fistula and rectal stenosis due to perianal Crohn’s disease by dual-port laparoscopic abdominoperineal resection: a report of two cases Matsuzawa Fumihiko mattuntun999@yahoo.co.jp Homma Shigenori +81-11-706-5927homma.s@nifty.com Yoshida Tadashi ta73825@dg7.so-net.ne.jp Shibasaki Susumu susumushi48@mist.ocn.ne.jp Minagawa Nozomi minanon@cam.hi-ho.ne.jp Shimokuni Tatsushi t.shimokuni19720421@gmail.com Sakihama Hideyasu akinokehai69@hotmail.com Kawamura Hideki h.kawamura@med.hokudai.ac.jp Takahashi Norihiko noripiko@med.hokudai.ac.jp Taketomi Akinobu taketomi@med.hokudai.ac.jp Department of Gastroenterological Surgery I, Hokkaido University Graduate School of Medicine, North 15, West 7, Kita-Ku, Sapporo, Hokkaido 060-8638 Japan 27 8 2016 27 8 2016 12 2016 2 1 8318 3 2016 4 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.Background The incidence of rectovaginal fistula in women with Crohn’s disease has been reported to be 3–10 %. Although rectovaginal fistulas can be managed medically and surgically, they have high rates of recurrence and complications. Rectal stenosis is another condition that occurs due to perianal Crohn’s disease. A novel, minimally invasive procedure, dual-port laparoscopic abdominoperineal resection using a multichannel port, has been shown effective in patients with lower rectal cancer and patients with medically uncontrolled ulcerative colitis. This report describes the use of the same method for two patients with Crohn’s disease-related rectovaginal fistula and rectal stenosis. Case presentation The first patient, a 22-year-old woman, was diagnosed with rectovaginal fistula and rectal stenosis due to perianal Crohn’s disease 2 years earlier. Induction therapy with infliximab and endoscopic balloon dilatation did not improve her symptoms. The second patient, a 33-year-old woman, was also diagnosed with rectovaginal fistula and rectal stenosis due to perianal Crohn’s disease, and medical treatment was also unsuccessful. Both patients underwent dual-port laparoscopic abdominoperineal resection using a multichannel port, with no perioperative and postoperative complications. Conclusion These findings show that this reduced port method can be used to successfully treat patients with Crohn’s disease-associated rectovaginal fistula and rectal stenosis. Keywords Rectovaginal fistulaRectal stenosisPerianal Crohn’s diseaseReduced port surgeryDual-port laparoscopic abdominoperineal resectionissue-copyright-statement© The Author(s) 2016 ==== Body Background Fistulas are abnormal tracts arising from mucosal ulcerations in the gastrointestinal tract that have progressed to penetrate other structures [1]. The cumulative incidence of fistulas in patients with Crohn’s disease (CD) has been reported to be 33 % after 10 years and 50 % after 20 years, with 83 % of all fistula episodes requiring surgery [2]. The incidence of rectovaginal fistulas (RVF) in women with CD has been reported to be approximately 3–10 %. Although medical and surgical management techniques have been described, these fistulas are difficult to manage because of their high rates of recurrence and complications. In addition to RVF, women with CD can also experience rectal stenosis, a major complication that is also difficult to manage [3]. We have described the successful use of laparoscopically assisted anterior resection, using a single incision and one additional port, for patients with rectal cancer [4]. Based on this technique, we have developed a novel, minimally invasive procedure, dual-port laparoscopic abdominoperineal resection using a multichannel port, and shown its effectiveness in patients with lower rectal cancer and patients with medically uncontrolled ulcerative colitis [5, 6]. This report documents the use of the same method in two patients with CD-related RVF and rectal stenosis. Case presentation Patients A 22-year old woman was diagnosed with colonic CD 6 years earlier. For 4 years, she experienced fluctuating disease, with periods of remission and recurrence. She was initially treated with maximal doses of 5-aminosalicylate (5-ASA), and corticosteroids, but later switched doses to azathioprine. Two years earlier, she started to pass gas from her vagina, and she was diagnosed with RVF due to CD. She was started on induction therapy using infliximab, but her symptoms did not improve. She also had a 2-year history of intestinal obstruction due to CD-associated rectal stenosis. She underwent endoscopic balloon dilatation to treat the stenosis, but experienced no improvement. Because of her inability to take nutrition through her intestinal tract, she was started on total parenteral nutrition. She consulted our department for the treatment of RVF and repeat intestinal obstruction. A lower gastrointestinal series showed the disappearance of haustra, along with stenosis in the transverse and descending colon (Fig. 1a), rectal stenosis and RVF (Fig. 1b). Dual-port laparoscopic abdominoperineal resection, from the transverse colon near the splenic flexure to the rectum, was performed. The resected specimen showed both the RVF and rectal stenosis (Fig. 2). The patient experienced no perioperative or postoperative complications. She was discharged 14 days after surgery and has not experienced any recurrence of CD for 3 years.Fig. 1 Preoperative lower gastrointestinal series of patient 1, showing stenosis and the disappearance of haustra in the transverse and descending colon (white arrow in a) and the rectovaginal fistula (black circle in b) Fig. 2 View of the resected specimen from patient 2, showing a fistula from the rectum to the vagina (black arrow) A 33-year old woman was diagnosed with small intestinal and colonic CD 14 years earlier. Six years ago, she underwent ileocecal resection for stenosis of the ileum. After surgery, she was started on infliximab but had to discontinue treatment because of the side effects of that drug. Three years ago, she started to pass gas from her vagina, and she was diagnosed with RVF due to CD. Medical therapy was unsuccessful, and she consulted our department. A preoperative lower gastrointestinal series showed the RVF, along with stenosis and the disappearance of haustra in the transverse and descending colon. She underwent the same operation as the first patient, with the same extent of resection. Pathological examination showed a fistula from the rectum to the vagina and stenosis from the transverse colon to the rectum. She did not experience any perioperative or postoperative complications. She was discharged 15 days after surgery and has not experienced recurrence of CD for 4 years. Operative procedures A multichannel port was inserted through a 25-mm skin incision in the right or left lower quadrant at the colostomy site. The colostomy site was determined by the preoperative marking, regardless of the surgical procedure. The choice of multichannel port site was based on the length of colon to be resected. An additional 5-mm trocar for the left hand was inserted at the umbilicus (Fig. 3). The surgeon and the assistant were positioned on the patient’s right side (Fig. 4). For anterior dissection, the uterus was suspended by two stitches of 3–0 monofilament thread penetrating the left and right lower abdomen near the pubic bone. The assistant retracted the rectosigmoid ventrally using a curved grasper, and the peritoneum was incised using an ultrasonic coagulation incision device from the level of the sacral promontory to the base of the inferior mesenteric artery. After identification of the left ureter, the vascular pedicle was lifted vertically by the assistant. The inferior mesenteric artery was isolated and ligated 1 cm distal to the aorta. The inferior mesenteric vein was also ligated. Pelvic dissection was performed along the presacral avascular plane down to the pelvic floor and the top of the anal canal. The rectum was completely mobilized to the level of the levator ani muscle. The oral side of the intestine, which was free of the mucosal disorder due to CD, was cut off. Perineal dissection was performed circumferentially to the pelvic cavity. The specimen was retrieved through the perineal wound. A drain was placed from the umbilicus into the pelvic bottom, and a colostomy was fashioned at the site of the multichannel port (Fig. 5).Fig. 3 Insertion of a multichannel port at the colostomy site through a 25-mm skin incision in the right lower quadrant. A 5-mm trocar was inserted via the umbilicus Fig. 4 Positioning of the surgeon and assistant on the patient’s right side Fig. 5 Creation of a colostomy at the site of the multichannel port and placement of the drain through the umbilicus Discussion This study reports the performance of dual-port laparoscopic abdominoperineal resection using a multiple port method. Neither patient experienced any perioperative or postoperative complications. In contrast, the rates of complications were reported to be much higher in patients with CD-related RVF who underwent conventional proctectomy, with 35 % experiencing delayed perineal wound healing, 17 % having intra-abdominal sepsis, and 15 % experiencing stomal complications [7]. Another series reported delayed or failed perineal wound healing in almost 50 % of patients [8]. The novel, minimally invasive procedure described here, dual-port laparoscopic abdominoperineal resection using a multiple port, has been found effective in patients with lower rectal cancer [5] and in patients with medically uncontrolled ulcerative colitis [6]. This procedure was completed successfully in all patients, without any intraoperative complications, and all postoperative outcomes were satisfactory. This report describes our use of this surgical method in two patients with CD-related RVF and rectal stenosis. The abdominal cavity was approached using two small incisions, with a multichannel port placed at the colostomy site through a 25-mm skin incision in the lower quadrant and a 5-mm trocar inserted via the umbilicus for postoperative placement of a drainage tube. Neither patient experienced any intraoperative or postoperative complications. We recently reported that the postoperative neutrophil count was lower after SLIS +1 port laparoscopy-assisted than after conventional laparoscopy-assisted anterior resection for rectal cancer [4]. Furthermore, the former group experienced a significant difference in body temperature on postoperative day 1, indicating a lower degree of inflammation. Our findings suggest that this procedure, involving small incisions and minimal invasiveness, may reduce the risk of complications and benefit not only patients with rectal cancer and ulcerative colitis but also patients with CD-related RVF and rectal stenosis. CD is the second most common cause of RVF after obstetrical trauma. The incidence of RVF in women with CD is approximately 3–10 %. RVF may cause significant clinical distress and social embarrassment. RVFs are extremely difficult to close medically [9], often leaving surgery as the only option [10]. Medical treatments have included antibiotics, corticosteroids, and immunosuppressants, but these agents are associated with low rates of long-term symptom control and unacceptably high rates of recurrence [11]. Infliximab, a monoclonal antibody to tumor necrosis-α (TNF-α), is a major advance in the treatment of fistulizing CD disease and has completely altered treatment strategies for perianal disease [12]. However, analysis of the results of the ACCENT II (A CD Clinical trial Evaluating infliximab in a New long-term Treatment regimen in patients with fistulizing CD) trial found that 56 % of patients on maintenance therapy with infliximab experienced RVF recurrence [10]. Another study showed that response to infliximab differed among patients with different types of CD fistula [13]. The closure rate after 4 to 6 weeks of treatment was 76 % for all external CD fistulas, but only 14 % for CD-associated RVF [13]. Patients who cannot be managed medically or are resistant or intolerant to infliximab can be managed surgically. Proctectomy was performed initially because of the high recurrence rate of CD-related RVF and difficulties treating rectal stenosis. To date, there have been no prospective, randomized, controlled trials assessing methods for the surgical correction of CD-related RVF [14]. Transvaginal, perineal, and transanal approaches, with or without transabdominal mobilization, can be used for local repair. Fecal diversion remains a problem, but protecting the fistula repair with a diverting stoma was reported to improve healing and reduce recurrence [15]. However, even with a diverting stoma, the cure rate remains less than satisfactory. The rates of recurrence of CD-related RVF have been reported to range from 25 to 50 %, higher than the recurrence rates of other CD fistulas [16–20]. Furthermore, most studies have considered only short-term outcomes. For example, of 12 patients with CD-related RVF who underwent local repairs, 7 (58 %) showed recurrence [16]. Proctectomy is still considered the only radical cure for Crohn’s-related RVF. Both patients in this report presented with RVF and rectal stenosis, with the latter causing repeat intestinal obstruction. Rectal strictures due to CD are as difficult to treat as RVF. For example, 66 % of patients with perianal CD and rectal strictures required a permanent stoma, with multivariate analysis showing that rectal stricture was independently predictive of the need for a permanent stoma [3]. This report has certain limitations. First, both patients had low degrees of adhesion, allowing conventional laparoscopic surgery. Second, this operation required high operative skill of the entire surgical team. The surgeon and team in this study had experience with over 500 laparoscopic colorectal resection procedures, including reduced port surgery for colorectal diseases including colorectal cancer. Conclusions CD-related RVF and rectal stenosis are difficult to manage medically and surgically. Dual-port laparoscopic abdominoperineal resection using a multichannel port, when performed by experienced surgeons, may be useful for selected patients with these conditions. Consent Written informed consent was obtained from both patients for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal. Abbreviations CDCrohn’s disease RVFrectovaginal fistulas. We thank the staff of Hokkaido University Hospital. Authors’ contributions SH, TS, HS, NT, and HK participated in the care of the patients. All authors participated in data collection. FM and SH evaluate these two patients based on previous literature and drafted the manuscript. NT and AT participated in revising the manuscript critically. All authors read and approved the final manuscript. Authors’ information FM is a doctoral student; SH, TY, SS, NM, TS, HS, HK, and NT are staff surgeons; and AT is a professor in the Department of Gastroenterological Surgery I of Hokkaido University Graduate School of Medicine. Competing interests The authors declare that they have no competing interests. ==== Refs References 1. Levy C Tremaine WJ Management of internal fistulas in Crohn’s disease Inflamm Bowel Dis 2002 8 106 111 10.1097/00054725-200203000-00007 11854609 2. Schwartz DA Loftus EV Jr Tremaine WJ Panaccione R Harmsen WS Zinsmeister AR The natural history of fistulizing Crohn’s disease in Olmsted County, Minnesota Gastroenterology 2002 122 875 880 10.1053/gast.2002.32362 11910338 3. Galandiuk S Kimberling J Al-Mishlab TG Stromberg AJ Perianal Crohn disease: predictors of need for permanent diversion Ann Surg 2005 241 796 802 10.1097/01.sla.0000161030.25860.c1 15849515 4. Kawamata F Homma S Minagawa N Kawamura H Takahashi N Taketomi A Comparison of single-incision plus one additional port laparoscopy-assisted anterior resection with conventional laparoscopy-assisted anterior resection for rectal cancer World J Surg 2014 38 2716 2723 10.1007/s00268-014-2642-8 24852437 5. Ishikawa T Homma S Shibasaki S Yoshida T Minagawa N Kawamura H Dual-port laparoscopic abdominoperineal resection using the intended stoma site as the multichannel port Surg Today 2015 45 1583 1587 10.1007/s00595-015-1224-1 26220049 6. Homma S Kawamata F Shibasaki S Kawamura H Takahashi N Taketomi A Does reduced-port laparoscopic surgery for medically uncontrolled ulcerative colitis do more harm than good? Asian J Endosc Surg 2016 9 24 31 10.1111/ases.12250 26490730 7. Yamamoto T Allan RN Keighley MR Audit of single-stage proctocolectomy for Crohn’s disease: postoperative complications and recurrence Dis Colon Rectum 2000 43 249 256 10.1007/BF02236990 10696900 8. Cohen JL Stricker JW Schoetz DJ Jr Coller JA Veidenheimer MC Rectovaginal fistula in Crohn’s disease Dis Colon Rectum 1989 32 825 828 10.1007/BF02554548 2791765 9. Present DH Rutgeerts P Targan S Hanauer SB Mayer L van Hogezand RA Infliximab for the treatment of fistulas in patients with Crohn’s disease N Engl J Med 1999 340 1398 1405 10.1056/NEJM199905063401804 10228190 10. Sands BE Blank MA Patel K van Deventer SJ Long-term treatment of rectovaginal fistulas in Crohn’s disease: response to infliximab in the ACCENT II Study Clin Gastroenterol Hepatol 2004 2 912 920 10.1016/S1542-3565(04)00414-8 15476155 11. Zhu YF Tao GQ Zhou N Xiang C Current treatment of rectovaginal fistula in Crohn’s disease World J Gastroenterol 2011 17 963 967 10.3748/wjg.v17.i8.963 21448347 12. Ricart E Panaccione R Loftus EV Tremaine WJ Sandborn WJ Infliximab for Crohn’s disease in clinical practice at the Mayo Clinic: the first 100 patients Am J Gastroenterol 2001 96 722 729 10.1111/j.1572-0241.2001.03612.x 11280541 13. Parsi MA Lashner BA Achkar JP Connor JT Brzezinski A Type of fistula determines response to infliximab in patients with fistulous Crohn’s disease Am J Gastroenterol 2004 99 445 449 10.1111/j.1572-0241.2004.04083.x 15056083 14. Hannaway CD Hull TL Current considerations in the management of rectovaginal fistula from Crohn’s disease Colorectal Dis 2008 10 747 756 10.1111/j.1463-1318.2008.01552.x 18462243 15. Marchesa P Hull TL Fazio VW Advancement sleeve flaps for treatment of severe perianal Crohn’s disease Br J Surg 1998 85 1695 1698 10.1046/j.1365-2168.1998.00959.x 9876077 16. Makowiec F Jehle EC Becker HD Starlinger M Clinical course after transanal advancement flap repair of perianal fistula in patients with Crohn’s disease Br J Surg 1995 82 603 606 10.1002/bjs.1800820509 7613925 17. Halverson AL Hull TL Fazio VW Church J Hammel J Floruta C Repair of recurrent rectovaginal fistulas Surgery 2001 130 753 758 10.1067/msy.2001.116905 11602908 18. Joo JS Weiss EG Nogueras JJ Wexner SD Endorectal advancement flap in perianal Crohn’s disease Am Surg 1998 64 147 150 9486887 19. Ozuner G Hull TL Cartmill J Fazio VW Long-term analysis of the use of transanal rectal advancement flaps for complicated anorectal/vaginal fistulas Dis Colon Rectum 1996 39 10 14 10.1007/BF02048261 8601343 20. Sonoda T Hull T Piedmonte MR Fazio VW Outcomes of primary repair of anorectal and rectovaginal fistulas using the endorectal advancement flap Dis Colon Rectum 2002 45 1622 1628 10.1007/s10350-004-7249-y 12473885

PMC005xxxxxx/PMC5002271.txt

==== Front 3 Biotech3 Biotech3 Biotech2190-5738Springer Berlin Heidelberg Berlin/Heidelberg 48610.1007/s13205-016-0486-7Original ArticleAscorbyl palmitate synthesis in an organic solvent system using a Celite-immobilized commercial lipase (Lipolase 100L) Sharma Shivika shivikasharma25@gmail.com Kanwar Kriti Kanwar Shamsher S. Department of Biotechnology, Himachal Pradesh University, Summer Hill, Shimla, 171 005 India 27 8 2016 27 8 2016 12 2016 6 2 18319 4 2016 2 8 2016 © The Author(s) 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.Ascorbyl palmitate was synthesized using a Celite-immobilized commercial lipase (Lipolase 100L) in dimethylsulfoxide (DMSO) as an organic solvent system. Lipase immobilized by surface adsorption onto Celite 545 matrix and subsequently exposed to 1 % glutaraldehyde showed 75 % binding of protein. The Celite-bound lipase was optimally active at 75 °C and pH 8.5 under shaking and showed maximum hydrolytic activity toward p-NPP as a substrate. The bound lipase was found to be stimulated only in the presence of Al3+ and EDTA. All surfactants (Tween-20, Tween-80 and Triton X-100) had an inhibitory effect on lipase activity. The optimization of various reaction conditions of ascorbyl palmitate was achieved considering one factor at a time. The esterification of ascorbic acid and palmitic acid was carried out with 1 M ascorbic acid and 2.5 M palmitic acid in DMSO at 75 °C for 18 h under shaking (120 rpm). Molecular sieves had an important effect on the ester synthesis resulting in an enhanced yield. The by-product (H2O) produced in the reaction was scavenged by the molecular sieves (20 mg/ml) added in the reaction mixture which enhanced the ester yield to 80 %. The characterization of synthesized ester was done through FTIR spectroscopy. Keywords CeliteLipaseGlutaraldehydeAscorbyl palmitate synthesisMolecular sievesissue-copyright-statement© King Abdulaziz City for Science and Technology 2016 ==== Body Introduction Lipases bring about a range of bioconversion reactions such as hydrolysis, inter-esterification, esterification, alcoholysis, acidolysis and aminolysis (Rajesh and Reddy 2013). Lipases find potential applications in bioprocesses largely due to their availability and stability in organic as well as in aqueous media (Sharma and Kanwar 2014). The use of lipase in immobilized form ensures that the enzyme remains stable in the organic medium, facilitating its subsequent recovery (Padilha et al. 2013). Lipase-catalyzed production of flavor esters by transesterification reactions is influenced by a number of transesterification variables, such as molarity of alcohol, reaction time, temperature and amount of immobilized enzyme (Garlapati and Banerjee 2013). There are many advantages of using enzyme in low water/organic solvent media such as better solubility of substrate and products, simple removal of solvent (most of the organic solvents have lower boiling point than water), reduction in water-dependent side reactions, easy removal of enzyme after reaction since it is not dissolved in organic solvents, better thermal stability of enzyme at high temperature, elimination of microbial contamination, absence of undesirable side reactions and an increased thermal stability of the enzyme in harsh conditions. The stability of biocatalyst in an organic solvent such as DMSO might be attributed to the fact that a thin layer of molecules remain tightly bound to the enzyme acting as a protective sheath along the enzyme hydrophilic surface, allowing retention of native confirmation. Also, the stimulation effect of the solvent on the enzyme modifies the oil–water interface to make enzymatic action easier without causing protein denaturation. Enhancement in enzymatic activity in the presence of organic solvents could be due to some alterations in the catalytic hydrophobic pocket of the enzyme (lipase) containing serine, aspartic acid and histidine groups. Lipases are the most sought enzymes from the viewpoint of solvent stability and possibly because of their esterification and transesterification reactions which are favored in the non-aqueous media. Ascorbyl esters are emerging food, cosmetic and pharmaceutical additives, which can be prepared in an eco-friendly way by using lipases as catalysts. Because they are amphiphilic molecules, which possess high free radical scavenging capacity, they can be applied as liposoluble antioxidants as well as emulsifiers and biosurfactants (Stojanovic et al. 2015). l-Ascorbic acid (vitamin C) is a natural antioxidant, but is highly polar and does not dissolve in fats and oil. The esterification process of converting ascorbic acid to its acid esters has been regarded as an effective solution for overcoming such problems. Furthermore, the esterified ascorbic acid products also have the antioxidant and surfactant functions with its potential application in high fat and food industries (Sun et al. 2013). The fatty acid and their esters obtained from oil and fats are used as acylating agent for the synthesis of hydrophobic vitamin C fatty acid ester (Karmee 2011). l-Ascorbyl palmitate is an amphipathic molecule that has both a hydrophilic (polar) head and a hydrophobic (a polar) tail and hence it is both lipohilic and water soluble (Fig. 1).Fig. 1 Schematic steps for the formation of ascorbyl palmitate using immobilized lipase Ascorbyl palmitate is preferred over ascorbic acid as an ingredient in food due to its lipophilic nature (Coppen 1999). It also acts as an anticarcinogenic compound and can strongly inhibit the DNA synthesis in cancer cells (Kageyama et al. 1999). Ascorbyl palmitate is also used as an antioxidant and preservative in food, vitamins, drugs and cosmetics (Wawire et al. 2011). It is also used to control the oxidation of lipids such as PUFA (Jacobsen 2010). Ascorbyl palmitate also acts synergistically with vitamin E. In most prevalent manufacturing processes, ascorbic acid is esterified with sulfuric acid and the product of the reaction is esterified with palmitic acid (Pokorny et al. 2001). Chemical synthesis has some drawbacks, such as the low or non-selectivity for the site of esterification, the requirement for protection and de-protection steps and product purification difficulties (Liang et al. 2012). However, vitamin C ester undergoes oxidation, degradation and rearrangement under the chemo-catalyzed reaction (Karmee 2011). Owing to a number of important applications of ascorbic acid or its derivatives, the aim of the present work was to synthesize ascorbyl palmitate in an organic medium using Celite-bound immobilized commercial lipase (Lipolase 100L). Materials and methods Chemicals and reagents ZnCl2, MnCl2, MgCl2, CoCl2, HgCl2, AlCl3, CaCl2, NH4Cl2, p-nitrophenyl palmitate (p-NPP), p-nitrophenyl formate (p-NPF), p-nitrophenyl caprylate (p-NPC), p-nitrophenyl acetate (p-NPA), p-nitrophenyl benzoate (p-NPB), p-nitrophenyl laurate (p-NPL), iso-propanol, Tween-20, Tween-80, Triton X-100, EDTA, sodium citrate, Tris buffer (HIMEDIA Laboratory Ltd., Mumbai, India), palmitic acid (Acros Ltd, New Jersy, USA), Celite 545, DMSO and ascorbic acid (S.D. Fine-Chem Ltd., Hyderabad, India) were obtained. All these chemicals/reagents were of analytic grade and used as received. Commercial lipase (Lipolase 100L) The commercial and purified lipase (Lipolase 100L) was manufactured by Novozyme A/S Denmark (MW 31,700 Da) and obtained from the Department of Biotechnology, Himachal Pradesh University, Summer Hill, Shimla (India), for further use. Assay of lipase activity The hydrolytic activity of the lipase was assayed by a colorimetric method using p-NPP. The reaction mixture contained 80 μl of p-NPP (10 mM, p-NPP prepared in iso-propanol) stock solution and an appropriate amount of matrix-bound biocatalyst or free lipase. The final volume of this reaction mixture was made to 3 ml with 0.02 M Tris buffer, pH 8.5 (for free and immobilized lipase) with gum acacia (0.1 % w/v). The test tubes were incubated for 10 min at 45 °C under continuous shaking in a water-bath shaker. The absorbance of p-nitrophenol released was measured at 410 nm (Shimazdu UV/visible spectrophotometer, Japan). The enzyme activity was defined as µmole(s) of p-nitrophenol released per min by 1 ml of enzyme or 1 g of Celite-immobilized enzyme (weight of the matrix included) under standard assay conditions. Immobilization of lipase (Lipolase 100L) on Celite 545 matrix Celite 545 (2 g) was incubated with 10 ml 0.02 M Tris HCl buffer of pH 8.5 for 24 h in a vial at 37 °C in a water bath. The enzyme (4 ml, 12.00 U/ml) was added to the matrix and incubated at 37 °C. Each of the matrices was then given five washings with Tris–HCl buffer (0.02 M), pH 8.5, to get rid of unbound enzyme. The matrix was then activated by 1 % cross-linking agent glutaraldehyde for 1 h at 37 °C in a water bath. Each activated matrix was further given five washings with Tris–HCl buffer (0.02 M), pH 8.5, to get rid of unbound activating agent. The weight of enzyme-incubated matrices was recorded and the activity was assayed using 5 mg of immobilized matrix using the standard protocol. The immobilized protein in the Celite matrix was determined by subtracting the unbound protein in the supernatant from the total protein used for immobilization as well as increased/decreased total activity was calculated by adding total activity of supernatant and matrix in comparison to total enzyme units in 4 ml of purified lipase incubated earlier. Swelling capacity The swelling capacity (Sw) of the selected Celite matrix in distilled water was found as follows; Sw=W2-W1/W1 W1: weight of dry matrix in g; W2: weight of wet matrix in g (i.e., net weight of matrix after suspending it in excess volume of water for 1 h at 75 °C). Celite matrix Immobilization by adsorption using inorganic matrices like Celite-545 has been most widely used for immobilization of many enzymes. Diatomaceous earth (Celite) is a naturally occurring, soft, chalk-like sedimentary rock that is easily crumbled into a fine white to off-white powder. This powder has an abrasive feel, similar to pumice powder and is very light, due to its high porosity. The use of a porous support material is desirable/recommended for immobilization of lipase, so that suitable amounts of lipase can be spread on a surface area without conformational changes. Optimization of hydrolytic properties of immobilized lipase The effect of pH, temperature, specificity toward the hydrolysis of p-nitrophenyl esters with varying C-chain length, salt ions, chelating agents and various detergents was studied using Celite-bound lipase. Effect of pH of reaction buffer on the activity of immobilized lipase The activity of the Celite-immobilized lipase (5 mg) and free lipase (5 µl) was assayed separately by incubating the reaction cocktail at different pH ranges, viz. 6.0, 6.5, 7.0, 7.5, 8.0, 8.5, 9.0, 9.5 and 10.0 in a water-bath incubator under shaking (120 rpm) for 10 min at 75 °C and the lipase activity was measured thereof. Effect of temperature on the activity of immobilized lipase To study the effect of reaction temperature on the lipase activity, the Celite-immobilized lipase (5 mg) and free lipase (5 µl) were incubated in the reaction buffer at selected temperatures (50–80 °C) under shaking (120 rpm) for 10 min. The lipase activity at each of the selected temperatures was determined and compared. Effect of p-nitrophenyl esters of varying C-chain length (substrate) on the activity of immobilized lipase Substrate specificity of the lipase was investigated using stock(s) of p-nitrophenyl fatty acid esters of varying chain length (p-NPA, p-NPB, p-NPC, p-NPF, p-NPL and p-NPP) prepared in iso-propanol (10 mM) in the reaction cocktail and the lipase activity was assayed under standard conditions. The Celite-immobilized lipase was reacted with each of these substrates prepared in 0.05 M Tris buffer (pH 8.5) for 10 min at 75 °C under shaking (120 rpm). Thereafter, assay for the lipase activity was done. Effect of salt ions on the activity of immobilized lipase To evaluate the effect of various salt ions on Celite-immobilized lipase, an attempt was made to study the effect of a few of the selected salt ions, viz. Zn2+, Mn2+, Mg2+, NH4 +, Co2+, Hg2+, Ca2+ and Al3+, on lipase activity. Each of the metal ions was separately included in the reaction mixture at a final concentration of 1 mM. The lipase activity was assayed under standard assay conditions at 75 °C (in Tris buffer 0.02 M) and pH 8.5 and the lipase activity was measured thereof. Effect of chelating agents on the activity of immobilized lipase The effect of chelating agents (EDTA and sodium citrate) on the activity of the Celite-immobilized lipase (5 mg) was studied by pre-incubating the Celite-bound lipase with the chelating agents at a final concentration of 1, 3 and 5 mM (in Tris buffer 0.02 M, 8.5 pH) for 10 min at 75 °C. The lipase activity was assayed under standard assay conditions at 75 °C and pH 8.5. The residual lipase activity was determined in each case and expressed as relative activity with respect to the control (without chelating agents). Effect of detergents on the activity of immobilized lipase The effect of each of the selected detergents (Tween-20, Tween-80 and Triton X-100) was also studied by including the detergents in the reaction mixture at a final concentration of 1, 5, 10 and 20 % (v/v) (in Tris buffer 0.02 M and pH 8.5). The reaction cocktail treated with lipase was incubated at 75 °C for 10 min and the lipase activity was measured thereof. Reusability of Celite-immobilized lipase The immobilized lipase (5 mg) was used repetitively up to the eighth cycle of hydrolysis at 75 °C under shaking. After the first cycle of reaction, the biocatalyst was recovered (by centrifuging and decanting the reaction mixture) and this biocatalyst was used to catalyze the fresh hydrolytic reaction. Ascorbyl palmitate synthesis using immobilized lipase The efficacy of the Celite-bound lipase to perform catalysis in solvent medium was tested by performing esterification reactions of ascorbic acid and palmitic acid in DMSO (2 ml) at optimal temperature (75 °C) in a chemical reactor under shaking (120 rpm). Optimization of reaction conditions for synthesis of ascorbyl palmitate by immobilized lipase Ascorbyl palmitate was synthesized using 1 M ascorbic acid, 2.5 M palmitic acid and Celite-bound lipase taken in a glass vial (20 ml capacity). The reaction was performed at 75 °C for 10 h in a chemical reactor. The effect(s) of incubation time, reaction temperature, relative molar concentration of reactants, biocatalyst concentration and concentration of molecular sieves on the rate of synthesis of ascorbyl palmitate was separately evaluated. The ascorbyl palmitate was separated on the basis of their solubility in hot water using a separating funnel. The ascorbic acid was soluble in hot water, while the formed ester (ascorbyl palmitate formed in different test tubes) was insoluble in hot water and separated out using a separating funnel. The amount of ester synthesized was determined and represented as % yield. The synthetic reactions in water-free medium were performed in triplicate and mean values were presented. Effect of relative molar concentration of reactants on the synthesis of ascorbyl palmitate It was studied by keeping the concentration of one of the reactants (ascorbic acid) constant at 1 M and varying the concentration of the other reactant (palmitic acid; 1, 2.5, 5 M) in an organic reaction mixture. The esterification was carried out at 75 °C for incubation time of 10 h under continuous shaking using Celite-bound lipase. Effect of reaction temperature on the synthesis of ascorbyl palmitate The reaction mixture (2 ml) contained 10 mg enzyme, 1 M ascorbic acid and 2.5 M palmitic acid in DMSO. The reaction mixture was incubated at 55, 65, 75 and 85 °C in the incubator shaker at 120 rpm for 10 h. The sample were extracted and weighed for the detection of ascorbyl palmitate. Effect of incubation time for the synthesis of ascorbyl palmitate The reaction mixture comprised Celite-immobilized lipase, 1 M ascorbic acid and 2.5 M palmitic acid taken in screw-capped Teflon-coated glass vials (5 ml capacity) in an organic system comprising DMSO, respectively. The glass vials were incubated at 75 °C in a chemical reactor at periodic (6 h) intervals, viz. 0, 6, 12, 18 and 24 h, under continuous shaking (120 rpm) for the synthesis of ascorbyl palmitic acid. Effect of biocatalyst concentration on the synthesis of ascorbyl palmitate The synthesis of ascorbyl palmitate was studied by taking different amounts of immobilized lipase (5, 10, 15, 20, 25 and 30 mg/ml) in a reaction mixture (2 ml) containing 1 M ascorbic acid and 2.5 M palmitic acid in an organic system at 75 °C. The reaction was carried out for 18 h and the sample was extracted and weighed for detection of ascorbyl palmitate. Effect of molecular sieves (3 Å × 1.5 mm) on the synthesis of ascorbyl palmitate Varying amounts of molecular sieves (10–100 mg) were added to the reaction mixture. The esterification was carried out using immobilized lipase at 75 °C for 18 h under continuous shaking (120 rpm). Characterization of ascorbyl palmitate The synthesis of ascorbyl palmitate was done by Fourier transform infrared spectroscopy and the FTIR spectrum was recorded on a Perkin Elmer spectrophotometer in transmittance mode in KBr. Results and discussion Immobilization of lipase on Celite-545 matrix The swelling capacity (in 0.05 M Tris buffer; pH 8.5) of Celite 545 in water was recorded as 1.9 times. A commercial lipase “lipolase 100L” (12.02 U/ml) was used for immobilization by adsorption onto the Celite-545 matrix. The matrix treated with 1 % (v/v) glutaraldehyde showed 75 % binding/retention of commercial lipase, while the untreated matrix resulted in 61 % binding/retention of lipase. Optimization of the hydrolytic properties of immobilized lipase Effect of pH of reaction buffer on the hydrolytic activity of Celite-bound lipase The maximum enzyme activity (16.03 ± 0.015 U/g) was recorded at pH 8.5. With further change in pH, there was a gradual decrease in enzyme activity. The free enzyme showed maximum activity (13.0 U/ml) at pH 7.5. The Celite-bound lipase showed a shift in the pH of the reaction system toward alkalinity (Fig. 2). In a study (Kumar and Kanwar 2011), NC-membrane bound lipase was more stable at high pH (pH 8.0–9.5), displaying better hydrolytic activity as compared to free lipase which showed less activity at pH 8.5. The lipase immobilized on the mesoporous material of high carbon content showed high activity at pH 7.0 (Claudia et al. 2014).Fig. 2 Effect of pH of the reaction buffer on the hydrolytic activity of Celite-bound lipase. The maximum enzyme activity (16.03 ± 0.015 U/g) was recorded at pH 8.5 Effect of temperature on the hydrolytic activity of bound lipase The enzyme activity for Celite-bound lipase showed maximum enzyme activity of 17.2 ± 0.01 U/g at 75 °C (Fig. 3). An increase in the incubation temperature resulted in a gradual decrease in enzyme activity. The free enzyme on the contrary showed maximum activity of 14.36 U/ml at 65 °C and least hydrolytic activity was recorded at 80 °C. Thus, Celite-bound lipase was used for further experiments. The previous study of the lipase from Bacillus sp. ITP-001 immobilized in a sol gel matrix showed optimum activity at 80 °C (Carvalho et al. 2013). Temperature also has an important effect on the physical state of substrate dispersion. Higher temperature leads to liquefaction that tends to make substrate more diffusible and easily acceptable to the enzyme. It appeared that the structure of the lipase becomes more fluid and open at an elevated temperature of 70 °C and above that was employed for achieving esterification.Fig. 3 Effect of temperature on the hydrolytic activity of bound lipase. The enzyme activity for Celite-bound lipase showed maximum enzyme activity of 17.2 ± 0.01 U/g at 75 °C Effect of C-chain length (substrate) of p-nitrophenyl esters on the hydrolytic activity of immobilized lipase The hydrogel-immobilized lipase showed a variable specificity toward the tested p-nitrophenyl esters. The Celite-bound lipase (5 mg) showed a preferential affinity toward esters of relatively longer C-chain length such as p-NPP and p-NPL. The maximum enzyme activity was observed at 75 °C (17.26 ± 0.02 U/g) with p-NPP as a substrate (Fig. 4). The immobilized lipase showed a variable specificity/hydrolytic activity toward various p-nitrophenyl esters. The substrate specificity of a lipase is usually determined by the size and the hydrophilicity/hydrophobicity of its pockets. Usually, a tunnel-like binding site was more likely to accept substrates with long-chain fatty acids. It seems that in the present study, among various p-nitrophenyl acyl esters, the high C-length (C: 16) ester (p-NPP) was more efficiently hydrolyzed than other esters. This indicated a preferential specificity of lipase toward longer carbon chain length substrates. Previously, a lipase from psychrotrophic Pseudomonas cepacia immobilized on a commercially available silica support showed a higher activity with p-NPA and very low with p-NPP (Goncalves et al. 1997).Fig. 4 Effect of C-chain length (substrate) of p-nitrophenyl esters on the hydrolytic activity of immobilized lipase. The maximum enzyme activity was observed at 75 °C (17.26 ± 0.02 U/g) with p-NPP as a substrate Effect of salt ions on the hydrolytic activity of immobilized lipase The different salt ions often stabilize or destabilize the enzyme structure and hence influence its activity. The results indicated that the lipase was found to be stimulated only in the presence of Al+3 (18.01 ± 0.01 U/g) in comparison to the control (17.26 ± 0.01 U/g), whereas other ions inhibited the activity of the lipase (Fig. 5). The presence of metal ions at 1 mM and 5 mM of Mg2+ and Ca2+ promoted lipase activity from Bacillus sphaericus MTCC 7542 (Tamilarasan and Kumar 2011).Fig. 5 Effect of salt ions on the hydrolytic activity of immobilized lipase. The lipase activity was found to be stimulated only in the presence of Al+3 (18.01 ± 0.01 U/g) The effect of chelating agents on the hydrolytic activity of immobilized lipase The enzyme activity (18.29 ± 0.015 U/g) was slightly increased in the presence of EDTA at 1 mM concentration with respect to the control (17.26 ± 0.01 U/g) (Table 1). However, there was a decrease in the enzyme activity at higher concentration of EDTA, i.e., at 3 mM (16.73 ± 0.02 U/g) and 5 mM (15.83 ± 0.01 U/g). Sodium citrate decreased the enzyme activity at all the tested concentrations, viz. 1 mM (15.97 U/g), 3 mM (14.73 ± 0.02 U/g) and 5 mM (9.83 ± 0.01 U/g) with respect to the control. An increase in the activity of lipase from Bacillus coagulans MTCC-6375 has been reported by 20 mM EDTA (Kanwar et al. 2005).Table 1 Effect of chelating agent on bound lipases Control 17.26 (U/g) Concentration 1 (mM) Concentration 3 (mM) Concentration 5 (mM) EDTA (U/g) 18.29 16.73 15.83 Sodium citrate (U/g) 15.92 14.73 9.83 Effect of detergents on the hydrolytic activity of immobilized lipase All the selected detergents (Tween-20, Tween-80 and Triton X-100) had an inhibitory effect on the enzyme activity (Fig. 6). When pre-exposed to Celite -bound lipase (5 mg) at 1, 5, 10 and 20 % (v/v) final concentration (in Tris buffer 0.02 M, 8.5 pH), all tested detergents reduced the enzyme activity to a considerable extent in comparison to the control (17.26 ± 0.02 U/g; Fig. 6) at a temperature of 75 °C. Previously, the presence of Tween 80 caused a sharp decline in the activity of NC-bound lipase (Kumar and Kanwar 2011).Fig. 6 Effect of detergents on the hydrolytic activity of immobilized lipase. All tested detergents reduced the enzyme activity to a considerable extent in comparison to the control (17.26 ± 0.02 U/g) Reusability of Celite-immobilized lipase While studying the effect of reusability under optimized conditions, the recorded lipase activities were 18.0, 16.3, 11.4, 9.3, 5.2, 3.8, 1.9 and 0.9 U/g, respectively, during the first to eighth cycle of reuse of immobilized lipase (5 mg) at 75 °C under shaking (Fig. 7). Thus, it was observed that the Celite-immobilized lipase retained more than 50 % of its original activity up to the fourth cycle of repetitive reaction and thereafter its activity declined sharply after each cycle of reuse. In a previous study, the nitrocellulose-bound lipase was repeatedly used as a biocatalyst to perform hydrolytic reactions. The nitrocellulose-bound lipase retained more than 50 % of its original activity after the fifth repetitive cycle of hydrolysis (Kumar and Kanwar 2011).Fig. 7 Reusability of Celite-immobilized lipase. The Celite-immobilized lipase retained more than 50 % of its original activity up to the fourth cycle of repetitive reaction and thereafter its activity declined sharply after each cycle of reuse Optimization of esterification conditions for the synthesis of ascorbyl palmitate by the Celite-bound biocatalyst Effect of relative molar concentration of reactants on the synthesis of ascorbyl palmitate The effect of varying molar concentrations of each of the reactants on ascorbyl palmitate synthesis by Celite-bound lipase was studied by keeping the concentration of one of the reactants (ascorbic acid) constant at 1 M and varying the concentration of the second reactant (1, 2.5 and 5 M). The highest yield of ester was 63 % when ascorbic acid and palmitic acid were used at the concentration of 1 M:2.5 M, respectively, in DMSO (Fig. 8). The esterification was carried out at 75 °C for incubation time of 10 h under continuous shaking using Celite-bound lipase. In another study, a molar ratio of 1:3 (ethanol:propionic acid) was optimum for the synthesis of ethyl propionate in hexane (Kumar et al. 2005). Also, the best reaction conditions for immobilized Pseudomonas stutzeri lipase TL were 55 °C, 1:5 ascorbic to palmitic acid molar ratio obtaining 57 % yield of ester (Santibanez et al. 2014).Fig. 8 Effect of relative molar concentration of reactants on the synthesis of ascorbyl palmitate. The highest yield of ester was 63 % when ascorbic acid and palmitic acid were used at a concentration of 1 M:2.5 M, respectively, in DMSO Effect of reaction temperature on the synthesis of ascorbyl palmitate The optimum temperature for the synthesis of ascorbyl palmitate was evaluated by incubating the reaction mixture at 55, 65, 75 and 85 °C in the incubator shaker for 10 h. The maximum yield of ascorbyl palmitate (69 %) was noticed at a temperature of 75 °C. At a temperature higher than 75 °C, a gradual decrease in the ester yield was observed indicating a denaturing effect on the Celite-bound biocatalyst (Fig. 9). Previously, the optimum temperature for butyl ferulate (62 mM) synthesis in DMSO using silica-bound lipase was found to be 45 °C (Chandel et al. 2011).Fig. 9 Effect of reaction temperature on the synthesis of ascorbyl palmitate. The maximum yield of ascorbyl palmitate (69 %) was noticed at a temperature of 75 °C Effect of incubation time for the synthesis of ascorbyl palmitate The optimum reaction time for the synthesis of ascorbyl palmitate by Celite-bound lipase was evaluated by incubating the reaction mixture at 75 °C for different time intervals of 0, 6, 12, 18 and 24 h under continuous shaking (120 rpm) for the synthesis of ascorbyl palmitic acid. The yield of ester was found to be highest after 18 h of reaction, resulting in 72 % of ester yield (Fig. 10). In a recent study, a reaction time of 8 h in case of synthesis of chain of coumarate esters (Sharma et al. 2014) was found.Fig. 10 Effect of incubation time for the synthesis of ascorbyl palmitate. The yield of ester was found to be highest after 18 h of reaction resulting in 72 % of ester yield Effect of biocatalyst concentration on the synthesis of ascorbyl palmitate The effect of the amount of biocatalyst used to catalyze the ester synthesis was studied by incubating the reaction mixture with varying concentrations of biocatalyst (5–30 mg) for 18 h at 75 °C. The maximum ester was formed with 10 mg of enzyme (74 %) (Fig. 11) and with any further increase in the amount of biocatalyst a similar amount of ester was produced. Previously, 10 mg/ml (i.e., 1 g/ml) of Celite-bound lipase was found to give an optimum yield of ethyl ferulate in DMSO (Kumar and Kanwar 2010). In a recent study, the production of ascorbyl palmitate from palmitic acid and ascorbic acid through the esterification catalyzed by Novozyme 435 was performed under microwave irradiation; the optimized biocatalyst concentration was 15 % (w/w in relation to ascorbic acid weight) of enzyme at 70 °C and the final 71 % yield of ascorbyl palmitate could be achieved (Ingrid et al. 2014).Fig. 11 Effect of biocatalyst concentration on the synthesis of ascorbyl palmitate. The maximum ester was formed with 10 mg of enzyme (74 %) Effect of molecular sieves (3 Å × 1.5 mm) on the synthesis of ascorbyl palmitate The reaction mixture (2 g) containing 10 mg enzyme, 1 M ascorbic acid and 2.5 M palmitic acid when incubated at 75 °C with varying concentrations of molecular sieves (10–100 mg) resulted in a marked increase in the amount of ester yield with 20 mg of molecular sieves (80 %) in comparison to the control (68 %) without molecular sieves (Fig. 12). Any further increase in the molecular sieves beyond 20 mg caused a gradual to drastic decrease in the amount of ester. An improvement in the rate of esterification has been previously reported for esterification of lauric acid and methanol in the presence of molecular sieves (Mustafa 1998).Fig. 12 Effect of molecular sieves on the synthesis of ascorbyl palmitate. The maximum increase in the amount of ester yield resulted in 20 mg of molecular sieves (80 %) in comparison to the control (68 %) without molecular sieves Characterization of ascorbyl palmitate FTIR spectrum of ascorbyl palmitate had a sharp peak at 1730.9 cm−1 which was ascribed to –C=O stretching of the ester group; when compared with the spectrum of ascorbic acid, this peak of ester was not present and clearly confirmed that the lipase-catalyzed esterification reaction had occurred. The other peak at 2939.5 cm−1 was due to the C–H stretching of the methyl group present in ascorbyl palmitate (Fig. 13).Fig. 13 FTIR spectrum: a ascorbic acid, b palmitic acid, c ascorbyl palmitate Conclusion In the present study, ascorbyl palmitate was synthesized in an organic medium using Celite-bound immobilized commercial lipase (Lipolase 100L). In an aqueous reaction system, the Celite-bound lipase showed maximum activity with p-nitrophenyl palmitate (C-16) at an alkaline pH 8.5 and temperature 75 °C. The esterification of ascorbic acid and palmitic acid was successfully achieved with 10 mg/ml immobilized biocatalyst at 75 °C for 18 h under shaking (120 rpm). Molecular sieves had an enhancing effect on the ester synthesis resulting in enhanced yield(s) at a concentration of 20 mg/ml. The financial support from the Department of Biotechnology, Ministry of Science and Technology, Government of India to the Department of Biotechnology, Himachal Pradesh University, Shimla (India), is thankfully acknowledged. Moreover, the authors do not have any conflict of interest among themselves or with their parent institution. ==== Refs References Carvalho NB Barbosa JMP Oliveira MVS Fricks AT Soares ALCMF Biochemical properties of Bacillus sp. ITP-001 lipase immobilized with a sol gel process Quim Nova 2013 6 52 58 10.1590/S0100-40422013000100010 Chandel C Kumar A Kanwar SS Enzymatic synthesis of butyl ferulate by silica-immobilized lipase in a non-aqueous medium J Biomater Nanobiotechnol 2011 2 400 408 10.4236/jbnb.2011.24049 Claudia B Sindy E Lorena W Andres I Monica M Carbonaceous-siliceous composite materials as immobilization support for lipase from Alcaligenes sp.: application to the synthesis of antioxidants Carbon 2014 74 96 103 10.1016/j.carbon.2014.03.008 Coppen PP Allen JC Hamilton RJ The use of antioxidants, Chapter 5 Rancidity in foods 1999 Gaithersburg Aspen Publishers, Inc 84 103 Garlapati VK Banerjee R Solvent-free synthesis of flavour esters through immobilized lipase mediated transesterification Enzym Res 2013 Goncalves APV Lopes JM Lemos F Ribeiro FR Prazeres DMF Chabral JMS Aires-Barros MR Effect of the immobilization supports on the hydrolytic activity of a cutinase from Fusarium solanipisi Enzym Microbial Technol 1997 20 93 101 10.1016/S0141-0229(96)00089-0 Ingrid CR Costa ICR Sutili FK da Silva GVV Leite SGF Miranda LSM de Souza ROMA Lipase catalyzed ascorbyl palmitate synthesis under microwave irradiation J Mol Catal B Enzym 2014 102 127 131 10.1016/j.molcatb.2014.02.002 Jacobsen C Challenges when developing omega-3 enriched foods Lipids 2010 17 8 14 Kageyama K Bradbury MJ Zhao L Blount AL Vale WW Urocortin messenger ribonucleic acid: tissue distribution in the rat and regulation in thymus by lipopolysaccharide and glucocorticoids Endocrinology 1999 140 5651 5658 10579329 Kanwar SS Kaushal RK Jawed A Chimni SS Evaluation of methods to inhibit lipase in colorimetric assay employing p -nitrophenyl palmitate Indian J Biochem Biophys 2005 42 233 237 23923547 Karmee SK The synthesis, properties and application of ascorbyl ester Lipid Technol 2011 23 27 229 10.1002/lite.201100146 Kumar A Kanwar SS Synthesis of ethyl ferulate in organic medium using Celite-immobilized lipase Bioresour Technol 2010 102 2162 2167 10.1016/j.biortech.2010.10.027 21041076 Kumar A Kanwar SS Catalytic potential of a nitrocellulose membrane-immobilized lipase in aqueous and organic media J Appl Polym Sci 2011 Kumar S Pahujani S Ola RP Kanwar SS Gupta R Enhanced stability of silica immobilized lipase from Bacillus coagulans BTS-3 and synthesis of ethyl propionate Acta Microbiol Immunol Hung 2005 53 217 229 Liang J Zeng W Yao P Yuanan W Lipase-catalyzed regioselective synthesis of palmitolyglucose ester in ionic liquids Adv Biol Chem 2012 2 226 232 10.4236/abc.2012.23027 Mustafa UU A study on the lipase catalyzed esterification in organic solvent Trends J Agric For 1998 22 1697 1702 Padilha GS de Barros M Alegrea RM Tambourgi EB Production of ethyl valerate from Burkholderia cepacia lipase immobilized in alginate Chem Eng Trans 2013 Pokorny J Yanishlieva N Gordon M Antioxidants in food: practical applications 2001 Cornwall Woodhead Publishing Ltd. Rajesh B Reddy IB Lipase from organic solvent tolerant Bacillus strain C5: isolation and identification Int J Sci Res 2013 2 26 28 Santibanez L Wilson L Illanes A Synthesis of ascorbyl palmitate with immobilized lipase from Pseudomonas stutzeri J Am Oil Chem Soc 2014 91 405 410 10.1007/s11746-013-2378-x Sharma S Kanwar SS Organic solvent tolerant lipases and applications Sci World J 2014 Sharma S Dogra P Chauhan GS Kanwar SS Synthesis of alkyl coumarate esters by lipase of Bacillus licheniformis SCD11501 J Mol Catal B Enzym 2014 101 80 86 10.1016/j.molcatb.2013.12.017 Stojanovic M Carevic M Mihailovic M Velickovic D Dimitrijevic A Milosavic N Bezbradica D Influence of fatty acid on lipase-catalyzed synthesis of ascorbyl esters and their free radical scavenging capacity Appl Biotechnol Biochem 2015 62 458 466 10.1002/bab.1296 Sun J Lim Y Liu SQ Biosynthesis of flavor esters in coconut cream through coupling fermentation and lipase-catalyzed biocatalysis Eur J Lipid Sci Technol 2013 115 345 350 Tamilarasan K Kumar MD Optimization of medium components and operating conditions for the production of solvent-tolerant lipase by Bacillus sphaericus MTCC 7542 Afr J Biotechnol 2011 10 15051 15057 10.5897/AJB11.1143 Wawire M Oey I Mathooko F Njoroge C Shitanda D Hendrickx M Thermal stability of ascorbyl acid and ascorbyl acid oxidase in African cowpea leaves (Vigna unguiculata ) of different maturities J Agric Food Chem 2011 59 1774 1783 10.1021/jf103469n 21309563

PMC005xxxxxx/PMC5002272.txt

==== Front 3 Biotech3 Biotech3 Biotech2190-5738Springer Berlin Heidelberg Berlin/Heidelberg 49610.1007/s13205-016-0496-5Original ArticleEnhanced production of an anti-malarial compound artesunate by hairy root cultures and phytochemical analysis of Artemisia pallens Wall. Pala Zarna 12Shukla Vishnu 13Alok Anshu 13Kudale Subhash 1Desai Neetin neetindesai@gmail.com 141 School of Biotechnology and Bioinformatics, D. Y. Patil University, Navi Mumbai, India 2 Department of Biological Sciences, BITS Pilani, Pilani Campus, Pilani, Rajasthan India 3 National Agri-Food Biotechnology Institute, Govt. of India, Mohali, Punjab India 4 Amity Institute of Biotechnology, Amity University, Mumbai, India 27 8 2016 27 8 2016 12 2016 6 2 18215 12 2015 16 8 2016 © The Author(s) 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.Artemisinin and its derivatives are still one of the most effective drugs for the treatment of malaria. Artemisia pallens commonly known as Dhavanam, is an aromatic herb belonging to the family Asteraceae. Increasing the artemisinin content of A. pallens by genetic engineering would improve the availability of this much needed drug. In the present study, Agrobacterium rhizogenes (strain NCIM 5140) mediated genetic transformation of Artemisia pallens were carried out for hairy root induction. The effect of different media (Half MS, MS, MS along with BAP 0.5 mg/l and MS along with Kinetin 0.5 mg/l) and type of explants (leaf and stem) on hairy root induction and culture were also studied. Maximum transformation efficiency (70.0 %) was observed in case of stem explants when it was co-cultivated with Agrobacterium rhizogenes and kept on half strength MS media. Artesunate is a derivative of artemisinin, was quantified using HPLC from dried aerial extract and hairy roots. The content of artesunate in hairy roots was increased up to twofold as compared to aerial part of Artemisia pallens. The maximum amount of artesunate found in hairy roots was 5.62 ± 0.16 μg/g of dry weight. Apart from artesunate the other phytochemicals like alkaloids, polyphenols, and flavonoids are important because they impart the medicinal properties in this plant. Therefore, we have also quantified total alkaloids, flavonoids and polyphenolic content in the aerial part of the plants. The total alkaloids and flavonoids content were found 1.72 ± 0.00 mg/g dry weight in aqueous extract and 3.8 ± 0.00 mg/g in methanolic extract in terms of colchicine and rutin equivalents, respectively. Similarly, total phenolic content is 3.70 ± 0.01 mg/g in ethanolic extract in terms of tannic acid equivalent. Electronic supplementary material The online version of this article (doi:10.1007/s13205-016-0496-5) contains supplementary material, which is available to authorized users. Keywords ArtesunateArtemisia pallensDhavanamMalariaHairy rootsAgrobacterium rhizogeneissue-copyright-statement© King Abdulaziz City for Science and Technology 2016 ==== Body Introduction Artemisinin is a highly potent anti-malarial drug found in the medicinal plant Artemisia annua (A. annua). World Health Organization recommends artemisinin and its derivatives for the treatment of malaria (White 2008). Artemisia annua is a plant species of genus Artemisia L. and is used as a traditional medicine for the treatment of malaria and other diseases in China. A. pallens also known as “Davana” is an important aromatic herb of genus Artemisia L. and is mostly found in southern region of India. These plants are industrially important due to its anti-microbial, insecticidal, antioxidant, and anti-malarial properties as well as perfumery compounds (Haider et al. 2014). A. pallens has been employed by the local people as a herbal medicine in curing many disease like diabetes and some skin infections (Haider et al. 2014). The leaves and flowers of the A. pallens yield an essential oil known as “oil of Davana”. Further, it also indicates the presence of artemisinin and its derivatives which has been used as a drug for the treatment of malaria (Shukla et al. 2015). Recently, Agrobacterium tumefaciens mediated genetic transformation of A. pallens has been optimized, which suggested that genetic engineering of A. pallens could be the effective strategy for the production of artemisinin and its derivatives (Alok et al. 2016). Artemisinin and its derivatives are produced only in the aerial parts (especially leaves and floral parts) of the plant whereas non significant amount or absent in roots of some species of Artemisia (Ferreira and Janick. 1996; Mannan et al. 2010; Wang et al. 2016). Hairy root culture of plants using the Agrobacterium rhizogenes, the causative agent of hairy root disease in several plants, has emerged as an important technique for the production of secondary metabolites (Sivakumar et al. 2010; Sujatha et al. 2013). Transformed hairy root cultures are biochemically and genetically stable model for scale-up of pharmaceutically important natural products (Souret et al. 2003). Hairy roots have been reported to yield higher amounts of artemisinin than intact plant tissue and cell suspension cultures (Sivakumar et al. 2010; Dilshad et al. 2015a). Growth of hairy roots can be scaled up using bioreactors and hence they can be exploited for commercial production of secondary metabolites (Liu et al. 1998; Patra and Srivastava 2014). Various biotic and abiotic elicitors such as endophytic fungi (Wang et al. 2001a), red light (Wang et al. 2001b) and methyl jasmonate (Baldi and Dixit 2008) have been reported to regulate cell metabolism of hairy root for enhanced production of artemisinin. Among the plethora of phytochemicals, terpenes (terpenoids or isoprenoids) constitute the largest and most diverse class of specialized metabolites. Terpenoids, are the plant secondary metabolites which play an important role in plant–microbe, and plant–plant interactions (Dudareva et al. 2006). Terpenes are one of the largest groups of natural products; more than 25,000 terpene structures have been reported in different plant species (Gershenzon and Dudareva 2007). Terpene compounds have many functional roles in plants such as in photosynthesis (plastoquinones, chlorophylls, carotenoids), respiration (ubiquinone) as well as in growth and development of plant (sterols, cytokinins, gibberellins, abscisic acid, brassinosteroids) (Pulido et al. 2012). Natural bioactive compounds like alkaloids, flavonoids and polyphenols are the most important secondary metabolites in plants having properties that affect appearance, taste, odor and oxidative stability (Singh 2012). These compounds posses various important biological properties. Therapeutic potential of the extract of Artemisia species is directly related to the total polyphenolic and flavonoid content in those plants. Artemisinin is the drug of choice for the treatment of malaria and other diseases (White 2008). The highest artimisinin content has been reported in the leaves of A. annua (0.44–1.00 %), a Chinese variety of Artemisia plants (http://www.mmv.org/) (Mannan et al. 2010). It was introduced and cultivated in different regions of India, but artemisinin production was decreased (Singh et al. 1988; Baldi and Dixit 2008). There is a great concern that the artemisinin production at the current rate will not meet the increasing demand by the pharmaceutical industry. In past, various efforts have been made to enhance the level of these molecules in plants by the genetic engineering approach (Farhi et al. 2011). There are various strategies which are now being used to meet the increasing demand of artemisinin (Durante et al. 2011; Farhi et al. 2011; Singh et al. 2016). Therefore, present study was aimed to investigate efficacy of the hairy root induction in cultures of A. pallens, an Indian Artemisia species. We quantified artesunate, an important derivative of artemisinin in the hairy roots of A. pallens and aerial extract of plant. To the best of our knowledge, this is the first report of A. rhizogenes mediated genetic transformation of A. pallens. Further, we also quantified the total content of alkaloids, flavonoids and phenolic comounds in different extracts prepared in aqueous, ethanolic and methanolic solvents. Materials and methods Plant material, sterilization and media preparation Seeds of A. pallens were collected from Kolhapur, Maharashtra. Seeds were rinsed twice with distilled water; and then sterilized with 0.1 % mercuric chloride for 5 minutes. Finally, the seeds were rinsed with sterile distilled water for 3–4 times. The seeds were germinated on MS medium with 3 % sucrose and 0.8 % agar and pH maintained at 5.8. Bacterial strain and culture conditions A. rhizogenes strain NCIM 5140 (ATCC 15834) was obtained from the National Chemical Laboratory, Pune, India and was used for the induction of hairy root. The bacterial culture was revived and maintained on YEB agar medium. Loop full of bacterial colonies were inoculated in 100 ml of liquid YEB medium and the culture was kept on a rotary shaker (100 rpm) at 30 °C overnight till the O.D. at 600 nm was about 0.8. A. rhizogenes mediated genetic transformation Stem and leaves were dissected from in vitro germinated seeds and used as explants. The explants were pre-cultured for 3 days on MS basal medium (Murashige and Skoog 1962). Overnight grown culture of A. rhizogenes was centrifuged at 5000 rpm for 10 minutes at 26 °C. Pellet was re-suspended in 10 ml liquid MS medium and the O.D. at 600 nm was adjusted about 0.6. The pre-cultured explants were now co-cultured with A. rhizogenes grown in the flask. The flask was wrapped with foil and kept on shaker at 90 rpm for 30 minutes. Explants were blotted dry on a sterile filter paper and transferred on MS basal media and kept in dark. After 3 days of co-cultivation, the explants were transferred to medium containing 400 mg/l cefotaxime to kill the residual Agrobacterium. The explants were again sub-cultured on medium containing 300 mg/l cefotaxime after a week. The concentration of cefotaxime was reduced gradually to 100 mg/l. Optimization of hairy root induction and growth on different media The co-cultivated explants were incubated on different media [Half MS, MS, MS supplemented with 6-Benzylaminopurine (BAP) 0.5 mg/l and MS supplemented with Kinetin 0.5 mg/l] for proper induction and growth of hairy roots. The MS basal medium contained 2 mg/l thiamine hydrochloride, 5 mg/l nicotinic acid, 10 mg/l pantothenic acid, 30 g/l sucrose, and 100 mg/l myo-inositol. All media were solidified with 0.8 % agar and the pH was adjusted to 5.8 with 1 N NaOH or 1 N HCl prior to sterilization at 121 °C for 20 minutes. All the cultures were maintained in complete darkness at 26 ± 1 °C. The response of explants (leaf and stem) in different media was recorded. The final and bulk hairy root induction was performed on the best responded media. The 100 mg/l cefotaxime was also used in all different media combinations to eliminate the residual bacteria. PCR analysis of hairy roots The Polymerase chain reaction (PCR) was used to detect the Ri T-DNA integration into genome of A. pallens (Eppendorf Mastercycler®, Germany). Genomic DNA from putative transgenic hairy roots as well as non-transgenic roots (in vitro plants roots) was extracted as per manufacturer protocol using HiPurA™ Plant Genomic DNA Miniprep Purification Kit (Himedia, India). The transgenic nature of the roots was confirmed by amplification of mannopine synthase 1′ (mas 1′) sequence using mas(For): 5′CGGTCTAAATGAAACCGGCAAACG3′ and mas(Rev) 5′GGCAGATGTCTATCGCTCGCACTCC3′ as mentioned by (Telke et al. 2011). Further confirmed by amplification of rolC gene using rolC(For):5′ATGGCTGAAGACGACCTGTGTT3′ and rolC(Rev):5′TTAGCCGATTGCAAACTTGCTC3′ as mentioned by (Jha et al. 2013). The PCR reactions were carried out in a total 25 µl volume and consisted of 50 ng of DNA, 2.5 μl 10× taq DNA polymerase buffer, 50 μM dNTPs, 0.2 μM primers and 0.5 U Taq DNA polymerase (Banglore Genei Pvt. Limited). DNA amplification was performed on an Eppendorf thermal cycler with following PCR cycle conditions: 94 °C for 2 minutes, 94 °C for 30 s, and 55 °C for 30 s, 72 °C for 90 s and final extension at 72 °C for 5 minutes. Amplification products were separated by electrophoresis on 1 % agarose gel in 1× TBE buffer, stained with ethidium bromide and visualized under UV trans-illuminator. Extraction and estimation of artesunate using HPLC The extraction process for the HPLC analysis of samples was optimized with different solvents. We choose methanol: water combination for the extraction of our samples as it gave the best response with both standard and the sample. The powdered plant material was repeatedly extracted three times with methanolic extract (methanol: water combination) on an orbital shaker for overnight incubation at room temperature. The pooled plant extract was centrifuged at 10,000 rpm for 10 min at room temperature and the supernatant was collected and filtered before HPLC analysis. The HPLC system equipped with a Waters 1525 binary pump and Waters 2487 dual wavelength absorbance detector was used for artesunate quantification. A known amount of artesunate (Sigma Aldrich) was dissolved in mobile phase, centrifuged and supernatant was injected in HPLC. The HPLC analysis of the samples was carried out as previously described protocol (Affum et al. 2013). The reverse phase C18 column of dimensions 0.39 mm × 150 mm (i.d.) and 5 μm particle size was used in the analysis. Before analysis, the column was equilibrated with a mobile phase prepared from acetonitrile and phosphate buffer at pH 3 (44:56 % v/v). 20 μl of the sample was injected into HPLC instrument with a flow rate of 1.0 ml/min for 10 min using an isocratic elution. The chromatograms of each sample were recorded at 216 nm. Artesunate content in the extracts was determined by comparing the peak areas of the sample with those of standard artesunate. Extraction of alkaloids, flavonoids and poyphenols Aerial part of A. pallens weighed 10 g and then it was crushed to fine powder in mortar and pestle using liquid nitrogen. Three different extracts were prepared in 100 ml 99.9 % ethanol, 99.9 % methanol, and disstiled water. This mixture of 10 g powdered plant material and 100 ml of its respective solvent was kept overnight for incubation. The mixture was then centrifuged at 10,000 rpm for 10 min at room temperature. The supernatant was collected in different tubes and labelled correctly. Determination of total alkaloids Alkaloid content of the different extracts was estimated using colchicine as standard and protocol followed as per mentioned by (Singh et al. 2004). Ethanol was dissolved and prepared 0.05 M 1,10-Phenanthroline monohydrate and 0.5 M HCl was used to prepare 0.025 M FeCl3. From each solution 1 ml were added to 1 % extract and final volume was made up to 10 ml with distilled water. The tubes were incubated at 70 ± 2 °C for 30 min. O.D. was taken at 510 nm using UV–VIS spectrophotometer. Once the O.D. is obtained then the standard graph for alkaloids is prepared and the concentrations of the samples are calculated following the Beer Lambert’s Law. The concentration of alkaloids in the test extracts was calculated from the calibration plot and expressed as milligrams of colchicines equivalent per gram dry weight of sample (mg CE/g). Measurement of total flavonoids Rutin was used as the standard compound for the estimation of flavonoids and protocol used by (Zhishen et al. 1999) was followed. A known volume of 1 % extract was placed in a 10 ml volumetric flask containing 3 ml distilled water and 0.3 ml NaNO2 (1:20) were added. After 5 min later, 3 ml AlCl3 (1:10) were added. Finally after 6 min, 2 ml 1 M NaOH was added and the total volume was made up to 10 ml with distilled water. The solution was mixed well again absorbance was measured against a blank at 510 nm using UV–VIS spectrophotometer. Rutin was used to make the standard calibration curve by making different dilutions. The total flavonoid content of each sample was determined from the standard curve of Rutin. The total flavonoid content was expressed as milligrams of rutin equivalent per gram dry weight of sample (mg RE/g). Determination of total poyphenols The total phenolic content of extract A. pallens was determined by using Folin-Ciocalteu reagent with slightly modified method as prescribed by (Ainsworth and Gillespie 2007). Different dilution of tannic acid was used for standard calibration curve. A volume of 0.5 ml of the 1 % extract was mixed with 2 ml of the Folin-Ciocalteu reagent (diluted 1:10 with de-ionized water) and finally neutralized with 4 ml of sodium carbonate solution (7.5 % w/v). The reaction mixture was incubated for 30 min at room temperature. The absorbance of the resulting blue color was measured at 765 nm using double beam UV–VIS spectrophotometer. The total polyphenolic contents were determined from the linear equation of a standard curve prepared with tannic acid. The total flavonoid content was expressed as milligrams of tannic acid equivalent per gram dry weight of sample (mg TAE/g). Quantification of different polyphenols using HPLC Separation of tannic acid, cathacol, ferulic acid and vaniline in the crude extract of A. pallens was achieved on HPLC system equipped with a PDA Detector. Xterra MSC-18 column (7.8 × 100 mm, 5 µm) with octadecylsilane as a solid support and 600e Multi Solvent Delivery System from Waters (USA) was used to separate the components. The mobile phase consists of methanol + acetonitrile:phosphate (30:15:55, v/v) (pH 3.5) and flow rate of the mobile phase was kept at 1.0 ml/min with isocratic elution. A linear gradient elution was used and detection was done at 280 nm in this method. The separated alkaloids, flavonoids and phenolic content were initially identified by a direct comparison of their retention times with those of standards. The contents of all compounds were calculated from the peak areas of HPLC chromatograms from the three replicate samples and the output was given in the units of ppm. The results were converted from ppm to mg/g. The data were analysed and processed using the installed Empower 2 software. Results and discussion Induction and establishment of hairy root cultures Agrobacterium rhizogenes strain, NCIM 5140 was used to determine the transformation efficiency towards hairy root induction in A. pallens. Leaf and stem of 4 weeks old in vitro germinated seedling were used as explants for hairy root induction. After co-cultivation of explants with A. Rhizogenes, the explants were kept on different media. The transformation efficiency of hairy root induction on different media is shown in Table 1. The best response of hairy root induction was found in all media when stem was co-cultivated as compared to leaf. The some leaf explants turned to callus like structure on MS media with thick and short hairy roots (Fig. 1a) whereas roots raised from stem were long and thin (Fig. 1b). Maximum transformation efficiency (70.0 %) was observed in case of stem explants when it was kept on half strength MS media for hairy root induction (Fig. 1c). The response of hairy root induction was very less (30.75 %) in case of leaf when it was incubated on MS media supplemented with 0.5 mg/l Kinetin (Fig. 1d). In all combinations hairy roots obtained from stem explants were very thin, long and appeared very weak whereas, roots obtained from the leaf explants were thick and short. After 5 weeks of growth, roots obtained from both the explants appeared same. Influence of A. rhizogenes on hairy root induction frequency has been documented earlier in several plant species. In A. vulgaris maximum of 92.6 % transformation frequency was observed in leaf explants followed by 64.3 % in node and 38.1 % in shoot tip using strain, A4GUS (Sujatha et al. 2013). Transformation efficiency depends upon type of explants and strain of A. rhizogenes. However, NCIM 5140 strain of A. rhizogenes was widely used for hairy root induction in many plants such as Linum album (Baldi et al. 2008), Abrus precatorius L. (Karwasara and Dixit 2009), Helianthus annuus L. (Jha et al. 2013) and Sesuvium portulacastrum L. (Lokhande et al. 2015).Table 1 Frequency of hairy root induction on different explants and media of A. pallens by Agrobacterium rhizogenes strains, NCIM 5140 Type of explants and different media No. of explants infected No. of explants transformed Transfomation efficiency (%) Leaf (half MS) 30 15 50.00 Stem (half MS) 50 35 70.00 Leaf (MS) 46.5 25 53.76 Stem (MS) 27 17 59.25 Leaf (MS + 0.5 mg/l BAP) 90 30 33.33 Stem (MS + 0.5 mg/l BAP) 80 50 62.5 Leaf (MS + 0.5 mg/l Kinetin) 65 20 30.75 Stem (MS + 0.5 mg/l Kinetin) 60 40 66.66 Fig. 1 Hairy root response after 30 days of co-cultivation on different media composition. a Leaf as explants on MS media, b stem as explants on MS media, c stem as explants on half strength MS media and d leaf as explants on MS media supplemented with 0.5 mg/l Kinetin PCR based analysis of transgenic hairy roots The transgenic nature of hairy roots was confirmed by PCR amplification of DNA from A. pallens hairy roots with mas gene specific primers, which showed the expected fragment size of 970 bp (Fig. 2b). 970-bp fragment amplification was observed in all the transformants; while it was absent in the non-transformed control roots. Rol C gene specific primers were also used for further confirmation of integration of T-DNA from the A. rhizogenes into the hairy roots. PCR analysis showed amplification of 500 bp band of Rol C gene integrated in hairy roots obtained from leaf and stem explants (Fig. 2a). Genomic DNA of strain NCIM 5140 was used as positive control where as negative control (non-transformed roots) there was no amplification (Fig. 2a, b). In Ri plasmids, the T-DNA region consists of two parts: TL-DNA and TR-DNA separated by a non-transferable DNA sequence. TR DNA consists of mas 1′ where as TL-DNA consist rol gene, which is is essential to induce hairy roots (Hansen et al. 1991). Therefore we used both gene specific primers to confirm transgenic nature of our hairy root culture. Similarly, both this primes are used to demonstrate transgenic nature of Sesuvium portulacastrum L. by the same strain A. Rhizogenes (Lokhande et al. 2015).Fig. 2 PCR detection of rol C and mas1′gene in hairy roots. a Lane L1, GeneRuler 1 kb DNA Ladder; lane L2, water control; lane L3, non-transformed roots (negative control); lane L4, rol C positive control (Ri plasmid); lane L5 to L7, rol C from transformed hairy roots. b Lane L1, NEB 1 kb DNA Ladder; lane L2, water control; lane L3, non-transformed roots (negative control); lane L4, mas1′positive control (Ri plasmid); lane L5 to L7, mas1′from transformed hairy roots Determination of artesunate in hairy roots through HPLC coupled with PDA The HPLC chromatograph of standard and artesunate from the dried aerial part and hairy root of A. pallens is depicted in Supplementary Fig. 1. The results confirmed a considerable amount of artesunate (5.62 ± 0.16 μg/g of DW) in the hairy roots where as in dried aerial part and (1.88 ± 0.10 μg/g of DW) was found Table 3. The content of artesunate in A. annua it was approx 10 μg/g (Dilshad et al. 2015a) whereas in A. carvifolia is found 2.24 μg/g (Dilshad et al. 2015b). So, it can be concluded that the hairy roots are a good source of accumulation of secondary metabolites, in this case artesunate which is an important anti-malarial compound. The accumulation of high content of artesunate in hairy roots of A. pallens could be due to the effect of rol gene integration on the secondary metabolite production as reported in many plant species (Bulgakov 2008). The transformation of rol ABC gene construct in A. dubia enhanced over production of artemisinin in i.e., up to tenfold than individual rol genes as we observed six- to sevenfold increase (Kiani et al. 2012). Recently, the transformation of A. annua with rol B gene showed two- to ninefold increase in artemisinin in different transgenic lines. They have also quantified artesunate in different transgenic line and they found 4- to 12-fold increase in artesunate content (Dilshad et al. 2015a). The artemisinin and its derivative such as artesunate, and dihydroartemisinin etc. are very effective for malaria, and therefore recently, theses metabolites were isolated and quantified from A. carvifolia Buch. These metabolites were quantified in the shoots extract of A. carvifolia at the following concentrations: artemisinin (8 μg/g), artesunate (2.24 μg/g), dihydroartemisinin (13.6 μg/g) and artemether (12.8 μg/g). Genetic transformation of A. carvifolia using rol B and C genes increased three- to sevenfold artemisinin, three- to tenfold for artesunate and 2.6- to 4-fold for dihydroartemisinin and artemether in transgenics lines (Dilshad et al. 2015b). Quantification of alkaloids, flavonoids and phenolics content Standard curve prepared was used for the determination of total alkaloids, flavonoids and phenolic content of A. pallens using different concentrations of colchicine, rutin and tannic acid respectively. The total alkaloids, flavonoids and phenolic content in different solvent of A. pallens have been presented in Table 2. Observation shows that the total alkaloids content was almost similar, however highest alkaloids content was found in aqueous extract (1.72 ± 0.00 mg/g CE of dried extract). The concentration of flavonoids is significantly high (3.8 ± 0.00 mg/g RE of dried extract) in methanolic extract as compared to ethanolic extracts (2.13 ± 0.56 mg/g RE of dried extract). Similarly, total phenolic content is highest in the ethanolic extract (3.7 ± 0.01 mg/g TAE of dried extract) followed by methanolic and significantly lower in aqueous extract (3.0 ± 0.00 mg/g TAE of dried extract).Table 2 The total alkaloids, flavonoids and phenolics content present in different extracts of A. Pallens Parameters Aqueous extract Ethanolic extract Methanolic extract Total alkaloids content (mg/g CE) 1.72 ± 0.00 1.68 ± 0.01 1.67 ± 0.00 Total flavoinds content (mg/g RE) 3.70 ± 0.1 2.13 ± 0.56 3.80 ± 0.00 Total phenolics content (mg/g TAE) 3.00 ± 0.00 3.70 ± 0.01 3.20 ± 0.01 Values are expressed as mean ± SE of three replicates Further, the different phenolics compounds such as tannic acid, catechol, vanillin, and ferulic acid was also quantified in plant extract using HPLC and the chromatograph is depicted in Supplementary Fig. 2. Phenolics are aromatic benzene ring compounds that have one or more hydroxyl groups. These individual phenolic compounds were extracted in methanolic extract (Table 3). The content of Tannic acid (3. 41 ± 0.04), Cathacol (5.37 ± 0.03) and Ferulic Acid (1.94 ± 0.00); whereas Vaniline was not detected (Table 4).Table 3 Quantification of artesunate (μg/g of DW) using HPLC of dried areial part and hairy roots of A. pallens Tissue type Content of artesunate (μg/g of DW) Hairy roots 5.62 ± 0.16 Aerial parts 1.88 ± 0.10 Table 4 Quantification of different polyphenols using HPLC of dried areal part of A. pallens Parameters Methanolic extract Tannic acid (μg/g) 3.41 ± 0.04 Cathacol (μg/g) 5.37 ± 0.03 Ferulic acid (μg/g) 1.94 ± 0.00 Vaniline (μg/g) ND Conclusion The overall goal of this study was to induce hairy roots in A. pallens and quantified the content of secondary metabolites produced in hairy roots and areal parts. A. pallens are of great medicinal importance in diabeteus mellitus. Apart from these the presence of artesunate also confirmed it can use for as an alternative system for the production of artemisinin. Further the total content of alkaloids flavanoids, and phenolic compounds was also quantified. These natural bioactive compounds like alkaloids, flavonoids and phenols are the important secondary metabolites in plants. These compounds have intrinsic properties that affect appearance, taste, odor and oxidative stability of plant based foods. To the best of our knowledge this is the first report on hairy root induction and quantification of artesunate from A. pallens. This study provides basic platform to initiate large scale production of artemisinin and its derivatives and needs further standardization for industrial application. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary material 1 (JPEG 1282 kb) Supplementary material 2 (JPEG 1236 kb) The authors express their gratitude for providing research facilities to School of Biotechnology and Bioinformatics, D. Y. Patil University, Navi Mumbai, India. We thank Dr. Sandeep Kale for providing HPLC facility at DBT-ICT Centre for Energy Biosciences, Institute of Chemical Technology, Mumbai, India. Compliance with ethical standards Conflict of interest The authors declare that they have no conflict of interest in the publication. ==== Refs References Affum AO Lowor S Osae SD Dickson A Gyan BA Tulasi D A pilot study on quality of artesunate and amodiaquine tablets used in the fishing community of Tema, Ghana Malar J 2013 12 220 10.1186/1475-2875-12-220 23809666 Ainsworth EA Gillespie KM Estimation of total phenolic content and other oxidation substrates in plant tissues using Folin–Ciocalteu reagent Nat Protoc 2007 2 875 877 10.1038/nprot.2007.102 17446889 Alok A, Shukla V, Pala Z, Kumar J, Kudale S, Desai N (2016) In-vitro regeneration and optimization of factors affecting Agrobacterium mediated transformation in Artemisia Pallens, an important medicinal plant. Physiol Mol Biol Plants 22:261–269. doi: 10.1007/s12298-016-0353-3 Baldi A Dixit VK Yield enhancement strategies for artemisinin production by suspension cultures of Artemisia annua Bioresour Technol 2008 99 4609 4614 10.1016/j.biortech.2007.06.061 17804216 Baldi A Srivastava AK Bisaria VS Improved podophyllotoxin production by transformed cultures of Linum album Biotechnol J 2008 3 1256 1263 10.1002/biot.200800173 18932162 Bulgakov VP Functions of rol genes in plant secondary metabolism Biotechnol Adv 2008 26 31824 10.1016/j.biotechadv.2008.03.001 Dilshad E Cusido R Palazon J Enhanced artemisinin yield by expression of rol genes in Artemisia annua Malar J 2015 14 424 10.1186/s12936-015-0951-5 26510528 Dilshad E Cusido RM Estrada KR Genetic transformation of Artemisia carvifolia Buch with rol genes enhances Artemisinin accumulation PLoS One 2015 10 e0140266 10.1371/journal.pone.0140266 26444558 Dudareva N Negre F Nagegowda DA Orlova I Plant volatiles: recent advances and future perspectives CRC Crit Rev Plant Sci 2006 25 417 440 10.1080/07352680600899973 Durante M Caretto S Quarta A De Paolis A Nisi R Mita G β-Cyclodextrins enhance artemisinin production in Artemisia annua suspension cell cultures Appl Microbiol Biotechnol 2011 90 1905 1913 10.1007/s00253-011-3232-4 21468706 Farhi M Marhevka E Ben-Ari J Algamas-Dimantov A Liang Z Zeevi V Edelbaum O Spitzer-Rimon B Abeliovich H Schwartz B Tzfira T Vainstein A Generation of the potent anti-malarial drug artemisinin in tobacco Nat Biotechnol 2011 29 1072 1074 10.1038/nbt.2054 22158354 Ferreira JFS Janick J Roots as an enhancing factor for the production of artemisinin in shoot cultures of Artemisia annua Plant Cell Tissue Organ Cult 1996 44 211 217 10.1007/BF00048526 Gershenzon J Dudareva N The function of terpene natural products in the natural world Nat Chem Biol 2007 3 408 414 10.1038/nchembio.2007.5 17576428 Haider SZ Mohan M Andola HC Constituents of Artemisia indica Willd. from Uttarakhand Himalaya: a source of davanone Pharmacogn Res 2014 6 257 259 10.4103/0974-8490.132607 Hansen G Larribe M Vaubert D Agrobacterium rhizogenes pRi8196 T-DNA: mapping and DNA sequence of functions involved in mannopine synthesis and hairy root differentiation Proc Natl Acad Sci U S A 1991 88 7763 7767 10.1073/pnas.88.17.7763 1909028 Jha P Jobby R Kudale S Biodegradation of phenol using hairy roots of Helianthus annuus L. Int Biodeterior Biodegradation 2013 77 106 113 10.1016/j.ibiod.2012.11.001 Karwasara VS Dixit VK Agrobacterium rhizogenes mediated genetic transformation of Abrus precatorius L. Pharmacogn Mag 2009 5 336 342 10.4103/0973-1296.58563 Kiani BH Safdar N Mannan A Mirza B Comparative Artemisinin analysis in Artemisia dubia transformed with two different Agrobacteria harbouring rol ABC genes Plant Omics J 2012 5 386 391 Liu CZ Wang YC Ouyang F Production of artemisinin by hairy rot cultures of Artemisia annua L. in bioreactor Biotechnol Lett 1998 20 265 268 10.1023/A:1005382020039 Lokhande VH Kudale S Nikalje G Hairy root induction and phytoremediation of textile dye, Reactive green 19A-HE4BD, in a halophyte, Sesuvium portulacastrum L. Biotechnol Rep 2015 8 56 63 10.1016/j.btre.2015.08.002 Mannan A Ahmed I Arshad W Survey of artemisinin production by diverse Artemisia species in northern Pakistan Malar J 2010 9 310 10.1186/1475-2875-9-310 21047440 Murashige T Skoog F A revised medium for rapid growth and bioassays with tobacco tissue cultures Physiol Plant 1962 15 473 497 10.1111/j.1399-3054.1962.tb08052.x Patra N Srivastava A Enhanced production of artemisinin by hairy root cultivation of Artemisia annua in a modified stirred tank reactor Appl Biochem Biotechnol 2014 174 2209 2222 10.1007/s12010-014-1176-8 25172060 Pulido P Perello C Rodriguez-Concepcion M New insights into plant isoprenoid metabolism Mol Plant 2012 5 964 967 10.1093/mp/sss088 22972017 Shukla V Pala Z Alok A Desai N Screening of different Artemisia spp. from Western Ghats of Maharashtra for an anti-malarial compound-artemisinin Am J Plant Sci 2015 06 1619 1632 10.4236/ajps.2015.69162 Singh R Total phenolic, flavonoids and tannin contents in different extracts of Artemisia absinthium J Intercult Ethnopharmacol 2012 1 101 104 10.5455/jice.20120525014326 Singh A Vishwakarma RA Husain A Evalutation of Artemisia annua strains for higher Artemisinin production Planta Med 1988 54 475 476 10.1055/s-2006-962515 17265330 Singh DK Srivastava B Sahu A Spectrophotometric determination of Rauwolfia alkaloids : estimation of reserpine in pharmaceuticals Anal Sci 2004 20 571 573 10.2116/analsci.20.571 15068309 Singh ND Kumar S Daniell H Expression of β-glucosidase increases trichome density and artemisinin content in transgenic Artemisia annua plants Plant Biotechnol J 2016 14 1034 1045 10.1111/pbi.12476 26360801 Sivakumar G Liu C Towler MJ Weathers PJ Biomass production of hairy roots of Artemisia annua and Arachis hypogaea in a scaled-up mist bioreactor Biotechnol Bioeng 2010 107 802 813 10.1002/bit.22892 20687140 Souret FF Kim Y Wyslouzil BE Scale-up of Artemisia annua L. hairy root cultures produces complex patterns of terpenoid gene expression Biotechnol Bioeng 2003 83 653 667 10.1002/bit.10711 12889030 Sujatha G Zdravković-Korać S Ćalić D High-efficiency Agrobacterium rhizogenes -mediated genetic transformation in Artemisia vulgaris : hairy root production and essential oil analysis Ind Crops Prod 2013 44 643 652 10.1016/j.indcrop.2012.09.007 Telke A Kagalkar A Jagtap U Biochemical characterization of laccase from hairy root culture of Brassica juncea L. and role of redox mediators to enhance its potential for the decolorization of textile dyes Planta 2011 234 1137 1149 10.1007/s00425-011-1469-x 21735196 Wang J Zhang Z Tan R Stimulation of artemisinin production in Artemisia annua hairy roots by the elicitor from the endophytic Colletotrichum sp. Biotechnol Lett 2001 23 857 860 10.1023/A:1010535001943 Wang Y Zhang H Zhao B Yuan X Improved growth of Artemisia annua L. hairy roots and artemisinin production under red light conditions Biotechnol Lett 2001 23 1971 1973 10.1023/A:1013786332363 Wang S Towler MJ Weathers PJ Root regulation of artemisinin production in Artemisia annua : trichome and metabolite evidence Planta 2016 White NJ Qinghaosu (artemisinin): the price of success Science 2008 320 330 334 10.1126/science.1155165 18420924 Zhishen J Mengcheng T Jianming W The determination of flavonoid contents in mulberry and their scavenging effects on superoxide radicals Food Chem 1999 64 555 559 10.1016/S0308-8146(98)00102-2

PMC005xxxxxx/PMC5002273.txt

==== Front AMB ExpressAMB ExpressAMB Express2191-0855Springer Berlin Heidelberg Berlin/Heidelberg 22710.1186/s13568-016-0227-7Original ArticleSophorolipids production from rice straw via SO3 micro-thermal explosion by Wickerhamiella domercqiae var. sophorolipid CGMCC 1576 Liu Xin-ge ++ 86-187-8883-542218788835422@163.com 1Ma Xiao-jing ++ 86-531-88364715maxj@hfut.edu.cn 1Yao Ri-sheng ++ 86-155-5698-2575yaors@hfut.edu.cn 1Pan Chun-yu ++ 86-155-5698-257515556982575@163.com 1He Hua-bing ++ 86-552-4928335804727862@qq.com 21 School of Biological and Medical Engineering, Hefei University of Technology, 193 Tunxi Road, Hefei, 230009 Anhui China 2 Anhui BBCA Chemical Equipment Co. LTD, Bengbu, 233010 China 27 8 2016 27 8 2016 2016 6 1 6031 7 2016 10 8 2016 © The Author(s) 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.A novel lignocellulose material, holocellulose from rice straw via the pretreatment of SO3 micro-thermal explosion, was developed to produce sophorolipids (SLs) with Wickerhamiella domercqiae var. sophorolipid CGMCC 1576. The influence factors of inoculum dose, yeast extract concentration and pH regulators (chemical regents used for adjusting/influencing pH) was investigated and discussed. Results showed that W. domercqiae can grow in the rice straw holocellulose hydrolysate, and acquire relative high SL yield of 53.70 ± 2.61 g/L in shake flask culture. Inoculum dose, yeast extract concentration and pH regulator made obvious influence on fermentation parameters, especially on final broth pH and SLs production. Furthermore, there is a strong negative linear correlation existing between final broth pH and lactonic SL or ratio of lac SL/tot SL. Additionally, comparison between SL production and non-glucose carbon sources, culture methods, microbes in previous reports was carried out. These results will be benefit for acquiring SL mixture with suitable lac SL/tot SL ratio for specific purpose and scope economically. Keywords SophorolipidRSHHLac SL/tot SLBroth pHCorrelationNational High Technology Research and Development Program of China2014AA021902Yao Ri-sheng National Natural Science Foundation of China (CN)31400049Ma Xiao-jing http://dx.doi.org/10.13039/501100002858China Postdoctoral Science Foundation2015T806462013M531501Ma Xiao-jing issue-copyright-statement© The Author(s) 2016 ==== Body Introduction Sophorolipids (SLs), an extracellular biosurfactant, are being applied in fields of detergent (Cox et al. 2013; Lee et al. 2014), petroleum (Sirola 2010), cosmetic (Hillion et al. 1998; Morya et al. 2013), pharmaceutical (Chen et al. 2006; Morya et al. 2013; Singh et al. 2016), food processing (Cooper and Paddock 1984; Liu et al. 2009), environment industries (Sirola 2010), nano material (Pandey et al. 2016) and fermentation engineering (Gross et al. 2015) for their potential characteristics. They have comparable surface activity, biodegradability, biocompatibility, and low toxicity compared with chemical surfactants (Banat et al. 2000; Lee et al. 2008). Nonpathogenic yeasts of Candida apicola, Rhodotorula bogoriensis, Torulopsis gropengiesseri, Candida bombicola, Wickerhamiella domercqiae, Torulopsis bombicola, Pichia anomala, Candida batistae, Candida rugosa, Candida stellata, Candida floricola etc. were reported to produce SLs (Kurtzman et al. 2010; Imura et al. 2010; Chandran and Das 2011; Bogaert and Soetaert 2011). Ordinary, SLs occur as a mixture of lactonic and acidic SL molecules, lactonic SLs show better surface tension reducing propriety and biological activities, while the acidic ones display better foam formation ability and solubility and allow further modification at the free carboxylic acid end (Concaix 2003; Chen et al. 2006; Hu and Ju 2001; Maingault 1999). Although SLs was first published in 1961 (Gorin et al. 1961), they gained increased attention as growing environmental awareness in the last two decades. Currently, both choosing new substrates to lower production cost of SLs and selectively synthesis of SLs by optimizing fermentation conditions are hot topics. SLs-producing yeasts can grow on variety of saccharic and lipidic feed-stocks to accumulate SL mixture. Glucose and oleic acid are common used substrates. One of major study on SLs is to explore new resource to replace glucose. Non-lignocellulose material, such as whey (Otto et al. 1999), soy molasses (Solaiman et al. 2004, 2007), honey (Pekin et al. 2005), cassava starch (Thaniyavarn et al. 2008), sugarcane molasses (Daverey and Pakshirajan 2009; Takahashi et al. 2011), sweet water (Wadekar et al. 2012), glycerol (Bhangale et al. 2014) had been studied previously. However, SL yields were mostly lower when different sugars or low-cost substrates were employed, and the cost-reducing effects were inconspicuous compared to SLs obtained from glucose. Lignocellulosic material is one of the most abundant renewable resources. Hydrolysate of delignified corncob residue (DCCR) was firstly introduced for SLs production by Ma et al. (2013). Total SL yield of 32.79 g/L and 36.25 g/L was obtained from DCCR without/with detoxification, respectively. This report expanded the range of raw materials for SLs production. Furthermore, to eliminate the wastewater appended in DCCR, Masaaki Konishi et al. (2015) developed SLs from corncob hydrolysate (CCR), with a SL yield of 49.20 g/L. They also found that excess sulfuric acid used during corncob pretreatment would increase the content of furfural in hydrolysate. Additionally, Samad et al. (2014) obtained 84.60 g/L of SLs from bagasse by continuous fermentation with a fermentation time of 240 h. Inhibition effects of lignin on SLs production were also verified. Hence, lignocellulosic materials are potential resources for SL production and pretreatment was essential for production improving. Although corncob is easier to be utilized than other raw materials, rice straw and wheat straw are the most abundant biomass resources to be exploited and used in the world (Binod et al. 2010). In order to make cellulose more accessible, pretreatment of straw was used to break the lignin seal and disrupt the crystalline structure of cellulose. The pretreatment method of “SO3 micro-thermal explosion” was proposed by our group and could be applied to pretreat rice straw, wheat straw and other lignocellulose. Preliminary work has proved that this method has advantages of high lignin removal rate, low consumption, less inhibitors and little wastewater generating (Yao et al. 2011; Li et al. 2012). For research influence factors on SLs production and lac SL/tot SL ratio in SL mixture, some studies regarding carbon sources, nitrogen sources, pH, cultivation methods and strains, etc. were available (Cooper and Paddock 1984; Zhou et al. 1992; Casas and García-Ochoa 1999; Göbbert et al. 1984; Stüwer et al. 1987). Our previous work also indicated that inorganic nitrogen sources could significantly inhibit lactonic SL production and addition of pH regulators could enhance lactonic SL production by increasing broth pH value for W. domercqiae. However, further research is needed to more accurately assess the potential beneficial and harmful effects of nitrogen source and pH regulator on SLs production. The aim of the present work was to explore inexpensive substrate for SL production and obtain SL mixture with appropriate lac SL/tot SL ratio for various use. Firstly, rice straw pretreated by the method of SO3 micro-thermal explosion was introduced for SLs production. Then, the fermentation process and the influence factors on SLs accumulation by W. domercqiae in rice straw holocellulose hydrolysate (RSHH) were investigated and discussed. Finally, the relationship between the final broth pH and lactonic SL or lac SL/tot SL ratio was fitted. The current study is an attempt to address substrates and main parameters of SL fermentation to arrive at an economically SL-producing procedure. Materials and methods Rice straw pretreatment and enzymatic hydrolysis Rice straw was obtained from a local farm in the suburb of Hefei, China, which was harvested in 2015 and pretreated according to our previous study. Firstly, rice straw was cut into small pieces of about 2–3 cm in length. Then holocellulose was obtained with the pretreatment of SO3 micro-thermal explosion, ammonia wash, alkaline wash, water wash and drying processes, successively (Yao et al. 2011). For enzymatic hydrolysis experiments, crude KDN cellulase solution (KDN Biotech Co., Ltd., Qingdao, China) with filter paper activity (FPA) of 67 FPU/mL was used. Before enzymatic hydrolysis, rice straw holocellulose (RSH) was desiccated at room temperature and subsequently milled to power in preparation. RSHH was obtained by hydrolyzing 10 % of RSH with the enzyme dose of 25 FPU/g of dry RSH in 0.2 M sodium acetate buffer (pH 4.8) at 45 °C and 150 rpm for 72 h. After hydrolysis, the liquid fraction was collected by centrifugation and the cellulase was inactivated at 80 °C for 0.5 h. All RSHH was stored at 4 °C prior to use. Microorganism and growth condition Wickerhamiella domercqiae var. Sophorolipid CGMCC 1576 was offered by Professor Song of Shandong University and now preserved in China General Microbiological Culture Collection Center (CGMCC). The seed inoculum was prepared by growing the organism in seed medium on a rotary shaker at 220 rpm for 16 h at 30 °C. The seed medium contained glucose 20, peptone 20 and yeast extract 10 (g/L). Then the seed culture was transferred to different fermentation media (50 mL in 300 mL flask) and cultivated for 7 days at 220 rpm at 30 °C. The chemical defined fermentation medium, which containing ingredients (w/v, g/L) of glucose 60.00, yeast extract 3.00, KH2PO4 1.00, Na2HPO4·12H2O 1.00, MgSO4·7H2O 0.05, and oleic acid 60.00 (v/v, mL/L), was used for SLs production and set as the control group. For hydrolysate fermentation medium, rice straw holocellulose hydrolysate (RSHH) which containing 60.00 g/L of glucose was used to replace glucose as carbon source. Then, effects of different seed inoculum dose, yeast extract concentration and pH regulator on SLs production and lac SL/tot SL ratio in RSHH fermentation medium was investigated. The experiment was carried out in triplicate, respectively. Different seed inoculum dose of 1, 2, 4 %, yeast extract in different concentration (g/L, 0.00, 0.05, 0.10, 0.15, 0.20, 0.25, 0.30) and different pH regulator (NaOH, K2CO3, CaO, AlCl3, C6H8O7, Na3C6H5O7·2H2O) were added at the beginning of the cultivation. The other ingredients in RSHH medium were the same as that in the control group. Analysis of chemical and elementary composition The chemical composition of rice straw and obtained holocellulose was analyzed by the method of NREL. The elementary analysis was determined by elemental analyzer (Elementar, Germany). The experiments of chemical composition and elementary analysis all were carried out in triplicate and the results reported are mean ± SD of the three independent experimental results. Analysis of scanning electron microscope images Scanning electron microscope (SEM, Hitachi, Japan) instrument was used to observe morphology variation of rice straw and confirm the pretreatment effects of SO3 micro-thermal explosion. The images were collected under the magnification of 1000 and 200 times, at the accelerating voltage of 20 kV. Determination of residual glucose, pH, biomass, SL production Glucose in hydrolysate and residual glucose in broth was measured by SBA-40C bio-sensor analyzer (Shandong Academy of Sciences, Shandong, China) with an injection volume of 25 μL. Broth pH value was measured by PHS-3C pH meter (Shanghai Shengci Instrument co., LTD, China). For biomass determination, 1.0 mL of culture broth was mixed with two volumes of n-butanol/ethanol/chloroform (10:10:1), and centrifuged at 8000 rpm for 10 min. The solid residue was washed twice with distilled water, and dried at 50 °C to a constant weight. All data were presented as the mean of three readings. For lactonic SL determination, two volumes of ethyl acetate were added to 0.5 mL fermentation broth firstly, after shaking and extracting, the lactonic SL in organic phase was measured by anthrone method. For total SL determination, 1.0 mL of ethanol was added to 0.5 mL broth to dissolve SLs, then the solution was centrifuged at 8000 rpm for 5 min, and total sugar content in the supernatant was quantified by anthrone method. Total SL production was calculated according to the glucose standard curve with glucose content of total sugar content minus residual glucose content (Ma et al. 2011). Lac SL/tot SL was calculated as a ratio of lactonic SL production on total SL production and stated as a percentage (%). The measurements were carried out in triplicate for each fermentation broth and all results reported are the mean of three independent experimental results. Results In the present study, the composition or structural characteristics of holocellulose have been detected and analyzed firstly. And major fermentation influencing factors of SLs production by W. domercqiae was studied. Several parameters including substrates, inoculum dose, yeast extract concentration and pH regulator was optimized to attempt to reduce the cost and make it more economical. Chemical, elemental and SEM analysis of rice straw and holocellulose The differences in chemical, elemental and SEM analysis of rice straw before/after pretreatment were compared and the results were shown in Tables 1, 2 and Fig. 1.Table 1 Chemical composition analysis of rice straw and holocellulose after pretreatment Water (%) Holocellulose (%) Lignin (%) Ash (%) Cellulose Hemicellulose Rice straw 11.19 ± 0.74 30.46 ± 0.62 25.56 ± 0.83 14.00 ± 0.75 9.76 ± 0.44 Holocellulose 6.60 ± 0.55 59.78 ± 0.77 26.39 ± 0.99 5.77 ± 0.68 1.38 ± 0.57 Table 2 Elemental composition analysis of rice straw and holocellulose after pretreatment C (%) H (%) N (%) O (%) S (%) Rice straw 37.80 5.27 1.18 43.78 0.11 Holocellulose 41.18 6.13 0.83 50.04 0.06 Fig. 1 SEM images comparison of rice straw before and after pretreatment. A Rice straw ×200; B holocellulose ×200; C rice straw ×1000; D holocellulose ×1000 We could see that rice straw contained 56.02 % of holocellulose and 14.00 % of lignin and lignin content reduced to 5.77 % in holocellulose after pretreatment. In other words, the delaminating rate of lignin reached as high as 58.79 % with the pretreatment of SO3 micro-thermal explosion. As shown in Table 1, the change of hemicellulose content of rice straw before and after pretreatment was minor, what is mean xylose content could be equal after hydrolyzing. Besides, the content of water and ash was reduced to 6.60 and 1.38 %, respectively. This phenomenon was because that the small molecule or soluble substances were vaporized and washed. The holocellulose content increased to 86.17 %, which is predominantly attributed to the decrease of lignin, water and ash. This outcome would be conductive to enzymatic hydrolyzing by cellulase to obtain more available sugars. Normally, chemical structure of lignocellulose mainly is carbon skeleton and main elements are hydrogen, oxygen, nitrogen and sulfur (Witczak et al. 2015). The decrease of nitrogen and sulfur will decrease the amount of acid rain formation, which is an environmental hazard. C content increased from 37.80 to 41.18 %, was due to the increase of holocellulose percentage, the same to the O and H content. In addition, SEM was carried out to investigate the capacity of lignin removing by pretreatment of SO3 micro-thermal explosion directly. The microscopic morphology of rice straw before and after pretreatment with different zoom factors (200× and 1000×). Regarding the rice straw (Fig. 1A, C), a compact structure was exposed. As is seen in Fig. 1, the surface of rice straw was covered orderly by different sizes of circular protuberances, which is called lignin, and fibers in rice straw were arranged intensively (Fig. 1A, B). On the contrary, flaking traces and some holes could be observed on the holocellulose surface (Fig. 1B, D). After pretreatment, few lignin protuberances were remained, and compact microstructure was damaged into disorganized (Fig. 1C, D), with the improvement of loose degree of spatial structure. Hence, the efficiency of enzymatic hydrolysis could be increased significantly after the pretreatment. In summary, the pretreatment we adopted will enhance hydrolysis and saccharification process by partly breaking rice straw structure. In the current work, glucose content in hydrolysate was increased from 15 to 75 g/L for rice straw before and after pretreatment. It was found that the saccharification rate was increased to 400 % in holocellulose compared to rice straw. These test results were in agreement with above analysis. Based on the above mentioned results, experiments were carried out to explore the possibility of SLs production from RSHH without detoxification. The pre-experiments results showed that W. domercqiae could utilize RSHH to produce SLs without any pretreatment in a relatively lower yield than the control group. Therefore, further optimization was necessary to obtained higher production and lower cost. Effects of inoculum dose on SL production in RSHH Effects of the inoculum dose on SL production were shown in Table 3, glucose in RSHH could be applied directly for SL production by W. domercqiae.Table 3 Effects of inoculum dose on final broth pH, cell growth and SL production Inoculum dose Final pH Residual glu (g/L) Biomass (g/L) Total SL (g/L) Lactonic SL (g/L) Lac SL/tot SL  (%) Control group 2.55 0.81 ± 0.08 6.53 ± 0.14 40.74 ± 1.60 18.83 ± 0.01 46.22 1 % 5.21 0.63 ± 0.42 5.57 ± 0.09 29.08 ± 3.48 5.28 ± 3.27 18.17 2 % 4.49 0.78 ± 0.21 6.73 ± 0.25 42.33 ± 1.65 8.76 ± 0.46 20.70 4 % 4.26 0.27 ± 0.04 8.53 ± 0.33 37.56 ± 0.27 14.44 ± 0.97 38.46 It was found that, with the increase of inoculum dose, the final broth pH was decreased and the biomass was increased. Sugars (mainly glucose) in RSHH was effectively utilized in the early exponential phase and almost used up at the end of the fermentation. The optimum inoculum dose in RSHH medium dose for total and lactonic SL production was 2 and 4 %, respectively. Lactonic SL production and proportion of lactonic SL to total SL increased with increasing inoculum dose, which demonstrated that high inoculum dose is more suitable for lactonic SL production in RSHH medium. The highest total SL production of 42.33 ± 1.65 g/L, 3.90 % higher than the control group, was obtained with the inoculum dose of 2 %. Therefore, the optimal inoculum dose was set as 2 % in the follow-up experiments. Effects of yeast extract content on SL production in RSHH Considering the effects of varying YE levels, the growth of yeasts, pH value of final broth and production of SL in RSHH fermentation medium containing different yeast extract (YE) contents were shown in Table 4.Table 4 Effects of yeast extract concentration on final broth pH, cell growth and SL production YE content (w/w, %) Final pH Residual glu (g/L) Biomass (g/L) Total SL (g/L) Lactonic SL (g/L) Lac SL/tot SL (%) Control group 2.55 0.81 ± 0.08 6.53 ± 0.14 40.74 ± 1.60 18.83 ± 0.01 46.22 0.00 4.35 1.77 ± 0.01 3.07 ± 0.78 41.53 ± 3.62 10.03 ± 0.84 24.15 0.05 4.03 0.24 ± 0.04 3.55 ± 0.02 49.66 ± 0.99 22.32 ± 2.37 44.94 0.10 4.16 0.14 ± 0.11 4.47 ± 0.28 51.42 ± 0.92 21.60 ± 0.43 42.01 0.15 4.40 0.03 ± 0.00 4.65 ± 0.35 53.70 ± 2.61 22.06 ± 5.33 41.08 0.20 4.46 0.08 ± 0.06 5.64 ± 0.13 50.09 ± 5.92 20.57 ± 1.77 41.06 0.25 4.66 0.18 ± 0.04 6.88 ± 0.46 46.05 ± 4.30 20.12 ± 1.34 43.70 0.30 4.77 0.18 ± 0.00 7.32 ± 0.46 45.83 ± 2.00 18.12 ± 0.48 39.53 As shown in Table 4, glucose in RSHH was almost exhausted in all cases. Yeast growth increased with the increase of yeast extract content, in accordance with the conclusion that sufficient nitrogen source are beneficial for microbe growth. Wickerhamiella domercqiae could survive and utilize sugars in RSHH even no extra nitrogen source adding (YE 0.00 %), which might due to the residual cellulase used in hydrolysis process was applied as nitrogen source. Total SL production increased with the increasing yeast extract concentration under 0.15 % and decreased with further increasing up to 0.30 %. The highest total SL yield of 53.70 ± 2.61 g/L and superior lactonic SL yield of 22.06 ± 5.33 g/L (the highest was 22.32 ± 2.37 g/L) was obtained with the YE addition of 0.15 %, respectively. Furthermore, yield of total SL still reached 41.53 ± 3.62 g/L without YE supplement, 1.94 % higher than that of control group. However, more acidic SL was synthesized and low ratio was obtained in this condition. All of the final broth pH of RSHH cultures was higher than that of control group and final broth pH value increased with the YE adding content. Except for the one without YE supplement (YE 0.00 %), maximum final broth pH, minimum total SL, lactonic SL and Lac SL/tot SL ratio of 4.77, 45.83 ± 2.00, 18.12 ± 0.48 g/L, and 39.53 % were obtained with 0.30 % of YE addition, respectively. Interestingly, the results also revealed that there could be certain relation between final broth pH, SL production and ratio within limits. Effects of pH regulator on SL production in RSHH In order to verify above speculation that final broth pH is negative related to lactonic SL and Lac SL/tot SL ratio, experiments on effects of different pH regulator on SLs production in RSHH were developed. Different pH regulators divided into acidifying agent, alkaline and buffer salt with the same addition was selected and applied to the RSHH medium, respectively. The influence of different pH regulator on SLs production was shown in Table 5.Table 5 Effects of pH regulators on final broth pH, cell growth and SL production pH regulator Final pH Residual glu (g/L) Biomass (g/L) Total SL (g/L) Lactonic SL (g/L) Lac SL/tot SL (%) Control group 2.55 0.81 ± 0.08 6.53 ± 0.14 40.74 ± 1.60 18.83 ± 0.01 46.22 NaOH 4.56 0.81 ± 0.01 9.65 ± 0.01 38.89 ± 1.09 14.91 ± 0.52 38.33 K2CO3 5.62 2.40 ± 0.14 5.72 ± 0.39 38.27 ± 4.03 8.14 ± 3.46 21.27 CaO 5.71 0.29 ± 0.02 9.92 ± 1.01 40.69 ± 0.88 11.96 ± 4.46 29.38 AlCl3 3.89 0.20 ± 0.02 7.05 ± 0.12 50.09 ± 3.64 23.58 ± 4.79 47.08 C6H8O7 4.18 0.17 ± 0.02 9.21 ± 0.08 41.01 ± 0.48 20.03 ± 2.56 48.84 Na3C6H5O7·2H2O 4.48 0.24 ± 0.00 5.87 ± 0.08 39.52 ± 1.96 16.13 ± 0.79 40.81 The results showed that different pH regulator have different effects on final broth pH and SL production. In general, comparable or higher SL production compared to the control group was acquired with different pH regulator. When the content of pH regulator was kept constant, Lewis acid like AlCl3 and mild acid like C6H8O7 were more suitable for lactonic SL production than alkali like NaOH and Lewis base (K2CO3, CaO, Na3C6H5O7·2H2O). Meanwhile, high Lac SL/tot SL ratio of 47.08 and 48.84 % was gained, respectively. It is worth noting that the final broth pH dropped significantly to 3.89 than the others when AlCl3 was added and lactonic and total SL production reached 23.58 ± 4.79 and 50.09 ± 3.64 g/L, with an increase of 25.2 % and 23.0 % compared to the control group, respectively. Besides, lactonic SL production and Lac SL/tot SL ratio of the medium contained K2CO3 and CaO was much lower, in accordance with the trend of final broth pH. The results demonstrated that different pH regulator regulates SL production by changing pH value of the culture again. Correlation coefficient analysis between final broth pH, lactonic SL and Lac SL/tot SL ratio in RSHH All the results above-mentioned suggested that there may be a certain correlation between final broth pH, lactonic SL and Lac SL/tot SL ratio in RSHH. The software of Minitab 17 was used to fit the correlation and the fitting curves and equations were displayed as followed (Fig. 2).Fig. 2 Correlation fitting curve between final broth pH and lactonic SL (a) or Lac SL/tot SL ratio (b) As shown in Fig. 2, the test results of different lactonic SL and ratio of Lac SL/tot SL are in good correlation with the variations of final broth pH values (p = 0.000 < 0.005). Both lactonic SL and Lac SL/tot SL ratio was negatively correlated to the final broth pH, with the linear fitting equation of y = 50.93 − 8.032x and y = 112.5 − 17.14x in the present study, separately. Although the mechanism and explicit relationship remains to be investigated, it did provide a good approach and idea to improve the output of different SLs. On the basis of above experiment results, choosing certain pH regulator to adjust broth pH and achieve large amount of lactonic SL, acidic SL or SLs with appropriate ratio would be possible. However, the complex regulation mechanism of broth pH on SL accumulation at different levels needs further explore and study. Discussion Generally, glucose is common used as hydrophilic carbon source for SL production. Considering glucose is rather rarely found in larger quantities in wastes, which means non-recyclable and a high costs in SL production. In the present comparison, SL yield in non-glucose medium with different hydrophobic carbon sources, nitrogen sources and culture methods by various microbes are compared and exhibited in Table 6.Table 6 Microorganisms, culture conditions, and SL production in non-glucose medium reported in previous references Microorganism Hydrophilic C source (g/L) Hydrophobic C source (g/L) N source (g/L) Culture T (°C)/t (h) Yield (g/L) Country Reference T. bombicola ATCC 22214 (Now known as C. bombicola ATCC 22214) Fructose (20) – YE (10.0) CF 30/72 4.10 Germany Göbbert et al. (1984) Mannose (20) 4.90 Saccharose (20) 3.20 Maltose (20) 2.00 Raffinose (20) 4.10 C. bombicola ATCC 22214 Sucrose (100) Sunflower oil YE (2.5) CF 30/144 33.00 Czechoslovakia Klekner et al. (1991) YE (5.0) 9.00 YE (10.0) 13.00 YE (20.0) 17.00 T. bombicola ATCC 22214 Lactose (100) Olive oil YE (2.5–3.0) b 30/192 46.46 Canada Zhou and Kosaric (1993) Galactose (100) Olive oil 24.41 Sucrose (100) Safflower oil 58.32 C. bombicola ATCC 22214 Lactose (100) Canola oil YE (4.0) CF 30/192 90–110 Canada Zhou and Kosaric (1995) C. bombicola ATCC 22214 Deproteinized whey concentrate [lactose (100)] Rapeseed oil YE (4.0) F-b 26/168 280.00 Germany Daniel et al. (1998a) C. bombicola ATCC 22214 Deproteinized cheese whey concentrate [lactose (110)] SCO & rapeseed oil – F-b 30/410 422.00 Germany Daniel et al. (1998b) C. bombicola ATCC 22214 Soy molasses [333 + 667, total sugar (300)] Oleic acid YE (2.5) F-b 26/168 21.00 USA Solaiman et al. (2004) C. bombicola ATCC 22214 Honey (100) Turkish corn oil YE (10.0) F-b 25/436 >400 Turkey Pekin et al. (2005) C. bombicola ATCC 22214 Biodiesel co-product stream (100 + 100) – YE (10.0) b 26/168 60.00 USA Ashby et al. (2005) C. bombicola ATCC 22214 Glycerol (100) Methyl-soyate YE (10.0) F-b 27/168 46.00 USA Ashby et al. (2006) ethyl-soyate 42.00 propyl-soyate 18.00 C. bombicola ATCC 22214 Soy molasses [333 + 667, total sugar (300)] Oleic acid YE (10.0) F-b 26/168 75.00 USA Solaiman et al. (2007) C. bombicola ATCC 22214 Sugarcane molasses (100) Soybean oil – b 30/120 23.25 India Daverey and Pakshirajan (2009) S. bombicola NRRL Y-17069 Deproteinized whey (90) & glu (10) Oleic acid YE (2.0) b 30/192 23.29 India Daverey and Pakshirajan (2010) CF 25.54 C. bombicola ATCC 22214 Sugarcane molasses (50) Synthetic dairy wastewater & soybean oil – b 30/192 38.76 India Daverey et al. (2011) S. bombicola NBRC 10243 Sugarcane molasses [total sugar (150)] – – F-b 25/120 14.40 Japan Takahashi et al. (2011) S. bombicola ATCC 22214 Sweetwater [glycerol (150)] Sunflower oil YE (4.0) CF 30/200 6.60 India Wadekar et al. (2012) W. domercqiae var. sophorolipid CGMCC 1576 Delignined corncob residue hydrolysate [glu (60)] Single cell oil YE (3.0) CF 30/168 42.06 China Ma et al. (2013) C. bombicola ATCC 22214 Sweet sorghum bagasse hydrolysate [glu (100)] Soybean oil YE (10.0) CF 25/240 84.60 USA Samad et al. (2014) corn fiber hydrolysate [glu (100)] 15.60 S. bombicola ATCC 22214 Glycerol 150 Castor oil – CF 30/200 2.70 India Bhangale et al. (2014) S. bombicola NBRC 10243 Corncob hydrolysate [glu (45)] Olive oil YE (1.0) CF 25/168 43.80 Japan Konishi et al. (2015) W. domercqiae var. sophorolipid CGMCC 1576 Rice straw holocellulose hydrolysate [glu (60)] Oleic acid YE (3.0) CF 30/168 53.70 China Present study T temperature, t time, F-b fed-batch, b batch, CF continuous fermentor, YE yeast extract, glu glucose As indicated in Table 6, yeasts, especially Candida bombicola ATCC 22214, are the most common strains for SLs production in various fermentation cultures. The production is significantly lower when only one carbon source was supplied than two types of carbon source were provided (Göbbert et al. 1984). The highest SL production obtained was 422 g/L by using deproteinized cheese whey concentrate and SCO & rapeseed oil as combined carbon sources, achieved by C. bombicola, with the fermentation time of 410 h and culture mode of fed-batch (Daniel et al. 1998b). Several of cheap substrates can act as hydrophilic carbon sources, such as cheese whey, soy molasses, honey, glycerol, sugarcane molasses, sweet water, DCCR, CCR and RSH, etc. The main aims of all the attempts are to reduce substrate costs and increase SL yields. However, lower yields always were observed when culture mode of batch was used. Fed-batch (F-b) and continuous fermentation (CF) are beneficial for SL accumulation than batch. Yeast extract with different contents are common used nitrogen source in the listed references. Currently, lignocellulosic materials are potential substrates for SL production and relative high yields could be obtained. Further optimization on the pretreatment of lignocellulose, culture modes and reactors would improve the fermentation results even more. Series of reports showed that blockage of the lignin droplets on the surface of the cellulose was the main cause of cellulase inhibition. Besides, the intensive structure of rice straw also hinders the attack of cellulase (Xing et al. 2012; Li et al. 2014). Nevertheless, some small molecule inhibitors would be generated during the pretreatment process and further detoxification treatment would be needed to remove the inhibiting effects on cell growth and product accumulation. However, for pretreatment of SO3 micro-thermal explosion on rice straw, our previous study indicated that the lignin content was dropped dramatically and there are no or little fermentation inhibitors in RSHH (Yao and Li 2013; Wang et al. 2015). And the pretreated rice straw had been directly applied for lipids (Yao et al. 2012), ethanol (Yao and Li 2013) etc. production. These findings are relative significant as it could simplify the lignocellulose utilization process since the detoxification treatment is not required. Various factors were reported to affect sophorolipid accumulation, including C/N ratio, nitrogen source addition, temperature and oxygen supply (Stüwer et al. 1987). Poor research was related to the correlation between final broth pH and sophorolipid production or lac SL/tot SL ratio. In view of this, the effects of inoculum dose, nitrogen source (yeast extract) and pH regulator additions on sophorolipid exting was studied systematically. As the increase of inoculum dose, the biomass was increased and the final broth pH was decreased. This phenomenon was usually ascribed to the increasing organic acid amount generated during more prosperous biological metabolism (Gupta and Prabhune 2012). Interestingly, lactonic SL and lac SL/tot SL ratio were negatively related to the final broth pH. It might be due to the induced effects of broth pH on lactonesterase activity or esterification process occurred in fermentation medium, and the specific mechanism was still remained explored. Nitrogen type and content were reported to have extraordinary influences on SLs production and composition (Cooper and Paddock 1984; Zhou et al. 1992; Casas and García-Ochoa 1999; Ma et al. 2011). Generally, SLs production was initiated at the time of nitrogen source was exhausted and production of lactonic SL strongly depended on nitrogen source (Göbbert et al. 1984). The maximum yield, 53.70 g/L total SL and 22.32 g/L lactonic SL was obtained with the YE addition of 0.15 and 0.05 %, respectively. It indicates that SL production and ratio of lac SL/tot SL can be changed by different YE levels for suitable use. Except for the one without YE supplement (YE 0.00 %), the final broth pH was increased with the YE addition increasing, within the scope of 4.03–4.77. The results demonstrated that YE of appropriate contents, reached to appropriate final broth pH, would be more suitable for total SL or lactonic SL production. Opposite to C. apicola (Stüwer et al. 1987) and W. domercqia (Ma et al. 2011), adjustment of pH by NaOH or Na3C6H5O7·2H2O showed a decrease in lactonic SL production by W. domercqia in RSHH. Furthermore, a strong negative correlation existing between final broth pH and lactonic SL or Lac SL/tot SL ratio was found. It demonstrated the feasibility of SLs obtainment with proper Lac SL/tot SL ratio for specific purpose and scope in a simple, cost efficient and sustainable way. Based on the above, inoculum dose, yeast extract addition and pH regulator can be regarded as factors involved in the complex regulation of sophorolipid accumulation, and the adjustment of final broth pH was the ultimate way to achieve suitable sophorolipid mixture. These factors affected the final broth pH may be closely related to biological metabolism of W. domercqia, biosynthesis of SL production and the enzymes activities in different conditions. Exploring the cheaper substrate to instead of glucose and optimizing the fermentation process was the main aim of this work, of course, which was the same meaning and purpose to the similar articles. For the overview of SL yield in non-glucose medium with different hydrophobic carbon sources, nitrogen source additions and culture methods, a lot of cheap wastes were studied containing sugars, sugar-producing biomass and lignocellulose wastes from crops. Considering the Sustainable development, harmfulness on environment and economic benefits etc. factors, DCCR, CCR and RSH etc. lignocellulosic materials from crops were regarded the most potential substrates. Besides, we believe that the recycling of crop waste was the inevitable trend in the future. Abbreviations SLsophorolipid Lac SLlactonic sophorolipid Acid SLacidic sophorolipid RSHHrice straw holocellulose hydrolysate DCCRdelignified corncob residue hydrolysate CCRcorncob residue hydrolysate SEMscanning electron microscope Gluglucose Xin-ge Liu and Xiao-jing Ma contributed equally to this work Authors’ contributions XJM and RSY conceived and designed the study, XGL performed the experiments. CYP provided the rice straw holocellulose. XGL write the paper, XJM, RSY and HBH reviewed and edited the manuscript. All authors read and approved the final manuscript. Acknowledgements We would like to express our gratitude to Professor Song of Shandong University for providing the strain. Competing interests The authors declare that they have no competing interests. Declarations The authors claim that none of the material in the paper has been published or is under consideration for publication elsewhere. Ethics approval and consent to participate (Not applicable) There was no involvement of human participants or animals in this study. Funding This work was supported by the National High Technology Research and Development Program of China (863 Program, No. 2014AA021902); the National Natural Science Foundation of China (No .31400049); and the China Postdoctoral Science Foundation (No. 2015T80646 and No. 2013M531501). ==== Refs References Ashby RD Nuñez A Solaiman DKY Foglia TA Sophorolipid biosynthesis from a biodiesel co-product stream J Am Oil Chem Soc 2005 82 625 630 10.1007/s11746-005-1120-3 Ashby RD Solaiman DKY Foglia TA The use of fatty acid esters to enhance free acid sophorolipid synthesis Biotechnol Lett 2006 28 253 260 10.1007/s10529-005-5527-y 16555009 Banat IM Makkar RS Cameotra SS Potential commercial applications of microbial surfactants Appl Microbiol Biotechnol 2000 53 495 508 10.1007/s002530051648 10855707 Bhangale A Wadekar S Kale S Bhowmick D Pratap A Production of sophorolipids synthesized on castor oil with glucose and glycerol by using Starmerella bombicola (ATCC 22214) Eur J Lipid Sci Technol 2014 116 336 343 10.1002/ejlt.201300236 Binod P Sindhu R Singhania RR Vikram S Devi L Nagalakshmi S Kurien N Sukumaran RK Pandey A Bioethanol production from rice straw: an overview Bioresour Technol 2010 101 4767 4774 10.1016/j.biortech.2009.10.079 19944601 Bogaert INAV Soetaert W Gloria SC Sophorolipids Biosurfactants 2011 New York Springer Casas JA García-Ochoa F Sophorolipid production by Candida bombicola : medium composition and culture methods J Biosci Bioeng 1999 88 488 494 10.1016/S1389-1723(00)87664-1 16232650 Chandran P Das N Characterization of sophorolipid biosurfactant produced by yeast species grown on diesel oil Int J Sci Nat 2011 2 63 71 Chen J Song X Zhang H Qu YB Miao JY Sophorolipid produced from the new yeast strain Wickerhamiella domercqiae induces apoptosis in H7402 human liver cancer cells Appl Microbiol Biotechnol 2006 72 52 59 10.1007/s00253-005-0243-z 16528516 Concaix FB. Use of sophorolipids comprising diacetyl lactones as agent for stimulating skin fibroblast metabolism. US Patent 6.596.265. 2003. Cooper DG Paddock DA Production of a biosurfactant from Torulopsis bombicola Appl Environ Microbiol 1984 47 173 176 16346455 Cox TF, Crawford RJ, Gregory LG, Hosking SL, Kotsakis P. Mild to the skin, foaming detergent composition. US Patent 8.63.490. 2013. Daniel HJ Otto RT Reuss M Syldatk C Sophorolipid production with high yields on whey concentrate and rapeseed oil without consumption of lactose Biotechnol Lett 1998 20 805 807 10.1023/B:BILE.0000015927.29348.1a Daniel HJ Reuss M Syldatk C Production of sophorolipids in high concentration from deproteinized whey and rapeseed oil in a two stage fed batch process using Candida bombicola ATCC 22214 and Cryptococcus curvatus ATCC 20509 Biotechnol Lett 1998 20 1153 1156 10.1023/A:1005332605003 Daverey A Pakshirajan K Production of sophorolipids by the yeast Candida bombicola using simple and low cost fermentative media Food Res Int 2009 42 499 504 10.1016/j.foodres.2009.01.014 Daverey A Pakshirajan K Sophorolipids from Candida bombicola using mixed hydrophilic substrates: production, purification and characterization Colloids Surf B 2010 79 246 253 10.1016/j.colsurfb.2010.04.002 Daverey A Pakshirajan K Sumalatha S Sophorolipids production by Candida bombicola using dairy industry wastewater Clean Technol Environ Policy 2011 13 481 488 10.1007/s10098-010-0330-4 Göbbert U Lang S Wagner F Sophorose lipid formation by resting cells of Torulopsis bombicola Biotechnol Lett 1984 6 225 230 10.1007/BF00140041 Gorin PAJ Spencer JFT Tulloch AP Hydroxy fatty acid glycosides of sophorose from Torulopsis magnoliae Can J Chem 1961 39 846 855 10.1139/v61-104 Gross RA Madamwars D Nerud F Shah V Sophorolipids as protein inducers and inhibitors in fermentation medium WO Patent 2015 073 371 Gupta R Prabhune AA Structural determination and chemical esterification of the sophorolipids produced by Candida bombicola grown on glucose and alpha-linolenic acid Biotechnol Lett 2012 34 701 707 10.1007/s10529-011-0818-y 22167634 Hillion G, Marchal R, Stoltz C, Borzeix F. Use of a sophorolipid to provide free radical formation inhibiting activity or elastase inhibiting activity. US Patent 5.756.471. 1998. Hu Y Ju LK Purification of lactonic sophorolipids by crystallization J Biotechnol 2001 87 263 272 10.1016/S0168-1656(01)00248-6 11334668 Imura T Masuda Y Minamikawa H Fukuoka T Konishi M Morita T Sakai H Abe M Kitamoto D Enzymatic conversion of diacetylated sophoroselipid into acetylated glucoselipid: surface-active properties of novel bolaform biosurfactants J Am Oil Chem Soc 2010 59 495 501 Klekner V Kosaric N Zhou QH Sophorose lipids produced from sucrose Biotechnol Lett 1991 13 345 348 10.1007/BF01027680 Konishi M Yoshida Y Horiuchi JI Efficient production of sophorolipids by Starmerella bombicola using a corncob hydrolysate medium J Biosci Bioeng 2015 119 317 322 10.1016/j.jbiosc.2014.08.007 25240400 Kurtzman CP Price NPJ Ray KJ Kuo TM Production of sophorolipid biosurfactants by multiple species of the Starmerella (Candida) bombicola yeast clade FEMS Microbiol Lett 2010 311 140 146 10.1111/j.1574-6968.2010.02082.x 20738402 Lee YJ Choi JK Kim EK Youn SH Yang EJ Field experiments on mitigation of harmful algal blooms using a sophorolipid-sellow clay mixture and effects on marine plankton Harmful Algae 2008 7 154 162 10.1016/j.hal.2007.06.004 Lee BG Lee HB Singh BK Noh EK Sang GJ Lee SA Kim J Park HW Guinguino LA Goo YA Kim EK Characterization of Sophorolipid derived from medium chain fatty acids and feasibility study of the Industrial detergent Odontology 2014 66 588 606 Li FH Hu HJ Yao RS Wang H Li MM Structure and saccharification of rice straw pretreated with microwave-assisted dilutelye Ind Eng Chem Res 2012 51 6270 6274 10.1021/ie202547w Li H Pu Y Kumar R Ragauskas AJ Wyman CE Investigation of lignin deposition on cellulose during hydrothermal pretreatment, its effect on cellulose hydrolysis, and underlying mechanisms Biotechnol Bioeng 2014 111 485 492 10.1002/bit.25108 24037461 Liu QZ Guo XP Zhu XQ Progress on biosynthesis and application of sophorolipids Food Drug 2009 11 51 55 Ma XJ Li H Shao LJ Shen J Song X Effects of nitrogen sources on production and composition of sophorolipids by Wickerhamiella domercqiae var. sophorolipid CGMCC 1576 Appl Microbiol Biotechnol 2011 91 1623 1632 10.1007/s00253-011-3327-y 21590287 Ma XJ Li H Wang DX Song X Sophorolipid production from delignined corncob residue by Wickerhamiella domercqiae var. sophorolipid CGMCC 1576 and Cryptococcus curvatus ATCC 96219 Appl Microbiol Biotechnol 2013 98 475 483 10.1007/s00253-013-4856-3 23532513 Maingault M. Utilization of sophorolipids as therapeutically active substances or cosmetic products, in particular for the treatment of the skin. US Patent 5.981.497. 1999. Morya VK Ahn C Jeon S Kim EK Medicinal and cosmetic potentials of sophorolipids Mini Rev Med Chem 2013 13 1761 1768 10.2174/13895575113139990002 23701653 Otto RT Daniel HJ Pekin G Müller-Decker K Fürstenberger G Reuss M Syldatk C Production of sophorolipids from whey: iI. product composition, surface active properties, cytotoxicity and stability against hydrolases by enzymatic treatment Appl Microbiol Biotechnol 1999 52 495 501 10.1007/s002530051551 10570796 Pandey PR Dhasaiyan P Prasad B Roy S Structural insight into self assembly of sophorolipids: a molecular dynamics simulation study Z Phys Chem 2016 230 819 836 10.1515/zpch-2015-0719 Pekin G Vardar-Sukan F Kosaric N Production of sophorolipids from Candida bombicola ATCC 22214 using turkish corn oil and honey Eng Life Sci 2005 5 357 362 10.1002/elsc.200520086 Samad A Zhang J Chen D Liang Y Sophorolipid production from biomass hydrolysates Appl Biochem Biotechnol 2014 175 2246 2257 10.1007/s12010-014-1425-x 25475889 Sirola K Enhanced biodegradation of hydrocarbons in soil by microbial biosurfactant, sophorolipid Appl Biochem Biotech 2010 160 780 790 10.1007/s12010-009-8580-5 Singh PK Prabhune AA Ogale SB Curcumin-sophorolipid complex WO Patent 2016 013 026 Solaiman DKY Ashby RD Nunez A Foglia TA Production of sophorolipids by Candida bombicola grown on soy molasses as substrate Biotechnol Lett 2004 26 1241 1245 10.1023/B:BILE.0000036605.80577.30 15289681 Solaiman DKY Ashby RD Zerkowski JA Foglia TA Simplified soy molasses-based medium for reduced-cost production of sophorolipids by Candida bombicola Biotechnol Lett 2007 29 1341 1347 10.1007/s10529-007-9407-5 17541506 Stüwer O Hommel R Haferburg D Kleber HP Production of crystalline surface-active glycolipids by a strain of Torulopsis apicola J Biotechnol 1987 6 259 269 10.1016/0168-1656(87)90057-5 Takahashi M Morita T Wada K Hirose N Fukuoka T Imura T Kitamoto D Production of sophorolipid glycolipid biosurfactants from sugarcane molasses using Starmerella bombicola NBRC 10243 J Oleo Sci 2011 60 267 273 10.5650/jos.60.267 21502725 Thaniyavarn J Chianguthai T Sangvanich P Roongsawang N Washio K Morikawa M Thaniyavarn S Production of sophorolipid biosurfactant by Pichia anomala Biosci Biotechnol Biochem 2008 72 2061 2068 10.1271/bbb.80166 18685212 Wadekar SD Kale SB Lali AM Bhowmick DN Pratap AA Utilization of sweetwater as a cost-effective carbon source for sophorolipids production by Starmerella bombicola (ATCC 22214) Prep Biochem Biotechnol 2012 42 125 142 10.1080/10826068.2011.577883 22394062 Wang H Pan CY Xu F Liu LJ Yao RS Enhanced saccharification for wheat straw with micro-thermal explosion technology of in situ SO3 reaction Chem Eng J 2015 286 394 399 10.1016/j.cej.2015.11.006 Witczak M Walkowiak M Cichy W Komorowicz M The application of elemental analysis for the determination of the elemental composition of lignocellulosic materials For Wood Technol 2015 205 200 205 Xing Y Bu LX Wang K Jiang JX Pretreatment of furfural residues peroxide to improve cellulose hydrolysis. Characterization of isolated lignin Cell Chem Technol 2012 46 249 260 Yao RS Hu HJ Deng SS Wang H Zhu HX Structure and saccharification of rice straw pretreated with sulfur trioxide micro-thermal explosion collaborative dilutes alkali Bioresour Technol 2011 102 6340 6343 10.1016/j.biortech.2011.02.073 21392977 Yao RS Zhang P Wang H Ss Deng Zhu HX One-step fermentation of pretreated rice straw producing microbial oil by a novel strain of Mortierella elongata PFY Bioresour Technol 2012 124 512 515 10.1016/j.biortech.2012.08.142 23010215 Yao RS, Li FH. Sulfur trioxide micro-thermal Explosion for rice straw pretreatment, In: Theo van de V, John K, editors. Cellullose-biomass conversion; 2013. Zhou QH Kosaric N Effect of lactose and olive oil on intra- and extracellular lipids of Torulopsis bombicola Biotechnol Lett 1993 15 477 482 10.1007/BF00129322 Zhou QH Klekner V Kosaric N Production of sophorose lipids by Torulopsis bombicola from safflower oil and glucose J Am Oil Chem Soc 1992 69 89 91 10.1007/BF02635883 Zhou QH Kosaric N Utilization of canola oil and lactose to produce biosurfactant with Candida bombicola J Am Oil Chem Soc 1995 72 67 71 10.1007/BF02635781

PMC005xxxxxx/PMC5002274.txt

==== Front J Headache PainJ Headache PainThe Journal of Headache and Pain1129-23691129-2377Springer Milan Milan 66710.1186/s10194-016-0667-0Research ArticleCognitive dysfunctions and psychological symptoms in migraine without aura: a cross-sectional study Santangelo Gabriella 13Russo Antonio 23Trojano Luigi luigi.trojano@unina2.it 14Falco Fabrizia 1Marcuccio Laura 23Siciliano Mattia 13Conte Francesca 23Garramone Federica 1Tessitore Alessandro 23Tedeschi Gioacchino gioacchino.tedeschi@unina2.it 231 Department of Psychology, Second University of Naples, Caserta, 81100 Italy 2 Headache Center, Department of Medical, Surgical, Neurological, Metabolic and Aging Sciences, Second University of Naples, Naples, 80138 Italy 3 Institute for Diagnosis and Care “Hermitage Capodimonte”, Naples, 80100 Italy 4 Salvatore Maugeri Foundation, Scientific Institute of Telese, Telese Terme, BN 82037 Italy 27 8 2016 27 8 2016 2016 17 1 765 7 2016 17 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.Background The occurrence of cognitive dysfunctions and psychological symptoms, as well as their mutual relationships, in migraine patients are still debated. The aim of the study was to characterize the cognitive profile and psychological symptoms (i.e. depression, anxiety and apathy) in drug-naïve migraine without aura (MwoA) patients. Methods Seventy-two consecutive MwoA patients, referred to the Italian University Headache Clinic and 72 healthy subjects (HCs) were enrolled. Patients, during an attack-free period, and HCs completed Montreal Cognitive Assessment (MoCA), Beck Depression Inventory-II (BDI-II), Self-version of Apathy Evaluation Scale (AES-S) and State and Trait Anxiety Inventory (STAI-Y-1 and 2). Clinical parameters of disease severity (i.e. disease duration, migraine attacks per month, mean pain intensity during migraine attacks, migraine disability and impact on daily life) were recorded. Results Although performance of MwoA patients on MoCA was above Italian cut-off threshold (<15.5) suggesting presence of cognitive impairment, MwoA patients achieved significantly lower scores than HCs on total MoCA scale (22.3 ± 2.7 versus 25.4 ± 2.3) and on its attention (4.9 ± 1.1 versus 5.6 ± 0.7), memory (1.8 ± 1.4 versus 3.1 ± 1.3), visuospatial (3.2 ± 0.9 versus 3.6 ± 0.6) and executive subscales (2.6 ± 1.1 versus 3.1 ± 0.8). In addition, we observed significant correlations between MoCA executive domain subscore and the attack-related disability score (MIDAS). As for behavioral profile, the percentage of depressive symptoms (4.2 %), high state and trait anxiety (13.9 and 9.7 %, respectively), and apathy (11.1 %) in MwoA patients were similar to that of HCs. No significant associations of behavioural symptoms with cognitive performance and clinical parameters were found. Conclusions Drug-naïve MwoA patients are characterized by subtle cognitive dysfunctions and low percentage of behavioural symptoms. The results support the importance of searching for subclinical cognitive disturbances in patients with MwoA, who deserve to be followed-up to verify whether they develop clinically relevant disorders over time. Keywords MigraineCognitive deficitsDepressionApathyMoCAMontreal cognitive assessmentissue-copyright-statement© The Author(s) 2016 ==== Body Background Migraine is one of the most common pain disorders with a prevalence of 5-20 % in the general population [1], higher in women than in men (with an average prevalence of 20.2 % versus 9.4 %) [2]. Until now, the relationship between migraine and cognitive deficits has been investigated by several cross-sectional studies providing divergent results (for a review see 3). Indeed, some studies did not find any cognitive difference between migraine patients and non-migraine subjects [4–7]. Conversely, other studies revealed that migraine patients are characterized by a poorer cognitive performance [8–15] during both interictal [11–13] or ictal [16, 17] phases when compared to HCs. In particular, it has been suggested that migraine patients might show selective defects in executive/attention and visuospatial domains [9, 11, 13–15]. The discrepancies among studies assessing cognitive functions in migraine may be ascribed to several reasons, possibly related to differences in patients’ characteristics (some studies enrolled both patients with or without aura), sample sizes or neuropsychological assessments [3]. Moreover, some studies did not control the possible influence of clinical variables, such as frequency of attacks, or relationship with pharmacological treatment. In fact, no neuropsychological study specified the role of pharmacological treatment even though some treatments for migraine (e.g., topiramate, amytriptiline) have been associated with cognitive dysfunctions [18, 19]. Last, only a few studies took into account the possible relationships between cognitive performances and associated psychological symptoms or behavioural disturbances, which may often occur in migraine patients [20]. Migraine patients may also present affective temperamental dysregulation, with high hopelessness that may be considered a significant risk factor for negative outcome [21]. Among behavioural disturbances, apathy appears not only as a symptom of depression but as a specific behavioural disturbance [22], with its neural (i.e. abnormalities of prefrontal cortex) and cognitive correlates (i.e. impaired executive/attention and visuospatial abilities) [23, 24]. Since previous neuroimaging studies revealed reduced functional connectivity in the fronto-parietal network in MwoA patients [25], and apathy is associated mainly with abnormalities of prefrontal cortex, it is possible that apathy in itself, not associated with clinically relevant depression, can occur in MwoA. Nonetheless, until now no study specifically assessed apathy, i.e. a loss of interest and motivation [22], in MwoA. On the basis of the above considerations, the present study aimed at investigating the cognitive profile in a homogeneous sample of drug-naïve migraine without aura (MwoA) patients, using the Montreal Cognitive Assessment (MoCA) [26], a widely available screening tool, easy to use in clinical practice. Due to its sensitivity to executive/attention and visuospatial impairments, MoCA seems particularly suitable as a neuropsychological assessment tool in migraine patients, who might show selective defects in such cognitive domains [9, 11, 13–15]. Moreover, we investigated possible relationships between cognitive performances and psychological symptoms such as depression, anxiety and apathy. Finally, we compared neuropsychological and psychological pattern in these patients with those of a group of healthy controls (HCs) matched for age and educational level to patients. Methods Subjects In the present study, consecutive patients with diagnosis of MwoA were recruited from migraine population referring to the outpatient Headache Clinic of the First Division of Neurology of the University of Naples, Italy, from September 2014 to June 2015. The inclusion criterion to participate in the present study was a diagnosis of MwoA according to the ICHD-III beta version criteria of the International Headache Society (IHS). Moreover, we did not enrol in the study patients who showed one of the following exclusion criteria: 1) other ICHD-III diagnosis (e.g., tension type headache, chronic migraine etc.), somatic or psychiatric disorders (e.g., major depression, or psychosis according to DSM-V criteria); 2) current or previous intake of any pharmacological anti-migraine preventive therapy [27] (i.e., we selected patients drug-naïve for anti-migraine preventive treatment); 3) symptoms compatible with acute confusional migraine during migraine attacks. Moreover, to avoid any possible interference related to migraine attacks or to pharmacological treatment on cognitive functions, all MwoA patients were migraine free, and not taking rescue medications, for at least 3 days before and after the neuropsychological assessment. For this aim, patients were interviewed 3 days after neuropsychological assessment to ascertain this point. After enrolment of MwoA patients, for each patient we selected a healthy individual with the same demographic features. MwoA patients and HCs were matched on age and education but not for gender, although the difference in gender distribution in the two samples was not statistically significant (Chi-squared 0.298; p = 0.585). HCs were recruited from among patients’ friends, employees at the clinic or university centres, and were included if they gave their written informed consent to participate on a voluntary basis and met the following selection criteria: lack of history of migraine or any other type of headache and/or current diagnosis of migraine according to clinical criteria; lack of history of or actual psychiatric diseases (e.g., major depression, or psychosis according to DSM-V criteria); no use of psychoactive drugs. All selected participants gave their written informed consent to participate to the study, which was approved by the Local Ethic Committee and was performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. Procedures The demographic and clinical aspects such as disease duration, migraine attacks per month, mean pain intensity during migraine attacks (by means of visual analogic scale -VAS) were recorded. Moreover, to obtain an accurate assessment of patients’ headache-related disability, MwoA patients completed the Migraine Disability Assessment Scale (MIDAS) [28] and the Headache Impact Test −6 (HIT-6) [29]. All patients and HCs completed Beck Depression Inventory-II (BDI-II) [30], Self-version of Apathy Evaluation Scale (AES-S) [22] and State and Trait Anxiety Inventory (STAI-Y-1 and STAI-Y-2, respectively) [31, 32]. The BDI-II is a questionnaire consisting of 21 items, designed to assess severity of depressive symptoms (e.g., sense of failure, guilt, social withdrawal, insomnia, or weight loss). The total score ranges from 0–63, with higher scores reflecting higher levels of depression. The questionnaire allows identifying patients with clinically significant depression according to well-known cut-off in general population (14–19 for mild depression, 20–28 for moderate depression and 29–63 for severe depression) [30]. The AES-S was developed to assess severity of apathy. The questionnaire includes 18 items evaluating four aspects of apathy: “behavioural”, “emotional”, “cognitive” and “other”. Items labeled “other” evaluate reduction or loss of motivation, initiative and accurate understanding of one’s problems. The total score ranges from 18 to 72. Clinically significant apathy was identified according to cut-off score of 38 [22]. The STAI-Y consists of two separate self-report scales for measuring state anxiety, intended as a transitory emotional state, and trait anxiety, consisting in stable tendency to attend to negative emotions such as fears or and anxiety across many situations. The total scores of both subscales (STAI-Y-1 and STAI-Y-2, respectively) range 20–80, and are converted to T-score; a T-score higher than 65 on STAI-Y-1 and STAI-Y-2 indicated high level of state and trait anxiety [31, 32]. All patients and HCs underwent the Italian version of Montreal Cognitive Assessment (MoCA) [26] to evaluate global cognitive status and several cognitive domains: memory, attention, language, and orientation, visuospatial and executive functions domains. The MoCA total score ranges from 0 (worst performance) to 30 (best performance). According to the Italian norms a value of age- and education-adjusted total MoCA score lower than 15.5 is suggestive for presence of cognitive decline; the Italian norms also provide cut-off values for five cognitive domains [33]. Statistical analysis The comparison between MwoA patients and HCs on demographic, neuropsychiatric and cognitive aspects was performed by means of multivariate analysis of variance (MANOVA) and Chi-squared, as appropriate. The performance of MwoA patients and HCs on MoCA was also compared to Italian normative data for MoCA and single cognitive domains to identify how many individuals had clinically relevant cognitive impairment [33]. Several procedures (i.e. Shapiro-Wilk normality test, kurtosis and skewness and a z-score obtained by dividing the skew values or excess kurtosis by their standard errors) to evaluate normal distribution of cognitive and behavioural variables. Within the sample of MwoA patients, the association between clinical, neuropsychiatric and cognitive variables was carried out by means of Spearman’s rank correlation coefficient. Although value of p < 0.05 was considered statistically significant, the Bonferroni correction for multiple comparisons was applied. All analyses were performed using SPSS version 20, (SPSS Inc., Chicago, IL, USA). Results Eighty consecutive patients with diagnosis of MwoA were screened, but only 72 MwoA patients (63 females and 9 males) were enrolled, since 8 patients experienced migraine attacks during the 3 days following the neuropsychological assessment. Moreover, 72 HCs were included in the present study (see Table 1 for a summary of demographic and clinical features).Table 1 Demographic and clinical aspects of migraineurs without aura (MwoA) and healthy subjects MwoA Healthy subjects χ 2/F P Gender (Females/Males) 63/9 66/6 0.298 0.585 Age (years) 34.9 ± 11.2 33.8 ± 11.9 0.421 0.193 Education (years) 12.1 ± 3.6 11.9 ± 3.8 0.112 0.738 Disease duration (years) 15.1 ± 11.6 (13.5; 16.7) - - - Attacks per month 6.1 ± 5.1 (4; 5) - - - MIDAS score 25.3 ± 19.5 (20; 25.75) - - - HIT-6 score 59.9 ± 8.7 (60.5; 10) - - - VAS score 11.1 ± 12.5 (8.2; 10.7) - - - Side of pain  Left 10 -  Right 6 -  Bilateral 56 - The values are expressed in Mean ± Standard Deviation, median and interquartile range are reported in brackets; MIDAS Migraine Disability Assessment Scale, HIT-6 the Headache Impact Test −6, VAS Visual Analogic Scale Neuropsychological assessment MANOVA showed that MwoA patients and HCs had different cognitive profiles (Wilks’ Lambda = 0.694, F = 8.559, df = 7, p < 0.001). MwoA patients performed significantly lower than HCs on the total MoCA score, and on attention, memory, visuospatial and executive domains (all p < 0.007 after Bonferroni correction). No significant difference between the two groups was found on language and orientation domains (Table 2).Table 2 Cognitive comparison between migraineurs without aura (MwoA) and healthy subjects MwoA Healthy subjects F P Mean ± SD Mean ± SD MoCA adjusted total score 22.3 ± 2.7 25.4 ± 2.3 56.606 <0.001 Cognitive Domains  Visuospatial functions 3.2 ± 0.9 3.6 ± 0.6 12.865 <0.001  Executive functions 2.6 ± 1.1 3.1 ± 0.8 11.297 0.001  Attention 4.9 ± 1.1 5.6 ± 0.7 14.776 <0.001  Language 5.2 ± 0.8 5.6 ± 0.6 7.452 0.007  Memory 1.8 ± 1.4 3.1 ± 1.3 29.660 <0.001  Orientation 5.9 ± 0.2 6 ± 0 5.299 0.023 SD Standard Deviation, MoCA Montreal Cognitive Assessment In bold are reported significant differences after Bonferroni correction With reference to available normative data for MOCA, no MwoA individual and no control achieved a total score below normal range. Behavioural assessment MANOVA did not reveal significant difference between MwoA patients and HCs on the behavioural profile (Wilks’ Lambda = 0.940; F = 1.077, df =8; p = 0.383), as assessed by BDI-II, AES-S and its subscales, STAY-1 and STAY-2 (Table 3).Table 3 Behavioural comparisons between migraineurs without aura (MwoA) and healthy subjects MwoA (Mean ± SD) Healthy Subjects (Mean ± SD) F P BDI-II 10.6 ± 9.4 8.9 ± 6.4 1.708 0.193 AES-S: behavioral subscale 7.8 ± 2.5 7.9 ± 2 0.033 0.856 AES-S: cognitive subscale 12.6 ± 3.2 13.5 ± 2.9 2.624 0.107 AES-S: emotive subscale 3.6 ± 1 3.6 ± 1.2 0.022 0.883 AES-S: “others” subscale 6.1 ± 1.6 6.1 ± 1.4 0 1.000 AES-S: total score 29.9 ± 5.5 30.5 ± 5.4 0.351 0.555 STAI-Y-1 40.6 ± 11.6 39.2 ± 12.1 0.453 0.502 STAI-Y-2 42.1 ± 9.7 39.6 ± 9.7 2.324 0.130 Severity of Depressive symptoms (cut-off score of BDI-II)  Severe 3 (4.2 %) 1 (1.7 %)  Moderate 7 (9.7 %) 2 (2.7 %)  Mild 8 (11.1 %) 10 (13.8 %)  No depression 54 (75 %) 59 (81.8 %) Levels of state anxiety (cut-off score of STAY-I-1)  High 10 (13.9 %) 8 (11.1 %) Levels of trait anxiety (cut-off score of STAY-I-2)  High 7 (9.7 %) 3 (4.2 %) Levels of apathy (cut-off score of AES-S)  High 8 (11.1 %) 7 (9.7 %) SD Standard Deviation, BDI-II Beck Depression Inventory-II, AES-S Self-version of Apathy Evaluation Scale, STAI-Y State-Trait Anxiety Inventory- Y With reference to available cut-off scores, severity of depressive symptoms (χ2 = 4.221; df = 2; p = 0.239), of levels of state and trait anxiety (both χ2 < 1) and of apathy (χ2 < 1) did not differ in MwoA and HC groups (Table 3). Correlation analysis between clinical, cognitive and neuropsychiatric variables in MwoA We did not observe significant relationship between total MoCA score or subdomain scores with behavioural scores (AES-S, BDI-II and STAI-Y-1 and STAI-Y-2) (Table 4).Table 4 Correlation between cognitive scores and behavioural scores in patients with migraine without aura BDI-II AES-S -Behaviour AES-S-Cognitive AES-S-Emotive AES-S-Others AES-S-total STAI-Y-1 STAI-Y-2 Cognitive domains r (p value) r (p value) r (p value) r (p value) r (p value) r (p value) r (p value) r (p value) Visuospatial −0.0162 (0.174) −0.062 (0.603) 0.012 (0.923) −0.133 (0.267) 0.017 (0.887) −0.028 (0.815) −0.012 (0.918) −0.071 (0.554) Executive −0.180 (0.130) −0.043 (0.722) −0.042 (0.727) 0.047 (0.697) 0.099 (0.409) 0.005 (0.965) −0.178 (0.134) −0.129 (0.282) Attention 0.016 (0.894) −0.096 (0.421) 0.009 (0.939) 0.134 (0.262) 0.099 (0.408) 0.045 (0.707) 0.048 (0.688) 0.081 (0.499) Language −0.088 (0.462) 0.063 (0.601) −0.220 (0.063) 0.032 (0.789) 0.283 (0.016) −0.040 (0.737) −0.026 (0.826) 0.049 (0.680) Memory −0.162 (0.175) −0.149 (0.213) −0.153 (0.199) 0.104 (0.386) 0.119 (0.320) −0.070 (0.560) −0.123 (0.304) −0.067 (0.574) Orientation 0.166 (0.164) 0.129 (0.280) 0.239 (0.043) 0.265 (0.024) 0.139 (0.245) 0.284 (0.015) 0.024 (0.840) 0.109 (0.361) MoCA total score −0.75 (0.533) −0.033 (0.785) 0.102 (0.393) 0.218 (0.065) 0.181 (0.128) −0.004 (0.972) −0.130 (0.276) −0.072 (0.583) BDI-II Beck Depression Inventory-II, AES-S Apathy Evaluation Scale-Self version, MoCA Montreal Cognitive Assessment, STAI-Y State-Trait Anxiety Inventory- Y, r indicates Pearson's correlation coefficient After Bonferroni correction, MoCA-executive domain score showed a significant, negative and moderate correlation with MIDAS score whereas the remaining cognitive measures did not correlate with clinical and neuropsychiatric variables (Table 5).Table 5 Correlation between behavioral, cognitive and clinical parameters in patients with migraine without aura Disease duration Attacks per month MIDAS HIT-6 VAS Cognitive domains rho (p value) rho (p value) rho (p value) rho (p value) rho (p value) Visuospatial −0.117 (0.342) 0.139 (0.252) 0.067 (0.589) 0.177 (0.149) 0.113 (0.359) Executive 0.177 (0.149) −0.307 (0.010) −0.341 (0.004) −0.092 (0.455) 0.098 (0.427) Attention 0.099 (0.421) −0.011 (0.931) 0.043 (0.729) 0.221 (0.070) 0.041 (0.742) Language 0.061 (0.620) −0.128 (0.290) −0.248 (0.041) −0.079 (0.524) 0.116 (0.348) Memory −0.205 (0.094) −0.008 (0.950) −0.112 (0.361) −0.151 (0.218) 0.196 (0.109) Orientation 0.197 (0.108) 0.111 (0.359) 0.088 (0.477) 0.187 (0.126) 0.055 (0.654) MoCA total score −0.021 (0.865) −0.006 (0.963) −0.093 (0.453) 0.160 (0.192) 0.317 (0.008) BDI-II 0.134 (0.275) 0.031 (0.799) 0.254 (0.037) 0.141 (0.251) −0.065 (0.601) AES-S Behaviour 0.056 (0.648) 0.232 (0.053) 0.061 (0.623) 0.110 (0.372) 0.029 (0.817) AES-S-Cognitive 0.007 (0.957) 0.187 (0.121) 0.210 (0.085) 0.298 (0.014) −0.037 (0.763) AES-S-Emotive 0.059 (0.632) 0.022 (0.855) −0.011 (0.930) −0.057 (0.645) −0.147 (0.231) AES-S-Others −0.030 (0.809) −0.078 (0.521) −0.017 (0.894) 0.010 (0.936) −0.261 (0.031) AES-S-Total 0.059 (0.634) 0.197 (0.102) 0.180 (0.141) 0.207 (0.090) −0.115 (0.352) STAI-Y-1 0.041 (0.742) 0.015 (0.900) 0.076 (0.535) −0.091 (0.462) −0.200 (0.102) STAI-Y-2 0.126 (0.307) 0.070 (0.563) 0.156 (0.205) −0.060 (0.626) −0.009 (0.944) MIDAS Migraine Disability Assessment Scale, HIT-6 the Headache Impact Test −6, VAS Visual Analogic Scale, BDI-II Beck Depression Inventory-II, AES-S Apathy Evaluation Scale-Self version, MoCA Montreal Cognitive Assessment, STAI-Y State-Trait Anxiety Inventory-Y, rho indicates Spearman rho correlation Significant correlation was reported in bold after Bonferroni correction (for cognitive variables: 0.004 (0.05/12); for behavioural variable: 0.003 (0.05/13) Discussion The present study revealed that MwoA patients had significantly lower scores than HCs on the total MoCA scale and in 4 out of 6 cognitive subdomains in all subscores (i.e., executive function, attention, visuospatial and memory domains). However, no MwoA patients achieved scores below the available cut-off values, thus suggesting that the reduced efficiency in selected cognitive domains did not correspond to a clinically relevant cognitive deterioration. As for the psychological profile, MwoA patients and HCs reported similar scores for depression (BDI-II), trait and state anxiety (STAI-Y-1 and STAI-Y-2), and apathy (AES-S). Cognitive scores in MwoA patients were not found to be associated with severity of psychological disturbances, whereas depression, trait and state anxiety and apathy were associated among them. Our first main finding demonstrated the occurrence of cognitive dysfunctions in MwoA patients who had never taken anti-migraine preventive drugs in the course of their life with respect to HCs. In particular, our results revealed lower scores in the subscales assessing verbal memory, attention, frontal and visuospatial functions, in line with several previous studies [8–15]. The low score on the memory scale (i.e., poor retrieval abilities) might be ascribed to defective strategic and organizational aspects of learning [9], also in consideration of the lower attentional and executive function scores in MwoA patients compared to HCs. However, previous studies evaluating migraine patients in the community did not find differences between MwoA patients and HCs [4–6]. The divergence between those studies and ours might be ascribed to the different patient selection procedures: in those studies MwoA patients were selected from a cohort of subjects (i.e., population-based register) on the basis of self-report measures. This procedure might have led to a misclassification of the non-migraine subjects, as suggested by Elkind and Scher [34], and to the enrolment of individuals with very mild migraine, in terms of frequency and intensity of attacks. Instead, we performed a study in a clinical setting where MwoA patients, identified by expert neurologists according to established clinical criteria, required medical intervention due to migraine and related disability. However, our results underlined that MwoA patients’ scores were not lower than the cut-off value reported in normative studies [25, 35]. We can, thus, suggest that MwoA patients with migraine symptoms needing medical consultations show “sub-clinical” neuropsychological impairments mainly affecting executive functions, i.e. a reduced efficiency of cognitive processes in the lack of clinically relevant cognitive deterioration. Such cognitive results would be compatible with recent findings showing reduced functional connectivity in the fronto-parietal network even in migraineurs without overt executive dysfunctions, i.e. with scores on executive tests above the normal cut-off values [25]. It is also worth mentioning that in MwoA patients reduced connectivity in prefrontal and temporal regions of the default mode network [35], in the absence of structural abnormalities, thus suggesting that brain areas involved in executive control may show signs of dysfunction even before development of clinically detectable cognitive impairments. In the present study we found no significant difference on severity of depression, apathy, and state and trait anxiety in MwoA patients compared to HCs. Prevalence of depression and anxiety was lower than that reported in previous studies in which depression, anxiety disorders and migraine were considered to be comorbid diseases [20]. Moreover, where previous study revealed a relationship between depression and migraine [36], we failed to find any association; the divergence among other studies and ours might depend on difference in inclusion and exclusion criterion and on employment of different tools to assess depression. However, our observation is in line with an epidemiological study supporting the idea that psychiatric comorbidity occurs less frequently in MwoA than in migraine with aura [37] or probable medication-overuse headache [38]. Clinically significant apathy occurred in 8 patients only (11 %), and we did not observe significant differences on apathy scale (AES-S) between patients and HCs. Although the finding of mild frontal/executive dysfunction in our patient sample might predict a higher frequency of apathy [23], it must be considered that we enrolled relatively young individuals and that apathy is thought to increase with age [39]. Since this is the first study assessing apathy in MwoA patients, we believe that this issue deserves further cross-sectional and longitudinal investigation in MwA and MwoA patients, who are both characterised by functional abnormalities in prefrontal cortex [25, 35]. Moreover, future studies in large samples of controls and migraineurs should investigate whether psychological disturbance modify the effect of migraine on cognitive impairment. In the present study, several clinical aspects of migraine (pain intensity, disease duration, and frequency of attack) did not influence cognitive performance on MoCA. The lack of significant relationships between frequency of migraine attacks and cognitive performance is in agreement with most previous studies [10, 11], as is the lack of a significant correlation between disease duration and cognitive dysfunctions [9, 11], thus supporting the hypothesis that subtle alterations in information processing mechanisms might be present also in the early stages of migraine [11]. Finally, among clinical aspects of migraine, we found that high levels of attack related disability were associated with reduced executive functioning as assessed by the MoCA executive subscore. This result may lend support to the idea that some aspects of cognitive dysfunction are relevant contributors to migraine attack-related disability [16]. Taken together, our findings suggest that MwoA is associated with cognitive dysfunctions and in particular, that altered executive functioning can be related to high migraine related disability. Thus, these results underscore the importance of a careful evaluation of cognitive function even with a screening tool, provided that sensitive to detect dysfunctions of attention, memory, visuospatial and executive domains, as the MoCA is. Moreover, no correlation between scales assessing neuropsychiatric symptoms and cognitive tests suggests that the poorer cognitive performance exhibited by MwoA patients was not conditioned by the emotional variables such as depression, apathy or anxiety, differently from what reported in other neurological diseases [23, 40]. Our study had several strengths and limitations. The inclusion of a sample of patients homogeneous for type of migraine (all MwoA patients) and drug-naïve for preventive pharmacological therapies can be considered as strength of the study, suggesting that cognitive dysfunctions can occur in the natural history of MwoA, independently from preventive migraine therapies. As a consequence, these data could be considered important to better understand the MwoA clinical spectrum. However, the cross-sectional nature of our study did not allow us to ascertain whether the mild cognitive impairments we could detect here are amenable to changes related to pharmacological treatment, and can herald clinically relevant cognitive impairments. Our patient group was a convenience sample of treatment seeking migraine suffers and thus, may represent a limitation about generalizability of the findings. Furthermore, our sample was composed mainly of women as migraine is strongly related to gender, but this might also limit the generalizability of the results. Furthermore, we could not evaluate whether pain prevalence on either side of the head may be associated with different cognitive profiles, because most of our patients had no preferred side for pain. This issue deserved to be explored in further studies including balanced groups of patients with unilateral side of pain. Finally, in the present paper we only aimed at obtaining a screening of general cognitive functioning, and for this purpose we adopted a relatively recent tool, i.e. MoCA, which has been developed to obtain an estimation of several cognitive domains, including those (such as executive functions) that are often overlooked in more common screening tools. The MoCA cannot be used as a diagnostic instrument, but to screen cognitive status of patients attending a migraine ambulatory and to provide a baseline assessment, which could be compared to future assessments. The cut-off score used in our study refers to age- and adjusted-scores and not to raw scores, and it is specifically foreseen by the available Italian normative study. The cut-off value is clearly lower than that reported in the original study based on a sample of 90 healthy Canadian controls (mean age 72.84 ± 7.03 years; mean education 13.33 ± 3.40), using a one-point correction for education (12 years). However, several factors likely contribute to this discrepancy such as: i) reference to age- and education-adjusted scores and not to raw scores; ii) use of population-based normative sample instead of convenience sample, as in the original study (note that even in English speaking countries population-based normative studies provided lower cut-off values than that proposed in the original study [41]; iii) cultural and linguistic biases related to translations of the test in different languages (as in Portuguese [42] and Japan language [43]. For this reasons it has been clearly stated that appropriate MoCA normative data have to be employed when interpreting MoCA scores [41]. It is also interesting to note that the cut-off value of 15.5 for age- and education-adjusted scores did not identify any control or migraineurs participant as affected by cognitive impairment, whereas applying the cut-off score of 26 cognitive impairment would be present in 68/72 (94.4 %) MwoA patients and in 44/72 (61.1 %) controls. Such percentages appear to be implausibly high in two samples of home-dwelling and active young adults, thus further reinforcing the need for cut-off points based upon country-specific normative data. Conclusions In conclusion, we believe that early identification of cognitive deficits in MwoA patients is relevant for future care planning. The MoCA seems suitable to screen such cognitive defects in clinical practice in MwoA patients. Abbreviations AES-SSelf-version of apathy evaluation scale BDI-IIBeck depression inventory-II HCshealthy controls HIT-6Headache impact test −6 MIDASMigraine disability assessment scale MoCAMontreal cognitive assessment MwAMigraine with aura MwoAMigraine without aura SPSSStatistical package for the social sciences STAI-Y-1State anxiety inventory STAI-Y-2Trait anxiety inventory VASVisual analogic scale Authors’ contributions GS and AR were involved in conception and design of the study; FF, LM, MS, FC, FG contributed to data acquisition. Analysis and data interpretation were performed by GS, AR and LT. The article was drafted by GS (for neuropsychological issue) and AR (for neurological issue) with input from LT and AT. All authors reviewed and approved the final manuscript. Competing interests The authors declare that they have no competing interests. ==== Refs References 1. Lipton RB Bigal ME Migraine: epidemiology, impact, and risk factors for progression Headache 2005 45 Suppl 1 S3 S13 10.1111/j.1526-4610.2005.4501001.x 15833088 2. Burch RC Loder S Loder E Smitherman TA The prevalence and burden of migraine and severe headache in the United States: updated statistics from government health surveillance studies Headache 2015 55 1 21 34 10.1111/head.12482 25600719 3. Rist PM Kurth T Migraine and cognitive decline: a topical review Headache 2013 53 4 589 598 10.1111/head.12046 23405909 4. Gaist D Pedersen L Madsen C Tsiropoulos I Bak S Sindrup S McGue M Rasmussen BK Christensen K Long-term effects of migraine on cognitive function: a population-based study of Danish twins Neurology 2005 64 600 607 10.1212/01.WNL.0000151858.15482.66 15728279 5. Jelicic M van Boxtel MP Houx PJ Jolles J Does migraine headache affect cognitive function in the elderly? Report from the Maastricht Aging Study (MAAS) Headache 2000 40 715 719 10.1046/j.1526-4610.2000.00124.x 11091288 6. Pearson AJ Chronicle EP Maylor EA Bruce LA Cognitive function is not impaired in people with along history of migraine: a blinded study Cephalalgia 2006 26 74 80 10.1111/j.1468-2982.2005.01001.x 16396669 7. Burker E Hannay HJ Halsey JH Neuropsychological functioning and personality characteristics of migrainous and nonmigrainous female college students Neuropsychology 1989 3 61 73 8. Hooker WD Raskin NH Neuropsychologic alterations in classic and common migraine Arch Neurol 1986 43 709 712 10.1001/archneur.1986.00520070065020 3729750 9. Le Pira F Zappalà G Giuffrida S Lo Bartolo ML Reggio E Morana R Lanaia F Memory disturbances in migraine with and without aura: a strategy problem? Cephalalgia 2000 20 475 478 10.1046/j.1468-2982.2000.00074.x 11037744 10. Calandre EP Bembibre J Arnedo ML Becerra D Cognitive disturbances and regional cerebral blood flow abnormalities in migraine patients: their relationship with the clinical manifestations of the illness Cephalalgia 2002 22 291 302 10.1046/j.1468-2982.2002.00370.x 12100092 11. Le Pira F Lanaia F Zappalà G Morana R Panetta MR Reggio E Reggio A Relationship between clinical variables and cognitive performances in migraineurs with and without aura Funct Neurol 2004 19 101 105 15274516 12. Mulder EJ Linssen WH Passchier J Orlebeke JF de Geus EJ Interictal and postictal cognitive changes in migraine Cephalalgia 1999 19 557 565 10.1046/j.1468-2982.1999.019006557.x 10448542 13. Camarda C, Monastero R, Pipia C, Recca D, Camarda R (2007) Interictal executive dysfunction in migraineurs without aura: relationship with duration and intensity of attacks. Cephalalgia 27:1094–1100 14. Schmitz N Arkink EB Mulder M Rubia K Admiraal-Behloul F Schoonman GG Kruit MC Ferrari MD van Buchem MA Frontal lobe structure and executive function in migraine patients Neurosci Lett 2008 440 2 92 96 10.1016/j.neulet.2008.05.033 18556120 15. Martins IP Gil-Gouveia R Silva C Maruta C Oliveira AG Migraine, headaches, and cognition Headache 2012 52 10 1471 1482 10.1111/j.1526-4610.2012.02218.x 22830358 16. Gil-Gouveia R Oliveira AG Martins IP The impact of cognitive symptoms on migraine attack-related disability Cephalalgia 2016 36 5 422 30 10.1177/0333102415604471 26350071 17. Gil-Gouveia R Oliveira AG Martins IP Cognitive dysfunction during migraine attacks: a study on migraine without aura Cephalalgia 2015 35 8 662 674 10.1177/0333102414553823 25324500 18. Sommer BR Mitchell EL Wroolie TE Topiramate: effects on cognition in patients with epilepsy, migraine headache and obesity Ther Adv Neurol Disord 2013 6 4 211 227 10.1177/1756285613481257 23858325 19. van Laar MW Volkerts ER Verbaten MN Trooster S van Megen HJ Kenemans JL Differential effects of amitriptyline, nefazodone and paroxetine on performance and brain indices of visual selective attention and working memory Psychopharmacology (Berl) 2002 162 4 351 363 10.1007/s00213-002-1116-0 12172688 20. Buse DC Silberstein SD Manack AN Papapetropoulos S Lipton RB Psychiatric comorbidities of episodic and chronic migraine J Neurol 2013 260 8 1960 1969 10.1007/s00415-012-6725-x 23132299 21. Serafini G Pompili M Innamorati M Gentile G Borro M Lamis DA Lala N Negro A Simmaco M Girardi P Martelletti P Gene variants with suicidal risk in a sample of subjects with chronic migraine and affective temperamental dysregulation Eur Rev Med Pharmacol Sci 2012 16 10 1389 1398 23104655 22. Marin RS Biedrzycki RC Firinciogullari S Reliability and validity of the apathy evaluation scale Psychiatry Res 1991 38 2 143 162 10.1016/0165-1781(91)90040-V 1754629 23. Santangelo G Vitale C Trojano L Picillo M Moccia M Pisano G Pezzella D Cuoco S Erro R Longo K Pellecchia MT Amboni M De Rosa A De Michele G Barone P Relationship between apathy and cognitive dysfunctions in de novo untreated Parkinson's disease: a prospective longitudinal study Eur J Neurol 2015 22 2 253 260 10.1111/ene.12467 24848193 24. Tessitore A Russo A Esposito F Giordano A Taglialatela G De Micco R Cirillo M Conte F d’Onofrio F Cirillo S Tedeschi G Interictal cortical reorganization in episodic migraine without aura: an event-related fMRI study during parametric trigeminal nociceptive stimulation Neurol Sci 2011 32 Suppl 1 S165 7 10.1007/s10072-011-0537-0 21533737 25. Russo A Tessitore A Giordano A Corbo D Marcuccio L De Stefano M Salemi F Conforti R Esposito F Tedeschi G Executive resting-state network connectivity in migraine without aura Cephalalgia 2012 32 14 1041 1048 10.1177/0333102412457089 22908362 26. Nasreddine ZS Phillips NA Bédirian V Charbonneau S Whitehead V Collin I Cummings JL Chertkow H The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment J Am Geriatr Soc 2005 53 695 699 10.1111/j.1532-5415.2005.53221.x 15817019 27. Evers S Afra J Frese A Goadsby PJ Linde M May A Sándor PS European Federation of Neurological Societies EFNS guideline on the drug treatment of migraine-revised report of an EFNS task force Eur J Neurol 2009 16 9 968 981 10.1111/j.1468-1331.2009.02748.x 19708964 28. Stewart WF Lipton RB Dowson AJ Sawyer J Development and testing of the Migraine Disability Assessment (MIDAS) questionnaire to assess headache-related disability Neurology 2001 56 6 Suppl 1 S20 28 10.1212/WNL.56.suppl_1.S20 11294956 29. Kosinski M Bayliss MS Bjorner JB Ware JE Jr Garber WH Batenhorst A Cady R Dahlof CG Dowson A Tepper S A six-item short-form survey for measuring headache impact: the HIT-6 Qual Life Res 2003 12 963 974 10.1023/A:1026119331193 14651415 30. Beck AT Steer RA Ball R Ranieri W Comparison of beck depression inventories -IA and -II in psychiatric outpatients J Pers Assess 1996 67 3 588 597 10.1207/s15327752jpa6703_13 8991972 31. Spielberger CD Gorsuch R Lushene R Vagg PR Jacobs GA Manual for the State-Trait Anxiety Inventory (Form Y) 1983 Palo Alto Consulting Psychologists Press 32. Pedrabissi L Charles SM Spielberger D STAI, State-Trait Anxiety Inventory-Forma Y 1989 Firenze Giunti Organizzazioni Speciali 33. Santangelo G Siciliano M Pedone R Vitale C Falco F Bisogno R Siano P Barone P Grossi D Santangelo F Trojano L Normative data for the Montreal Cognitive Assessment in an Italian population sample Neurol Sci 2015 36 4 585 591 10.1007/s10072-014-1995-y 25380622 34 Elkind MS Scher AI Migraine and cognitive function: some reassuring news Neurology 2005 64 4 590 591 10.1212/01.WNL.0000154671.56854.90 15728275 35 Tessitore A Russo A Conte F Giordano A De Stefano M Lavorgna L Corbo D Caiazzo G Esposito F Tedeschi G Abnormal connectivity within executive resting-state network in migraine with aura Headache 2015 55 6 794 805 10.1111/head.12587 26084236 36 Rist PM Schürks M Buring JE Kurth T Migraine, headache, and the risk of depression: prospective cohort study Cephalalgia 2013 33 12 1017 1025 10.1177/0333102413483930 23588795 37 Breslau N Schultz LR Stewart WF Lipton RB Lucia VC Welch KM Headache and major depression: is the association specific to migraine? Neurology 2000 54 2 308 313 10.1212/WNL.54.2.308 10668688 38 Lampl C Thomas H Tassorelli C Katsarava Z Laínez JM Lantéri-Minet M Rastenyte D Ruiz de la Torre E Stovner LJ Andrée C Steiner TJ Headache, depression and anxiety: associations in the Eurolight project J Headache Pain 2016 17 59 10.1186/s10194-016-0649-2 27245683 39 Brodaty H Altendorf A Withall A Sachdev P Do people become more apathetic as they grow older? A longitudinal study in healthy individuals Int Psychogeriatr 2010 22 3 426 436 10.1017/S1041610209991335 20003630 40 Goretti B Viterbo RG Portaccio E Niccolai C Hakiki B Piscolla E Iaffaldano P Trojano M Amato MP Anxiety state affects information processing speed in patients with multiple sclerosis Neurol Sci 2014 35 4 559 563 10.1007/s10072-013-1544-0 24072658 41 Rossetti HC Lacritz LH Cullum CM Weiner MF Normative data for the Montreal Cognitive Assessment (MoCA) in a population-based sample Neurology 2011 77 13 1272 1275 10.1212/WNL.0b013e318230208a 21917776 42 Freitas S Simões MR Alves L Santana I Montreal Cognitive Assessment (MoCA): normative study for the Portuguese population J Clin Exp Neuropsychol 2011 33 989 996 10.1080/13803395.2011.589374 22082082 43 Narazaki K Nofuji Y Honda T Matsuo E Yonemoto K Kumagai S Normative data for the Montreal Cognitive Assessment in a Japanese community-dwelling older population Neuroepidemiology 2013 40 23 29 10.1159/000339753 23075757

PMC005xxxxxx/PMC5002275.txt

==== Front Rice (N Y)Rice (N Y)Rice1939-84251939-8433Springer US New York 11510.1186/s12284-016-0115-4Original ArticleParental Genome Imbalance Causes Post-Zygotic Seed Lethality and Deregulates Imprinting in Rice Zhang Hong-yu 1Luo Ming +61262464801ming.luo@csiro.au 2Johnson Susan D. 3Zhu Xiao-wei 14Liu Lei 1Huang Fang 1Liu Yu-tong 1Xu Pei-zhou 1Wu Xian-jun +8613708203775wuxjsau@126.com 11 Rice Research Institute, Sichuan Agricultural University, Wenjiang Chengdu, People’s Republic of China 2 Commonwealth Scientific and Industrial Research Organization (CSIRO), Agriculture Flagship, PO Box 1600, Canberra, 2601 ACT Australia 3 Commonwealth Scientific and Industrial Research Organization (CSIRO), Agriculture Flagship, Waite Campus, PMB 2, Glen Osmond, SA 5064 Australia 4 Institute of genetics and developmental biology, Chinese Academy of Sciences, 100101 Beijing, China 27 8 2016 27 8 2016 2016 9 1 434 5 2016 12 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.Background Reproductive isolation between rice of different ploidy levels is manifested as endosperm and embryo abortion in seeds produced by interploidy crosses. Genomic imprinting is considered to be the underlying mechanism establishing the post-zygotic hybridization barrier. We characterized disrupted seed development in reciprocal crosses between a diploid Japonica rice and a tetraploid Indica rice. Results Triploid seeds from these crosses had aborted development and could not germinate in soil but could be rescued in culture medium with significantly more seeds developing to seedlings in the 4n × 2n (♀-♂) cross with excess maternal genomes than in the 2n × 4n cross with excess paternal genome. Consistent with previous findings, precocious endosperm cellularization and bigger embryos were observed in the seeds from the maternal excess cross, whereas absence of cellularization and arrested globular embryos were found in the seeds from the paternal excess cross, supporting the idea that endosperm cellularization is an important transition for embryo development. Moreover, we found that starch granules were persistently deposited in the pericarp parenchyma cells of the paternal excess cross, while pericarp starch gradually decreased and relocated to the developing endosperm in balanced and maternal excess crosses in which cellularization and starch deposition occur in endosperm, suggesting that parental genome balance influences pericarp starch relocation via cellularization and starch deposition. Loss of imprinting, or altered expression of imprinted genes and epigenetic regulators, OsFIE2 and OsMET1b were observed, implying the potential role of imprinting and epigenetic mechanisms in regulating the differential parental genome dosage effects on endosperm development. Conclusions Our results support the hypothesis that the maternal genome dosage promotes endosperm cellularization and the paternal genome dosage delays or inhibits cellularization via contributing different sets of imprinted genes. Electronic supplementary material The online version of this article (doi:10.1186/s12284-016-0115-4) contains supplementary material, which is available to authorized users. Keywords PolyploidEndospermEmbryoCellularizationReproductionPolycombDNA methylationImprintingTriploid blockissue-copyright-statement© The Author(s) 2016 ==== Body Background Crosses between two parents with different ploidy levels result in a change of parental genome ratios by adding either additional maternal or paternal copies to the embryo and endosperm. Additionally, absolute ploidy levels and the relative genome dosages of the two fertilization products to the maternal tissues are also altered (Table 1). The consequences of interploidy hybridization on seed development have been studied in a wide range of species (Watkins 1932; Howard 1939; Brink and Cooper 1947; Hakansson 1956; Nishiyama and Inomata 1966; Lin 1984; Ramsey and Schemske 1998; Scott et al. 1998; Gutierrez-Marcos et al. 2003; Stoute et al. 2012; Kradolfer et al. 2013a; Sekine et al. 2013). Endosperm development is more affected than embryo and shows contrasting phenotypes between reciprocal interploidy crosses: with precocious cellularization and poor proliferation occurring in the maternal excess cross, and delayed or interrupted cellularization and over proliferation occurring in the paternal excess cross in each species. This suggests that similar mechanisms operate in different species to regulate post-zygotic incompatibility between plants of different ploidy levels. The analysis of the crosses between maize 2n or 4n pollen donors and the 2n seed parent indeterminate gametophyte which gives varying numbers of polar nuclei to the endosperm suggested that a 2maternal (m):1paternal (p) genomic ratio rather than a specific ploidy level of endosperm is critical for normal seed development (Lin 1984).Table 1 Ploidy levels and parental genome ratios of rice seed components, and seed setting and embryo rescue rates from balanced and unbalanced crosses Crosses Maternal ploidy Embryo ploidy (m:p) Endosperm ploidy (m:p) Seed setting rate Embryo rescue rate Nip2n × Nip2n 2n 2n (1:1) 3n (2:1) 27.42 % (164/598) 30.43 % (42/138) TH4n × TH4n 4n 4n (2:2) 6n (4:2) 14.04 % (183/1303) 37.35 % (31/83) TH4n × Nip2n 4n 3n (2:1) 5n (4:1) 17.26 % (231/1338) 48.86 % (86/176) Nip2n × TH4n 2n 3n (1:2) 4n (2:2) 10.05 % (255/2537) 4.69 % (16/341) The 2 m:1p requirement was believed to be an indication of existence of genomic imprinting in endosperm (Haig and Westoby 1989; 1991). Genomic imprinting (or imprinting) refers to the selective expression of a parental copy of a gene depending on its parent-of-origin due to the differential epigenetic modifications of parental genomes. It has been hypothesized that maternally expressed genes (MEGs) and paternally expressed imprinted genes (PEGs) have differential functions in regulate endosperm development (Haig and Westoby 1991; Scott et al. 1998). In plants, imprinted genes are predominantly found in endosperm but not in embryo (Luo et al. 2011; Waters et al. 2011; Wolff et al. 2011). Over dosage of MEGs in maternal excess cross and PEGs in the paternal excess cross are thought to contribute to the contrasting endosperm phenotypes between reciprocal interploidy crosses (Haig and Westoby 1991; Scott et al. 1998). Thus imprinting provides a likely explanation for the parent-of-origin specific effects on endosperm development in interploidy crosses. Several hypotheses have been proposed to explain the evolution and function of imprinting and its role in interploidy cross (Haig and Westoby 1989; Birchler 1993; Dilkes and Comai 2004; Josefsson et al. 2006; Jullien and Berger 2010; Berger et al. 2012). Dicot Arabidopsis has been used as model to analyse the parental genome imbalance on seed development and the underlying molecular mechanisms. In reciprocal crosses using 2n and 4n in certain ecotypes or using 2n and 6n, embryos arrest with an undisturbed basic body plan but accompanied with precocious endosperm cellularization or discrupted cellularization based on the directions of the crosses (Scott et al. 1998). The FERTILIZATION INDEPENDENT SEED (FIS) Polycomb Repressive Complex (PRC2) complex which catalyzes histone H3 lysine 27 trimethylation (H3K27me) and maintenance DNA methyltransferase 1 (MET1) play essential and antagonistic roles in regulating endosperm cellularization via controlling expression of key imprinted genes in Arabidopsis (Luo et al. 2000; Adams et al. 2000; Erilova et al. 2009; Kradolfer et al. 2013a; 2013b; Schatlowski et al. 2014). Seed defects or altered cellularization in the maternal excess cross or in the paternal excess cross are partially suppressed by the loss of function of a maternally expressed imprinted gene (MEG) PRC2 MEDEA (MEA), or by a paternally derived DNA hypomethylated genome, respectively (Erilova et al. 2009; Kradolfer et al. 2013a; Schatlowski et al. 2014), suggesting a connection between interploidy seed defects and deregulated function of PRC2 and MET1. Loss of imprinting and/or altered expression at MEA, FIS2 and PHERES 1 loci have been observed in interploidy crosses (Erilova et al. 2009; Jullien and Berger 2010). The connection between loss of imprinting and endosperm abortion is yet to be investigated. In the monocot rice, differential parental genome dosage effects on seed development in the reciprocal crosses between a 2n Japonica line and its own 4n line have recently been observed (Sekine et al. 2013). Endosperm development is inhibited with precocious cellularization but accompanied by swollen embryos in the 4n × 2n maternal excess cross, while seeds in the reciprocal paternal excess crosses display disrupted cellularization with arrested globular embryos. Starch granules are not seen in the disrupted endosperm of the 2n × 4n cross, suggesting cellularization is a prerequisite for the starch accumulation in endosperm. It has been implied that transitory pericarp starch in parenchyma cells plays an important role in seed development, as this type of starch is broken down and relocated to developing seed (Duffus and Rosie 1973; Sato 1984; Caley et al. 1990). It remains to be determined how relocation of the pericarp starch responds to the altered cellularization and starch deposition in endosperm with parental genome imbalance. The rice genome contains two FERTILISATION INDEPENDENT ENDOSPERM (FIE) genes encoding members of the PRC2 complex (Luo et al. 2009; Nallamilli et al. 2013; Li et al. 2014) and two MET1 genes (Yamauchi et al. 2009; Hu et al. 2014). OsFIE2 and OsMET1b are the major genes potentially regulating endosperm development and other processes (Luo et al. 2009; Yamauchi et al. 2009; Nallamilli et al. 2013; Hu et al. 2014; Li et al. 2014). Given that PRC2 and MET1 are required for establishing postzygotic hybridization barrier via controlling imprinted genes in Arabidopsis (Erilova et al. 2009; Kradolfer et al. 2013a; Schatlowski et al. 2014) and differential expression of imprinted genes between parental genomes in rice are likely controlled by differential H3K27me and DNA methylation (Rodrigues et al. 2013; Du et al. 2014), it is interesting to know if the rice PRC2 and MET1 counterparts have the same functions in regulating postzygotic hybridization barrier as in Arabidopsis, and whether loss of imprinting also occurs in the rice endosperm with parental genome imbalance. Subspecies hybridization between Indica and Japonica rice at 2n gives viable seeds with no apparent endosperm defect (Luo et al. 2011). In this study, we carried out a histological study on the seeds derived from reciprocal crosses between a 2n Japonica rice and an Indica 4n rice. We show similar patterns of disruption of endosperm and embryo development to the seeds from the interploidy reciprocal crosses in the Japonica background (Sekine et al. 2013). We observed that the deposition of starch granules in the pericarp parenchyma cells were different in the paternal excess cross compared with those in other crosses. With the sequence polymorphisms between the two parents, we were able to analyze the status of imprinted gene expression and observed that several imprinted genes lost imprinting or expression, along with the altered expression of two potential imprinting regulators, the PRC2 gene OsFIE2 and the MET1 ortholog OsMET1b in the parental genome unbalanced endosperm. This study provides useful clues for further understanding the mechanisms underlying reproductive isolation between rice of different ploidy levels. Results Seed Development is Perturbed in the Reciprocal Interploidy Crosses Using a diploid Japonica cultivar Nipponbare (Nip2n) and a tetraploid Indica line Tetra-Haitian (TH4n), we performed both self- and cross-pollinations to examine the effects of parental genome imbalance on seed development. We used the root tips to count the chromosomes to confirm ploidy levels for both parents and the progeny (Additional file 1: Figure S1). Panicles from the crosses, TH4n × TH4n, Nip2n × Nip2n, Nip2n × TH4n (paternal excess), and TH4n × Nip2n (maternal excess) were harvested for analysis of seed set, viability and development. There was a difference in the rates of seed setting among the four sets of crosses after cross-pollination, with crosses involving a 4n parent showing lower seed set (Table 1). All the seeds were aborted in both unbalanced crosses at maturity (Additional file 1: Figure S1). The outcomes of TH4n reciprocally crossed to H2n, the 2n progenitor line of TH2n were similar to the interploidy crosses between subspecies (Additional file 1: Figure S1E) and the Nip2n crossed with H2n Indica gave normal seeds, suggesting that seed abortion was not due to the difference of genetic background between Japonica and Indica but specific to the unbalanced crosses. Seeds derived from the rice unbalanced crosses could not be germinated on soil but the developing embryos were able to be rescued on nutrient-enriched medium. We rescued developing embryos at days 5, 8, 11, and 14 post-pollination on MS media, with an average of 48.86 % embryos in TH4n × Nip2n and 4.69 % embryos in Nip2n × TH4n being rescued (Table 1). The resulting seedlings reached maturity without any obvious developmental abnormality, except for the sterility expected from a 3n rice plant. Root tip chromosome counting confirmed the seedlings from the unbalanced crosses were triploid (Additional file 1: Figure S1C and D). Interestingly, embryos from the maternal excess TH4n × Nip2n cross had a much higher germination rate than those from the paternal excess Nip2n × TH4n cross, suggesting that postzygotic hybridization barriers between the two unbalanced reciprocal crosses are asymmetric and embryos are less affected by adding extra maternal genome copies. Distinct Endosperm and Embryo Development Patterns Result from the Maternal and Paternal Excess Crosses To understand the effects of parental genome imbalance on seed development, developing seeds from the balanced and unbalanced crosses were sectioned at 2, 3, 4 and 5 DAP (days after pollination). The development of embryo and endosperm in the Nip2n × Nip2n cross followed the same pattern as reported previously (Fig. 1a–d; Brown et al. 1996; Itoh et al. 2005). At 2 DAP, the embryo was at the globular stage (Fig. 1a). The endosperm was a syncytium, containing nuclei lining the entire peripheral cytoplasm, with a large central cavity. At 3 DAP, the embryo was slightly elongated and endosperm cellularization was well under way. The division of the endosperm cells was rapidly progressing toward the centre of the embryo sac, leaving a smaller cavity (Fig. 1b). By 4 DAP, the embryo differentiated a coleoptile primordium and accumulation of starch granules could be seen in the endosperm cells (Fig. 1c). At 5 DAP, the embryo differentiated further and a significant number of starch granules had accumulated in the endosperm (Fig. 1d). Seed development observed in another balanced cross (Fig. 1m–p), TH4n × TH4n, was similar to that in the Nip2n × Nip2n cross. By 2 DAP, the embryo was globular and the endosperm was uncellularized (Fig. 1m). Similarly, at 3 DAP, the peripheral endosperm had cellularized (Fig. 1n). The development of embryo and endosperm at later stages (TH4n × TH4n) followed the same patterns as seen in the Nip2n × Nip2n seeds (Fig. 1o and p).Fig. 1 Sections of developing rice seeds in parental genome balanced and unbalanced crosses showing endosperm and embryo development. a-d Sections of Nip2n × Nip2n developing seeds at 2 DAP a, 3 DAP b, 4 DAP c, and 5 DAP d. Endosperm (en) and embryo (em) are indicated by arrows. Endosperm is not cellularized until 3 DAP. Starch granules (black dots) can be seen in the cellularized endosperm. e-h Sections of Nip2n × TH4n developing seeds at 2 DAP e, 3 DAP f, 4 DAP g, and 5 DAP h. Endosperm is never cellularized. Embryo arrests at globular stage. i-l Sections of TH4n × Nip2n developing seeds at 2 DAP i, 3 DAP j, 4 DAP k, and 5 DAP l. Precocious cellularization of endosperm occurs at 2 DAP and stop growing at 4 DAP. Embryo appears to be enlarged at 5 DAP. m-p Sections of TH4n × TH4n developing seeds at 2 DAP m, 3 DAP n, 4 DAP o, and 5 DAP p. Development of endosperm and embryo is similar to that in Nip2n × Nip2n The seed derived from the paternal excess cross (Nip2n × TH4n) showed a different pattern of development. At 2 DAP, the seed contained a globular embryo and syncytial endosperm similar to the seed from the balanced crosses (Fig. 1e). However, the embryo arrested at the globular stage (Fig. 1e–h). Endosperm nuclei may have undergone additional divisions but remained uncellularized. At 5 DAP, endosperm nuclei were no longer visible and the embryo sac was full of clear liquid. No starch granules were observed in the endosperm at any stages in the paternal excess cross. The seed derived from the maternal excess cross (TH4n × Nip2n) followed a developmental pattern different from that seen in the above crosses (Fig. 1i–l). By 2 DAP, the endosperm had already cellularized, whereas the embryo was at the globular stage, thus the onset of cellularization in the maternal excess cross was earlier than that in the balanced cross (Fig. 1i). At 3 DAP, starch grains became obvious in cellularized endosperm (Fig. 1j). By 4 and 5 DAP, the endosperm apparently stopped growing but pericarp continued to expand, leaving the arrested endosperm detached from pericarp. The embryos at 5 DAP appeared to be enlarged with a swollen scutellum (Fig. 1k and l). Compared with TH4n × TH4n, the TH4n × Nip2n showed delayed degeneration of the nucellus (Fig. 2l).Fig. 2 Sections of developing rice seeds in parental genome balanced and unbalanced crosses showing deposition of starch grains in pericarp cells. a-d Sections of Nip2n × Nip2n developing seeds focusing on pericarp at 2 DAP a, 3 DAP b, 4 DAP c, 5 DAP d. pericarp (pe), integument (in), nucellar epidermis (ne) and nucellus (nu) are indicated by arrows. e-h Sections of Nip2n × TH4n developing seeds focusing on pericarp at 2 DAP e, 3 DAP f, 4 DAP g, 5 DAP h. i-l Serial sections of TH4n × Nip2n developing seeds focusing on pericarp at 2 DAP i, 3 DAP j, 4 DAP k, and 5 DAP l. m-p Serial sections of TH4n × TH4n developing seeds focusing on pericarp at 2 DAP m, 3 DAP n, 4 DAP o, and 5 DAP p In summary, seed development between the two balanced crosses is similar, following the patterns as previously described (Brown et al. 1996; Itoh et al. 2005). In the maternal excess cross, endosperm was precociously cellularized and stopped growing in early seed development stage, with embryos appearing enlarged and persistent nucellus. However, in the paternal excess cross, the embryo arrested at the globular stage and the endosperm remained uncellularized and became aborted. The developmental patterns of embryo and endosperm in the interploidy crosses in this study are similar to those observed in a study of interploidy crosses in the Japonica background (Sekine et al. 2013). Pericarp Starch Granules in Interploidy Crosses Starch granules in the pericarp of a developing seed are relocated to rapidly growing endosperm tissues thus serving as temporary starch storage in cereals (Caley et al. 1990; Duffus and Rosie 1973; Sato 1984). The contrasting outcomes of endosperm and embryo development, as well as starch accumulation between interploidy crosses, prompted us to investigate the developmental patterns of starch granule deposition in the pericarp. The pericarp starch granules located in parenchyma cells accumulated in a similar manner up to 4 DAP in both the Nip2n × Nip2n and Nip2n × TH4n seeds (Fig. 2a–c and e–g). However, at 5 DAP when the Nip2n × Nip2n seed was rapidly expanding and accumulating starch in the cellularized endosperm (Fig. 1d), the number of starch granules decreased in the pericarp (Fig. 2d). In contrast, the number of starch granules remained unchanged in the pericarp of the Nip2n × TH4n seed (Fig. 2h) where no starch accumulation in the abortive endosperm was observed (Fig. 1h). There was no obvious difference in accumulation of pericarp starch during seed development between the TH4n × Nip2n and TH4n × TH4n crosses (Fig. 2i–p), where both underwent endosperm cellularization (Fig. 1i–p). The starch granules almost completely disappeared in the pericarp parenchyma cells in the TH4n pollinated with either Nip2n or TH4n at 5 DAF (Fig. 2l and p). This suggests that the translocation of the transitory pericarp starch into endosperm is affected by the status of endosperm development with cellularization and/or starch accumulation being the signal for translocation of starch from pericarp, possibly via a feedback mechanism. Expression of PRC2 Gene OsFIE2 and DNA Methyltransferase Gene OsMET1b in the Endosperm of Interploidy Crosses Arabidopsis PRC2 genes and MET1 play important roles in modulating endosperm cellularization and postzygotic hybridization barriers (Erilova et al. 2009; Kradolfer et al. 2013a; Schatlowski et al. 2014). We investigated the expression levels of OsFIE2 and OsMET1b, two major epigenetic modification genes showing function seed development (Luo et al. 2009; Yamauchi et al. 2009; Nallamilli et al. 2013; Hu et al. 2014; Li et al. 2014) in the endosperm derived from the interploidy crosses and the parents Nip2n, TH4n and H2n (Fig. 3). We used 4 DAP endosperm as the 5 DAP endosperm in unbalanced endosperm started to show degradation and extract 3 DAP endosperm may result in maternal tissue comtamination. In the balanced crosses, OsFIE2 was expressed similarly between the three parents in the 4 DAP endosperm (Fig. 3). However, OsFIE2 expression was decreased in the paternal excess cross and elevated in the maternal excess cross, compared to the parents. For OsMET1b, expression levels were similar in the parents and the paternal excess crosses, but was slightly decreased in the maternal excess crosses, indicating that the parental genome imbalance influences the expression levels of these two epigenetic regulators.Fig. 3 Quantitative RT-PCR of OsFIE2 and OsMET1b. a Relative expression levels of OsFIE2 in 4 DAP endosperm from different crosses. b Relative expression level of OsMET1b in 4 DAP endosperm from different crosses Imprinted Genes Were Deregulated in the Interploidy Crosses The imprinted Arabidopsis PRC2 genes are deregulated in interploidy crosses (Jullien and Berger 2010). In rice, the only imprinted PRC2 gene is OsFIE1 (Luo et al. 2009; 2011). We examined the imprinting status of OsFIE1 and other imprinted genes in 4 DAP endosperm derived from unbalanced crosses between TH4n and Nip2n. We also used the reciprocal crosses between Nip2n and H2n as controls. The imprinted genes were identified previously using genome-wide transcriptome analysis of crosses between Nip (Japonica) and 93-11, an Indica rice line (Luo et al. 2011). Due to potential maternal contamination when isolating the poorly developed endosperm in the interploidy crosses, we used three endosperm-specific PEGs for this analysis, in addition to 13 MEGs (Luo et al. 2011). Among the 13 MEGs, two (09 g03500 and 11 g27470) became biallelic in the maternal excess cross, two (01 g10080 and 05 g26040) became biallelic in the paternal excess cross, four showing imprinting in the maternal excess cross lost expression in the paternal excess cross, and the remaining five including OsFIE1 maintained the imprinted expression in both interploidy crosses (Table 2; Additional file 2: Figure S2). For the PEGs, one was imprinted in both interploidy crosses, one became biallelic, and one lost expression in the paternal excess cross (Table 2; Additional file 2: Figure S2). The imprinted genes showing loss of imprinting or expression encode proteins of diverse functions (Table 2; Luo et al. 2011).Table 2 Deregulation of imprinting in interploidy crosses Gene ID Biasa Expa Nip2n × TH4n TH4n × Nip2n Function 01 g10080 mat g bi mat Expressed protein 02 g55560 mat g no-exp mat Protein phosphatase 2C 05 g26040 mat en bi mat Pumilio RNA binding protein 05 g34310 mat en no-exp mat No apical meristem protein 05 g40790 mat en mat mat CCR4-NOT transcription factor 06 g33640 mat en no-exp mat CAPIP1 07 g27359 mat en mat mat Expressed protein 07 g34620 mat g no-exp mat Expressed protein 07 g42390 mat en mat mat Intron derived transcripts 08 g04290 (OsFIE1) mat en mat mat PRC2 member 09 g03500 mat en mat bi ZOS9-01-C2H2 zinc finger 09 g34880 mat en mat mat bZIP domain protein 11 g27470 mat en mat bi Hypothetical protein 07 g17460 pat en pat pat OsFBL36 10 g04890 pat en bi pat Expressed protein 12 g08780 pat en no-exp pat YUCCA like protein aImprinting and expression status in Luo et al’s (2011) data set (bias > 0.9), and imprinting is confirmed using Nip2n and H2n reciprocal crosses (Additional file 2: Figure S2); exp, expression; mat, maternally expressed; pat, paternally expressed; bi, biallelic; en, endosperm-specific; g, non-tissue specific; H2n, the 2n progenitor of TH4n; See Figure S2 for Sanger sequencing results Discussion We have investigated the effects of parental genome imbalance on seed development in reciprocal interploidy crosses between a 2n Japonica and a 4n Indica rice. The developmental patterns of embryos and endosperm in this study are similar to those observed in a previous study using rice lines of different ploidy levels in the same genetic background (Sekine et al. 2013), suggesting genetic backgrounds in our study did not affect the outcomes of rice interploidy crosses. The increased ploidy levels per se in the triploid seeds are unlikely the direct cause for the disruptive development of embryo and endosperm development in interploidy cross, as polyploid embryo and endosperm with balanced parental genomes are usually viable. Although we cannot rule out the possibility that the relative genome dosages of the two fertilization products to the maternal tissues or even to the female gametophyte (Birchler 2014) cause the seed defects of interploidy cross, it has been speculated that the unbalanced parental genomes cause arrested embryo and disruptive seed development primarily via the unbalanced contributions of parentally expression-biased genes (imprinted genes) and altered endosperm development, as indicated by a detailed genetic analysis in maize and Arabidopsis (Lin 1984; Scott et al. 1998). This notion is also supported by the studies in Arabidopsis where knocking-out endosperm specific imprinted genes or regulators of imprinting can alleviate endosperm abortion and promote seed viability in interploidy crosses (Erilova et al. 2009; Kradolfer et al. 2013a; 2013b; Schatlowski et al. 2014). Our observations support the differential roles of the maternal and paternal genomes in endosperm development (Haig and Westoby 1989; 1991; Dilkes and Comai 2004): excess maternal genome drives the precocious endosperm cellularization and excess paternal genome delays or disrupts cellularization, implying the possible roles of imprinting involved in mediating parent-of-origin effects on seed development in different species. Our results also suggest that parental genomes have different roles in influencing the translocation of pericarp starch, which is classified as “transitory starch” and plays an important role in grain development (Duffus and Rosie 1973; Sato 1984; Caley et al. 1990). The lack of translocation of pericarp starch coincided with lack of cellularization and starch deposition in the endosperm of the 2n × 4n cross. Interestingly, compared with the 4n × 4n cross, pericarp starch deposition and translocation appears to be the same in the 4n × 2n cross, in which the precocious cellularization and starch deposition occur in endosperm. It is likely that parental genome balance influences pericarp starch relocation via cellularization. The decline of pericarp starch coincides with the rapid expansion of the developing seeds and fast accumulation of starch in the developing endosperm, suggesting the vital role of this transitory starch in grain filling. Degradation of nucellus is required to facilitate the nutrient supply for young embryo and endosperm growth (Yin and Xue 2012). We observed a delayed degradation of the nucellus in seed of the 2n × 4n cross in which precocious cellularization and poor growth of endosperm occur, supporting that degradation of nucellus and endosperm development are linked processes. Our results also support the idea that endosperm cellularization is a critical developmental transition for embryo differentiation and patterning (Hehenberger et al. 2012). Rice embryo development follows a series of stages identified as zygotic, globular, coleoptiles, juvenile vegetative and maturation (Itoh et al. 2005). The basic plant body plan is established when the embryo reaches the juvenile stage. Our results suggest that the development of endosperm plays a critical role in the differentiation and development of the embryo. Because disruption of endosperm development is the primary cause of seed abortion in parental genome unbalanced crosses (Watkins 1932; Howard 1939; Brink and Cooper 1947; Hakansson 1956; Nishiyama and Inomata 1966; Lin 1984; Ramsey and Schemske 1998; Scott et al. 1998; Gutierrez-Marcos et al. 2003; Stoute et al. 2012; Kradolfer et al. 2013a; Sekine et al. 2013), in the rice paternal excess cross, lack of endosperm cellularization appears hampering further differentiation and development of the embryo, and leads to the embryo arresting at globular stage. In contrast, in the rice maternal excess cross, precocious cellularization leads to the formation of an enlarged embryo showing overgrowth of the scutellum, which may allow the embryo store more nutrients, compensating for the poor reserve of nutrients of the premature endosperm. This suggests that endosperm not only serves as a supplier of nutrients but may also provide signaling molecules at various developmental stages for the differentiation and growth of the embryo. Consistent with this, we have seen better embryo rescue in the maternal excess crosses than in the paternal excess crosses in which embryo arrested early due to lack of endosperm cellularization. Compared with rice embryo, Arabidopsis embryo appears to be more tolerant to interploidy crosses (Scott et al. 1998). The interspecies differences in outcomes of interploidy crosses between rice and Arabidopsis may be related to the intrinsic differences in endosperm development. Rice and Arabidopsis are similar during early endosperm development. However, after cellularization, the endosperm in Arabidopsis disappears and the cotyledons of the embryo replace the nutrient storage function of endosperm. Hence, the endosperm only plays a transient role in supporting embryo development in Arabidopsis. In contrast, rice endosperm continues to proliferate and exists as a major component of the seed. The persistent endosperm is an ideal tissue that provides nutrients and, likely, other signaling molecules for embryo development and germination. Our results showed that expression of the PRC2 gene OsFIE2, a potential imprinting regulator was suppressed in the paternal excess cross, in which endosperm failed to cellularize. This is consistent with the role of H3K27me in rice endosperm development as knocking-down of OsFIE2 by RNAi results in abnormal endosperm development lacking of cellularization (Nallamilli et al. 2013; Li et al. 2014. Conversely, the elevated expression of OsFIE2 and slight down-regulation of OsMET1b were observed in the maternal excess cross, in which endosperm precociously cellularized. Whether these results imply that the rice MET1 and PRC2 genes function as the Arabidopsis counterparts to be involved in regulating postzygotic hybridization barriers needs further investigation. Our results also showed loss of imprinting or expression of imprinted genes in the parental genome unbalanced crosses in rice as in Arabidopsis (Erilova et al. 2009; Jullien and Berger 2010), suggesting that deregulation of imprinting is a common phenomenon across species. It appears that more imprinted genes are affected in the paternal excess cross than in the maternal cross. Whether deregulation of imprinted genes causes endosperm abortion, or vice versa in the interploidy crosses remains to be elucidated. Together, our results provide useful clues for further analysis of the roles of imprinted genes and epigenetic factors in endosperm development. It has been shown that the seed defects in interploidy crosses (Scott et al. 1998; Sekine et al. 2013) are similar to the seed defects in the interspecific crosses (Josefsson et al. 2006; Ishikawa et al. 2011). Increasing the ploidy level of one parent in an interspecific cross was able to improve the viability of hybrid seeds (Johnston et al. 1980; Josefsson et al. 2006), suggesting that there are common mechanisms underlying endosperm abortion in response to both interploidy and interspecific crosses (Haig and Westoby 1991; Schatlowski and Köhler 2012). Imprinting and epigenetic mechanisms regulating imprinting are the promising mechanisms controlling both interspecific and interploidy postzygotic hybridization barrier (Haig and Westoby 1991; Erilova et al. 2009; Kradolfer et al. 2013a; Schatlowski et al. 2014). We reason that low conservation of imprinted genes across species (Luo et al. 2011; Waters et al. 2011; Wolff et al. 2011) contributes to reproductive isolation between related species. Conclusion Our analysis of the differential parental genome dosage effects on the endosperm development in the reciprocal interploidy crosses suggests distinct roles of parental genomes in seed development, with maternal genome functions to promote the endosperm cellularization and paternal genome functions to delay or inhibit cellularization potentially. Starch deposition in endosperm and cellularization regulated by the balance of parental genomes serves as a feedback signal for pericarp starch relocation. Our results support the idea that endosperm cellularization is a critical developmental transition for embryo differentiation and patterning. It is likely that endosperm acts as a sensor in response to post-zygotic incompatibility in interploidy cross via a common mechanism in different species. Methods Rice Materials and Growth Conditions Diploid Indica rice Haitan (H2n), a local cultivar was treated with Colchicine following Beachell and Jones (1945) to generate a tetraploid line designated as Tetra-Haitan (TH4n). Japonica rice Nipponbare (Nip2n) was used as a crossing parent. Diploid and tetraploid lines were reciprocally crossed with each other to obtain seeds with different balances of parental genomes. Chromosome Counting Tillers were separated from rice plants. After removing the old roots, the tillers were washed and immersed in clean water to allow the generation of new roots. Roots of 2-3 cm were used for chromosome counting following the methods of Waninge (1965) and Martínez-Gómez et al. (2005). Embryo Rescue Embryos were rescued following the procedure described by Jena and Khush (1984a, b). Embryos of different stages in interploidy or intraploidy crosses were placed on 1/4 MS medium in the dark until root and shoot growth were initiated, and then incubated under light for ~12 days until leaves formed. Rescued seedlings were transferred to full MS liquid medium to promote root growth and then transplanted to soil. Microscopy Pollinated ovaries for serial sectioning were fixed in glutaraldehyde and embedded in Spurr’s resin as described previously (Koltunow et al. 1998). Ovaries were also fixed in FAA and cleared using methyl salicylate in order to examine a greater number of samples, as per Koltunow et al. (2011). All plant material was harvested around 4 pm to avoid the potential fluctuation of starch accumulation during the day. Sections and whole mount ovules were examined using a Zeiss Axioplan microscope and digital images were collected using a Spot II camera. For each stage, at least 8 seeds for each of the four crosses were examined either by clearing and observation under DIC Microscopy, or by sectioning. RT-PCR and qRT-PCR and Sequencing Total RNA was extracted using an RNeasy Mini kit (Qiagen,http://www.qiagen.com/) and treated with RNase-free DNase (Promega,http://www.promega.com/). cDNA was synthesized from 100 ng total mRNA using PRIME Script reverse transcriptase (Takara, http://www.takara-bio.com/) according to the manufacturer’s instructions. Primers used in this study are shown in Additional file 3: Table S1. The TaKaRa SYBR Prenmix Ex Taq™II was used for real time PCR, and data were collected using the BIO-RAD CFX96™ Real-Time System. Relative expression levels from the real-time PCR analysis were normalized against UBIQUITIN and EEF-1A. Bio-Rad CFX Manager V1.6.541.1028 was used to analyse gene expression level. Additional files Additional file 1: Chromosome number counting and seed phenotypes of interploidy crosses. (A-D) Chromosome numbers of root tip cells of Nip2n (A), TH4n (B), 3n plant of Nip2n × TH4n (C), and 3n plant of TH4n × Nip2n (D). (E) Seed phenotypes of balanced and unbalanced crosses. (DOCX 2648 kb) Additional file 2: Sanger sequencing results of selected imprinted genes. (PDF 712 kb) Additional file 3: List of RT-PCR primers. (DOCX 18 kb) Acknowledgement This work was supported by the Department of Innovation, Industry, Science and Research (DIISR) Australia/India grant (no.TA01-0001 to Anna.M.Koltunow), a Science and Industry Endowment Fund (SIEF) grant (no. RP01-006 to Anna.M.Koltunow). Grants (2013HH0023, 12ZA123, NSFC31301049) from Chinese central or local governments to support Hongyu Zhang and Xianjun Wu are acknowledged. Authors’ Contributions ML, HZ, SDJ, X-JW participated in the design of the study and its coordination. X-WZ, LL participated in microscopy and embryo rescue. FH, P-ZX participated in RT-PCR and qRT-PCR. Y-TL participated in writing. ML, X-JW as corresponding authors. All authors contributed to the manuscript draft, read and approved the final manuscript. Authors’ Information Hongyu Zhang, Peizhou Xu: Assistant Professor at Sichuan Agricultural University. Ming Luo: Research Scientist, Susan D. Johnson: Project officer at Commonwealth Scientific and Industrial Research Organisation (CSIRO). Xiao-wei Zhu: a student at Chinese Academy of Sciences. Lei Liu, Fang Huang, Yu-tong Liu: is studying at Sichuan Agricultural University. Xian-jun Wu: Professor at Sichuan Agricultural University. Competing Interests The authors declare that they have no competing interests. ==== Refs References Adams S Vinkenoog RSpielman M Dickinson HG Scott RJ Parent-of-origin effects on seed development in Arabidopsis thaliana require DNA methylation Development 2000 127 2493 2502 10804189 Beachell HM Jones JW Tetraploids Induced in Rice by Temperature and Colchicine Treatments Agron J 1945 37 165 175 10.2134/agronj1945.00021962003700030001x Berger F Vu TM Li J Hypothesis: selection of imprinted genes is driven by silencing deleterious gene activity in somatic tissues Cold Spring Harb Symp Quant Biol 2012 77 23 29 10.1101/sqb.2012.77.014514 23250991 Birchler JA Dosage analysis of maize endosperm development Annu Rev Genet 1993 27 181 204 10.1146/annurev.ge.27.120193.001145 8122901 Birchler JA Interploidy hybridization barrier of endosperm as a dosage interaction Front Plant Sci 2014 5 5 281 25018757 Brink RA Cooper DC The endosperm in seed development Bot Rev 1947 13 479 541 10.1007/BF02861549 Brown RC Lemmon BE Olsen OA Development of the endosperm in rice (Oryza sativa L.): Cellularization J Plant Res 1996 109 301 313 10.1007/BF02344477 Caley CY Duffus CM Jeffcoat B Photosynthesis in the pericarp of developing wheat grains J Exp Bot 1990 41 303 307 10.1093/jxb/41.3.303 Dilkes BP Comai L A differential dosage hypothesis for parental effects in seed development Plant Cell 2004 16 3174 3180 10.1105/tpc.104.161230 15579806 Du M Luo M Zhang R Finnegan EJ Koltunow AM Imprinting in rice: the role of DNA and histone methylation in modulating parent-of-origin specific expression and determining transcript start sites Plant J 2014 79 232 242 10.1111/tpj.12553 24819479 Duffus C Rosie R Starch hydrolysing enzymes in the developing barley grain Planta 1973 109 153 160 10.1007/BF00386123 24474061 Erilova A Brownfield L Exner V Rosa M Twell D Scheid OM Hennig L Köhler C Imprinting of the Polycomb group gene MEDEA serves as a ploidy sensor in Arabidopsis PLoS Genet 2009 5 e1000663 10.1371/journal.pgen.1000663 19779546 Gutierrez-Marcos JF Pennington PD Costa LM Dickinson HG Imprinting in the endosperm: a possible role in preventing wide hybridization Philos Trans R Soc Lond B Biol Sci 2003 358 1105 1111 10.1098/rstb.2003.1292 12831476 Haig D Westoby M Parent specific gene expression and the triploid endosperm Am Nat 1989 134 147 155 10.1086/284971 Haig D Westoby M Genomic imprinting in endosperm: its effects on seed development in crosses between species and between different ploidies of the same species, and its implications for the evolution of apomixis Philos Trans Biol Sci 1991 333 1266 1 13 10.1098/rstb.1991.0057 Hakansson A Seed development of Brassica oleracea and B. rapa after certain reciprocal pollinations Hereditas 1956 42 373 396 10.1111/j.1601-5223.1956.tb03024.x Hehenberger E1 Kradolfer D Köhler C Endosperm cellularization defines an important developmental transition for embryo development Development 2012 139 2031 2039 10.1242/dev.077057 22535409 Howard H The size of seeds in diploid and autotetraploid Brassica oleracea L J Genet 1939 38 325 340 10.1007/BF02982177 Hu L Li N Xu C Zhong S Lin X Yang J Zhou T Yuliang A Wu Y Chen YR Cao X Zemach A Rustgi S von Wettstein D Liu B Mutation of a major CG methylase in rice causes genome-wide hypomethylation, dysregulated genome expression, and seedling lethality Proc Natl Acad Sci U S A 2014 111 10642 10647 10.1073/pnas.1410761111 25002488 Ishikawa R Ohnishi T Kinoshita Y Eiguchi M Kurata N Rice interspecies hybrids show precocious or delayed developmental transitions in the endosperm without change to the rate of syncytial nuclear division Plant J 2011 65 798 806 10.1111/j.1365-313X.2010.04466.x 21251103 Itoh J Nonomura K Ikeda K Yamaki S Inukai Y Rice plant development: from zygote to spikelet Plant Cell Physiol 2005 46 23 47 10.1093/pcp/pci501 15659435 Jena KK Khush GS Embryo rescue of interspecific hybrids and its scope in rice improvement Rice Genet Newsl 1984 1 133 134 Johnston SA Dennijs TPM Peloquin SJ Hanneman RE The significance of genic balance to endosperm development in interspecific crosses TheorAppl Genet 1980 57 5 9 10.1007/BF00276002 Josefsson C Dilkes B Comai L Parent-dependent loss of gene silencing during interspecies hybridization Curr Biol 2006 16 1322 1328 10.1016/j.cub.2006.05.045 16824920 Jullien PE Berger F Parental genome dosage imbalance deregulates imprinting in Arabidopsis PLoS Genet 2010 6 e1000885 10.1371/journal.pgen.1000885 20333248 Koltunow AM Johnson SD Bicknell RA Sexual and apomictic development in Hieracium Sex Plant Reprod 1998 11 213 230 10.1007/s004970050144 Koltunow AM Johnson SD Rodrigues JC Okada T Hu Y Tsuchiya T Wilson S Fletcher P Ito K Suzuki G Mukai Y Fehrer J Bicknell RA Sexual reproduction is the default mode in apomictic Hieracium subgenus Pilosella, in which two dominant loci function to enable apomixis Plant J 2011 66 890 902 10.1111/j.1365-313X.2011.04556.x 21418351 Kradolfer D Hennig L Köhler C Increased maternal genome dosage bypasses the requirement of the FIS Polycomb Repressive Complex 2 in Arabidopsis seed development PLoS Genet 2013 9 e1003163 10.1371/journal.pgen.1003163 23326241 Kradolfer D Wolff P Jiang H Siretskiy A Köhler C An imprinted gene underlies postzygotic reproductive isolation in Arabidopsis thaliana Dev Cell 2013 26 525 535 10.1016/j.devcel.2013.08.006 24012484 Li S Zhou B Peng X Kuang Q Huang X Yao J Du B Sun MX OsFIE2 plays an essential role in the regulation of rice vegetative and reproductive development New Phytol 2014 201 66 79 10.1111/nph.12472 24020752 Lin BY Ploidy barrier to endosperm development in maize Genetics 1984 107 103 115 17246209 Luo M Bilodeau P Dennis ES Peacock WJ Chaudhury A Expression and parent-of-origin effects for FIS2, MEA, and FIE in the endosperm and embryo of developing Arabidopsis seeds Proc Natl Acad Sci 2000 97 10637 10642 10.1073/pnas.170292997 10962025 Luo M Platten D Chaudhury A Peacock WJ Dennis ES Expression, imprinting, and evolution of rice homologs of the polycomb group genes Mol Plant 2009 2 711 723 10.1093/mp/ssp036 19825651 Luo M Taylor JM Spriggs A Zhang H Wu X A genome-wide survey of imprinted genes in rice seeds reveals imprinting primarily occurs in the endosperm PLoS Genet 2011 7 e1002125 10.1371/journal.pgen.1002125 21731498 Martínez-Gómez P Sánchez-Pérez R Vaknin Y Dicenta F Gradziel TM Improved technique for counting chromosomes in almond Sci Hortic 2005 105 139 143 10.1016/j.scienta.2005.01.019 Nallamilli BR Zhang J Mujahid H Malone BM Bridges SM Peng Z Polycomb group gene OsFIE2 regulates rice (Oryza sativa) seed development and grain filling via a mechanism distinct from arabidopsis PLoS Genet 2013 9 e1003322 10.1371/journal.pgen.1003322 23505380 Nishiyama I Inomata N Embryological studies on cross-incompatibility between 2x and 4x in Brassica Jpn J Genet 1966 41 27 42 10.1266/jjg.41.27 Ramsey J Schemske DW Pathways, mechanisms, and rates of polyploid formation in flowering plants Annu Rev Ecol Systematics 1998 29 467 501 10.1146/annurev.ecolsys.29.1.467 Rodrigues J Ruan R Nishimura T Sharma M Sharma R Ronald P Fischer R Zilberman D Imprinted expression of genes and small rna is associated with localized hypomethylation of the maternal genome in rice endosperm Proc Natl Acad Sci U S A 2013 110 7934 7939 10.1073/pnas.1306164110 23613580 Sato K Starch granules in tissues of rice plants and their changes in relation to plant growth JARQ 1984 18 78 86 Schatlowski N Köhler C Tearing down barriers: understanding the molecular mechanisms of interploidy hybridizations J Exp Bot 2012 63 6059 6067 10.1093/jxb/ers288 23105129 Schatlowski N Wolff P Santos-González J Schoft V Siretskiy A Scott R Tamaru H Köhler C Hypomethylated pollen bypasses the interploidy hybridization barrier in Arabidopsis Plant Cell 2014 26 3556 3568 10.1105/tpc.114.130120 25217506 Scott R Spielman M Dickinson HG Parent-of-origin effects on seed development in Arabidopsis thaliana Development 1998 125 17 3329 3341 9693137 Sekine D Ohnishi T Furuumi H Ono A Yamada T Kurata N Kinoshita T Dissection of two major components of the post-zygotic hybridization barrier in rice endosperm Plant J 2013 76 792 799 10.1111/tpj.12333 24286595 Stoute AI Varenko V King GJ Scott RJ Kurup S Parental genome imbalance in Brassica oleracea causes asymmetric triploid block Plant J 2012 71 3 503 516 22679928 Waninge J A modified method of counting chromosomes in root tip cells of wheat Euphytica 1965 14 249 250 10.1007/BF00149508 Waters AJ Makarevitch I Eichten SR Swanson-Wagner RA Yeh CT Xu W Schnable PS Vaughn MW Gehring M Springer NM Parent-of-origin effects on gene expression and DNA methylation in the maize endosperm Plant Cell 2011 23 4221 4233 10.1105/tpc.111.092668 22198147 Watkins AE Hybrid sterility and incompatibility J Genet 1932 25 125 162 10.1007/BF02983249 Wolff P Weinhofer I Seguin J Roszak P Beisel C Donoghue MT Spillane C Nordborg M Rehmsmeier M Kohler C High resolution analysis of parent-of-origin allelic expression in the Arabidopsis endosperm PLoS Genet 2011 7 e1002126 10.1371/journal.pgen.1002126 21698132 Yamauchi T Johzuka-Hisatomi Y Fukada-Tanaka S Terada R Nakamura I Iida S Homologous recombination-mediated knock-in targeting of the MET1a gene for a maintenance DNA methyltransferase reproducibly reveals dosage-dependent spatiotemporal gene expression in rice Plant J 2009 60 386 396 10.1111/j.1365-313X.2009.03947.x 19519802 Yin LL Xue HW The MADS29 transcription factor regulates the degradation of the nucellus and the nucellar projection during rice seed development Plant Cell 2012 24 1049 1065 10.1105/tpc.111.094854 22408076

PMC005xxxxxx/PMC5002276.txt

==== Front SpringerplusSpringerplusSpringerPlus2193-1801Springer International Publishing Cham 307710.1186/s40064-016-3077-5ResearchA novel computational approach to approximate fuzzy interpolation polynomials Jafarian Ahmad jafarian5594@yahoo.com 1Jafari Raheleh jafari3339@yahoo.com 2Mohamed Al Qurashi Maysaa Maysaa@ksu.edu.sa 3Baleanu Dumitru dumitru@cankaya.edu.tr 451 Department of Mathematics, Urmia Branch, Islamic Azad University, Urmia, Iran 2 Departamento de Control Automático, CINVESTAV-IPN (National Polytechnic Institute), Mexico City, Mexico 3 Department of Mathematics, King Saud University, Riyadh, 11495 Saudi Arabia 4 Department of Mathematics, Faculty of Art and Sciences, Cankaya University, 06530 Balgat, Ankara, Turkey 5 Institute of Space Sciences, Magurele-Bucharest, Romania 27 8 2016 27 8 2016 2016 5 1 142820 6 2016 15 8 2016 © The Author(s) 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.This paper build a structure of fuzzy neural network, which is well sufficient to gain a fuzzy interpolation polynomial of the form yp=anxpn+⋯+a1xp+a0 where aj is crisp number (for j=0,…,n), which interpolates the fuzzy data (xj,yj)(forj=0,…,n). Thus, a gradient descent algorithm is constructed to train the neural network in such a way that the unknown coefficients of fuzzy polynomial are estimated by the neural network. The numeral experimentations portray that the present interpolation methodology is reliable and efficient. Keywords Fuzzy neural networksFuzzy interpolation polynomialCost functionLearning algorithmissue-copyright-statement© The Author(s) 2016 ==== Body Background Artificial neural networks (ANNs) are mathematical or computational models based on biological neural networks. Neural networks consist of universal approximation potentiality, and they function best when the system has a high endurance to error when used to model. Recently, there have been rapid growth of ANNs which was utilized in various fields (Abbasbandy and Otadi 2006; Chen and Zhang 2009; Guo and Qin 2009; Jafarian and Jafari 2012; Jafarian et al. 2015a, b; Jafarian and Measoomynia 2011, 2012; Song et al. 2013; Wai and Lin 2013). One of the vital roles of ANN is finding FIPs as it proposed in this research. Interpolation theory is one of the basic tool in applied and numerical mathematics. Interpolation has been used extensively, because it is one of the noteworthy techniques of function approximation (Boffi and Gastaldi 2006; Mastylo 2010; Rajan and Chaudhuri 2001). Using Newton’s divided difference scheme, a new technique was established in Schroeder et al. (1991) for polynomial interpolation. The problem related to multivariate interpolation has grabbed the attention of researchers world wide (Neidinger 2009; Olver 2006). There are various multivariate interpolation methods. In Olver (2006) they used a multivariate Vandermode matrix and its LU factorization, and Neidinger (2009) utilized the Newton-form interpolation. We recall that sparse grid interpolation is a further technique. In recent years this procedure is widely executed for the provision of an average approximation to a smooth function (Xiu and Hesthaven 2005). Utilizing the Lagrange interpolating polynomials, this approach introduces a polynomial interpolant on the basis of amounts of the function at the points in an amalgamation of product grids of minute dimension (Barthelmann et al. 2000). Existing trends on interpolation networks, have been revealed in Llanas and Sainz (2006), Sontag (1992). Numerable proof based on the notation that single hidden layer FNNs taking into account m+1 neurons, is able to learn m+1 isolated data (xi,fi)(fori=0,…,m) with zero error has been established in Ito (2001). The detailed introduction and survey of major results can be extracted from Refs. Szabados and Vertesi (1990), Tikhomirov (1990). This paper is inclined to the motive in order to deliver a fuzzy modeling technique by the utilization of FNNs for finding a FIP of the form 1 yp=anxpn+⋯+a1xp+a0, where ajϵR(forj=0,…,n), which interpolates the fuzzy data (xj,yj)(forj=0,…,n). The proposed network is a formation, abiding of three layers whereas the extension principle of Zadeh (2005) elaborately describes the input-output connection of each unit. In the latest model, the unrevealed coefficients of fuzzy polynomial can be approximated by employing a cost function. Moreover, a learning technique which is associated with gradient decent procedure is formulated for the adjustment of connection weights to any achievable degree of precision. This paper starts with a summary explanation of fuzzy numbers and fuzzy interpolation, then we provide the method of FNN for finding the crisp solution of the FIP. Two numerical examples are proposed to establish the validity and performance of the justified approach in “Numerical examples” section. Finally, “Concluding remarks” section presents the conclusions. Method description Basically, the interpolation theory has a wide range of applications in mathematical analysis. In numerical analysis, the interpolation is a method or operation of finding from a few given terms of a series, as of numbers or observations, other intermediate terms in conformity with the law of the series. Generally, the interpolation techniques are in phase with elementary model of an interpolating function which can be stated as: 2 s:R→R,s(x)=∑j=1nyj·ϕj(x), with the basis function ϕj(x):R→R that elates in interpolation criteria: ϕj(xk)=1,fork=j,0,fork≠j. Suppose that x^1,…,x^n be n fuzzy points in En whereas, a fuzzy number y^j∈E is in direct relation with each x^j for j=1,…,n. The sought function can be portrayed as follows: 3 s^:En→E:s^x^=∑j=1ny^j·ϕ^jx^, where the ϕ^j:En→E for j=1,…,n exhibit fuzzy functions that compensate the stipulation of interpolation: ϕ^jx^k=1,fork=j,0,fork≠j. Fuzzy interpolation polynomial The interested are vested in finding FIP of the form 4 yp=anxpn+⋯+a1xp+a0, where ajϵR(forj=0,…,n), that interpolates the fuzzy data (xj,yj)(forj=0,…,n). Taking into account a three layer FNN construction which is displayed by Fig. 1. The input-output connection of each unit of the offered neural network can be portrayed as mentioned below, when the α-level sets of the fuzzy input xp is nonnegative, i.e., 0≤[xp]lα≤[xp]uα:Input unit 5 [o]α=[xp]lα,[xp]uα,p=0,…,n. Hidden units 6 [Oj]α=f[netj]lα,[netj]uα=[o]lαj,[o]uαj,j=1,…,n. Output unit 7 [yp]α=F[Net]lα+a0,[Net]uα+a0=[Net]lα+a0,[Net]uα+a0,p=0,…,n, We have [Net]lα=∑jϵM[Oj]lα·aj+∑jϵC[Oj]uα·aj, and [Net]uα=∑jϵM[Oj]uα·aj+∑jϵC[Oj]lα·aj, where M={aj≥0},C={aj<0} and M∪C={1,…,n}. Fig. 1 Fuzzy neural network equivalent to fuzzy interpolation polynomial Cost function The input signals xp(forp=0,…,n) are represented to the network and then yn(xp) which is an representing the network output upon the presentation of aj(forj=0,…,n), is calculated. Defining of cost function over the model parameters makes it a good forecaster. The mean squared error is termed to be as one of the vastly popular usable cost function. Now, let the α-level sets of the fuzzy target output dp are exhibited as: [dp]α=[dp]lα,[dp]uα,α∈[0,1], A cost function which is required to be diminished is stated for each α-level sets as depicted: 8 ep(α)=epl(α)+epu(α),p=0,…,n, where 9 epl(α)=α·[dp]lα-[yp]lα22, 10 epu(α)=α·[dp]uα-[yp]uα22. Generally the summed up error of the suggested neural network is extracted by: 11 e=∑α∑p=0nep(α). Obviously, e⟶0 means [yp]α⟶[dp]α. Fuzzy neural network learning approach Suppose connection weights aj(forj=0,…,n) are randomly actuated by crisp numbers. Now tweaked rule is illustrated as (Ishibuchi et al. 1995): 12 aj(t+1)=aj(t)+Δaj(t),Δaj(t)=-η·∂ep(α)∂aj+γ·Δaj(t-1), where t denotes the number of moderation, η signifies the rate of learning and γ implies as the stationary momentum term. We calculate ∂ep(α)∂aj as follows: 13 ∂ep(α)∂aj=∂epl(α)∂aj+∂epu(α)∂aj. Hence complexities lies in the calculation of the derivatives ∂epl(α)∂aj and ∂epu(α)∂aj. So we have: ∂epl(α)∂aj=∂epl(α)∂[yp]lα·∂[yp]lα∂[Net]lα·∂[Net]lα∂aj=-α·[dp]αl-[yp]αl.∂[Net]lα∂aj,j=1,…,n, and ∂epl(α)∂aj=∂epl(α)∂[yp]lα·∂[yp]lα∂aj=-α·[dp]αl-[yp]αl,j=0, where ∂[Net]lα∂aj=[Oj]lα,aj≥0,[Oj]uα,aj<0, also we have ∂epu(α)∂aj=∂epu(α)∂[yp]uα·∂[yp]uα∂[Net]uα·∂[Net]uα∂aj=-α·[dp]αu-[yp]αu·∂[Net]uα∂aj,j=1,…,n, and ∂epu(α)∂aj=∂epu(α)∂[yp]uα·∂[yp]uα∂aj=-α·[dp]αu-[yp]αu,j=0, where ∂[Net]uα∂aj=[Oj]uα,aj≥0,[Oj]lα,,aj<0, Upper bound approximation Theorem 1 Supposep:R→Ris a continuous function, hence for each compact setϑ⊂E0(the set of all the bounded fuzzy set), andψ>0,there arem∈N,anda0,ai∈R,i=1,2,…,m,which imply 14 ∀x^∈ϑand∀x˘∈R,dpx˘,∑i=1mpix^ai+a0<ψ, whereψis a finite number. Proof The proof of theorem can be followed from the below results. □ If p:R→R, by applying the methodology of the extension principle, p can be extended to the fuzzy function that denotes by p:E0→E as follows: 15 ∀u∈E0,p(u)(y)=⋁px^=yux^y∈R, p is termed as expanded function. Also, cc(R) implies the bounded set of closed intervals of R. clearly 16 u∈E0⟹∀α∈(0,1],[u]α∈cc(R). Moreover 17 Supp(u)∈cc(R). Henceforth, we let 18 Supp(u)=[s1(u),s2(u)]. Theorem 2 Supposep:R→Ris a continuous function, hence for each compact setϑ⊂E0,ϱ>0and arbitraryε>0,existm∈N,anda0,ai∈R,i=1,2,…,m,implicate 19 ∀x^∈ϑ,dpx^,∑i=1mpix^ai+a0<ϱ, whereϱis a finite number. The bottom and top bounds of theα-level set of fuzzy function diminish toϱ,but the center goes toε. Proof Because ϑ⊂E0 is a compact set, and so by Lemma 3, it can be supposed that V⊂R be the compact set associated to ϑ.∀ε>0, therefore by the final outcome in Cybenko (1989), exist m∈N, and a0,ai∈R,i=1,2,…,m, which imply 20 ∀x^∈V,p(x^)-∑i=1mpi(x^)ai+a0<ε, holds. Let q(x^)=∑i=1mpi(x^)ai+a0,x^∈R, then 21 ∀x^∈V,px^-qx^<ε. Theorem 4 implies the validity of (19). □ Lemma 3 Ifϑ⊂E0be a compact set, henceϑis uniformly support-bounded, i.e. exists a compact setV⊂R,implicates∀u∈ϑ,Supp(u)⊂V. Theorem 4 Supposingϑ⊂E0be compact, V the corresponding compact set ofϑ,andp,q:R→Rare the continuous functions that compensate the relation mentioned below 22 ∀x^∈V,px^-qx^<k,k>0. Then∀u∈ϑ,d(p(u)-q(u))≤k. Proof See Liu (2000). □ Numerical examples The following examples has been used to narrate the methodology proposed in this paper. Example 5 The connection between three tanks and pipeline which is denoted by a constant H is represented by Fig. 2. It is a requirement to pump water in order to transfer it from one tank to the further two tanks. The mentioned system suffice the relation mentioned below H=A0⊕A1F1⊕A2F2⊕A3F3 here F1=2x,F2=xx,F3=x3 are considered to be the flow quantity, where x is taken to be the elapsed time. The height of the pipe is mentioned by the term H,A0,A1,A2 and A3 are the pump characteristic coefficients, to be mentioned A0=2,A1=4,A2=3,A3=5 In below, four real uncertain data have been mentioned x={6,(1,3,4),3,(2,3,4,6)} The iteration of data is continued for 19 times. H=1139.9472,(15.6568,162.3859,357.3137),162.3893,(58.4852,162.3859,357.3137,1139.9456) Fig. 2 Pumping of water in order to transfer it from one tank to the further two tanks We use x0=5,x1=7,x2=6,x3=8,η=1×10-2 and γ=1×10-2 for FNN. The approximation results are depicted in Table 1. The precision level of the solutions x0(t),x1(t),x2(t) and x3(t) are shown in Fig. 3, t implies the iterative numbers. It is eminent that by incrementing the iterations, the cost function diminishes to zero. The convergency criteria of the approximated solutions are portrayed using Figs. 4, 5, 6 and 7. For the purpose of attaining the exact solutions, the iterations in the figures have to be increased.Table 1 Neural network approximation for the coefficients t x0(t) x1(t) x2(t) x3(t) Error for FNN 1 4.9018 6.9215 5.9307 7.9121 58,756.65 2 4.5321 6.6450 5.5480 7.6010 6479.790 3 4.0231 6.2056 5.1250 7.2212 1741.483 4 3.6850 5.8401 4.7851 6.7945 577.7597 5 3.2032 5.4001 4.3365 6.3330 210.8822 ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ 15 2.0008 4.0007 3.0008 5.0006 0.366883 16 2.0007 4.0005 3.0006 5.0005 0.151818 17 2.0005 4.0004 3.0005 5.0003 0.062895 18 2.0004 4.0003 3.0004 5.0002 0.026075 19 2.0003 4.0002 3.0003 5.0001 0.010815 Fig. 3 The error between the approximate solution and the exact solution Fig. 4 The approximated solution approaches to the exact one Fig. 5 The approximated solution approaches to the exact one Fig. 6 The approximated solution approaches to the exact one Fig. 7 The approximated solution approaches to the exact one Example 6 Contemplate the sequential interpolation points: ((1,2,3);(-54,-29,-12)),((3,4,6);(-177,-87,-54)),((2,3,5);(-128,-54,-29)) The exact solution for the given problem can be stated as: y=-4x2-5x-3. This constrained is resolved by utilizing the technique of neural network suggested in this context, assuming x0=-0.5,x1=-2.5,x2=-1.5,η=3×10-2 and γ=3×10-2. The approximation results are depicted in Table 2. The precision level of the solutions x0(t),x1(t) and x2(t) are shown in Fig. 8, t implies the number of iterations.Table 2 Neural network approximation for the coefficients t x0(t) x1(t) x2(t) Error for FNN 1 −0.5915 −2.5895 −1.5784 2330.5296 2 −0.9910 −2.9033 −1.9664 1896.6752 3 −1.3356 −3.3346 −2.3696 999.56201 4 −1.8050 −3.8798 −2.7561 401.56201 5 −2.2257 −4.1035 −3.1100 95.188500 ⋮ ⋮ ⋮ ⋮ ⋮ 13 −2.9996 −4.9995 −3.9996 0.86688366 14 −2.9998 −4.9996 −3.9998 0.54635274 15 −2.9999 −4.9998 −3.9999 0.23614301 16 −3.0000 −4.9999 −4.0000 0.06896850 17 −3.0000 −5.0000 −4.0000 0.02003805 Fig. 8 The error between the approximate solution and the exact solution Concluding remarks This research introduces a new methodology in order to find a FIP which interpolates the fuzzy data (xj,yj)(forj=0,…,n). To achieve this goal, a FNN equivalent to FIP was built, and a fast learning algorithm was defined for approximating the crisp unknown coefficients of the given polynomial. The proposed method was based on approximating FNN and the MATLAB software is used for the simulations. The innovative method was validated with two examples. The simulation results clearly illustrated the efficiency and computational advantages of the proposed technique. In particular, the error of approximation is minute. Authors' contributions All authors contributed equally to this work. All authors read and approve the final manuscript. Acknowledgements The research is supported by a grant from the “Research Center of the Center for Female Scientific and Medical Colleges”, Deanship of Scientific Research, King Saud University. The authors are also thankful to visiting professor program at King Saud University for support. Competing interests The authors declare that they have no competing interests. ==== Refs References Abbasbandy S Otadi M Numerical solution of fuzzy polynomials by fuzzy neural network Appl Math Comput 2006 181 1084 1089 Barthelmann V Novak E Ritter K High dimensional polynomial interpolation on sparse grids Adv Comput Math 2000 12 273 288 10.1023/A:1018977404843 Boffi D Gastaldi L Interpolation estimates for edge finite elements and application to band gap computation Appl Numer Math 2006 56 1283 1292 10.1016/j.apnum.2006.03.014 Chen Lh Zhang Xy Application of artificial neural networks to classify water quality of the yellow river Fuzzy Inf Eng 2009 9 15 23 10.1007/978-3-540-88914-4_3 Cybenko G Approximation by superpositions of a sigmoidal function Math Control Signals Syst 1989 2 303 314 10.1007/BF02551274 Guo B Qin L Tactile sensor signal processing with artificial neural networks Fuzzy Inf Eng 2009 54 54 62 10.1007/978-3-540-88914-4_8 Ishibuchi H Kwon K Tanaka H A learning of fuzzy neural networks with triangular fuzzy weghts Fuzzy Sets Syst 1995 71 277 293 10.1016/0165-0114(94)00281-B Ito Y Independence of unscaled basis functions and finite mappings by neural networks Math Sci 2001 26 117 126 Jafarian A Measoomynia S Solving fuzzy polynomials using neural nets with a new learning algorithm Appl Math Sci 2011 5 2295 2301 Jafarian A, Jafari R (2012) Approximate solutions of dual fuzzy polynomials by feed-back neural networks. J Soft Comput Appl. doi:10.5899/2012/jsca-00005 Jafarian A, Measoomynia S (2012) Utilizing feed-back neural network approach for solving linear Fredholm integral equations system. Appl Math Model. doi:10.1016/j.apm Jafarian A Jafari R Khalili A Baleanud D Solving fully fuzzy polynomials using feedback neural networks Int J Comput Math 2015 92 742 755 10.1080/00207160.2014.907404 Jafarian A Measoomy S Abbasbandy S Artificial neural networks based modeling for solving Volterra integral equations system Appl Soft Comput 2015 27 391 398 10.1016/j.asoc.2014.10.036 Liu P Analyses of regular fuzzy neural networks for approximation capabilities Fuzzy Sets Syst 2000 114 329 338 10.1016/S0165-0114(98)00248-6 Llanas B Sainz FJ Constructive approximate interpolation by neural networks J Comput Appl Math 2006 188 283 308 10.1016/j.cam.2005.04.019 Mastylo M Interpolation estimates for entropy numbers with applications to non-convex bodies J Approx Theory 2010 162 10 23 10.1016/j.jat.2009.02.002 Neidinger RD (2009) Multivariable interpolating polynomials in newton forms. In: Joint mathematics meetings, Washington, DC, pp 5–8 Olver PJ On multivariate interpolation Stud Appl Math 2006 116 201 240 10.1111/j.1467-9590.2006.00335.x Rajan D Chaudhuri S Generalized interpolation and its application in super-resolution imaging Image Vis Comput 2001 19 957 969 10.1016/S0262-8856(01)00055-5 Schroeder H Murthy VK Krishnamurthy EV Systolic algorithm for polynomial interpolation and related problems Parallel Comput 1991 17 493 503 10.1016/S0167-8191(05)80151-6 Song Q Zhao Z Yang J Passivity and passification for stochastic TakagiSugeno fuzzy systems with mixed time-varying delays Neurocomputing 2013 122 330 337 10.1016/j.neucom.2013.06.018 Sontag ED Feedforward nets for interpolation and classification J Comput Syst Sci 1992 45 20 48 10.1016/0022-0000(92)90039-L Szabados J Vertesi P Interpolation of functions 1990 Singapore World Scientific Tikhomirov VM Gamkrelidze RV Approximation theory, analysis II Encyclopaedia of mathematical sciences 1990 Berlin Springer Wai RJ Lin YW Adaptive moving-target tracking control of a vision-based mobile robot via a dynamic petri recurrent fuzzy neural network IEEE Trans Fuzzy Syst 2013 21 688 701 10.1109/TFUZZ.2012.2227974 Xiu D Hesthaven JS High-order collocation methods for differential equations with random inputs SIAM J Sci Comput 2005 27 18 39 10.1137/040615201 Zadeh LA Toward a generalized theory of uncertainty (GTU) an outline Inf Sci 2005 172 1 40 10.1016/j.ins.2005.01.017

PMC005xxxxxx/PMC5002277.txt

==== Front ImmunogeneticsImmunogeneticsImmunogenetics0093-77111432-1211Springer Berlin Heidelberg Berlin/Heidelberg 92610.1007/s00251-016-0926-xReviewMammalian CD1 and MR1 genes Reinink Peter 12Van Rhijn Ildiko +1 617 525 1023i.vanrhijn@uu.nl 121 Division of Rheumatology, Immunology and Allergy, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA USA 2 Department of Infectious Diseases and Immunology, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands 28 7 2016 28 7 2016 2016 68 8 515 523 15 4 2016 22 6 2016 © The Author(s) 2016 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.All higher vertebrates share the fundamental components of the adaptive immune system: the B cell receptor, the T cell receptor, and classical MHC proteins. At a more detailed level, their immune systems vary considerably, especially with respect to the non-polymorphic MHC class I-like proteins. In mammals, the CD1 family of lipid-presenting proteins is encoded by clusters of genes of widely divergent sizes and compositions. Another MHC class I-like protein, MR1, is typically encoded by a single gene that is highly conserved among species. Based on mammalian genomes and the available data on cellular expression profiles and protein structure, we review MR1 genes and families of CD1 genes in modern mammals from a genetic and functional perspective. Understanding the CD1 and MR1 systems across animal species provides insights into the specialized functions of the five types of CD1 proteins and facilitates careful consideration of animal models for human diseases in which immune responses to lipids and bacterial metabolites play a role. Keywords CD1Lipid antigensMammalsMR1http://dx.doi.org/10.13039/501100003246Nederlandse Organisatie voor Wetenschappelijk Onderzoek824.02.002Van Rhijn Ildiko issue-copyright-statement© Springer-Verlag Berlin Heidelberg 2016 ==== Body Whereas MHC genes and the genes for T cell receptors appeared simultaneously during evolution and are present in all extant jawed vertebrates, the genes for CD1 and MR1 have a more limited distribution. Keeping the phylogenetic relationships among animal species in mind, we will describe the CD1 and MR1 genes in extant species. The evolution of these genes is beyond the scope of this review. CD1 and MR1 proteins bind lipids and vitamin B metabolites respectively, and present these to T cells, as opposed to the classical MHC proteins, which present peptide antigens to T cells. CD1 and MR1 genes have not been found in fish, while fish do contain MHC class I and II genes (Dascher 2007). Reptiles (Yang et al. 2015), birds (Miller et al. 2005; Salomonsen et al. 2005), and marsupials (Baker and Miller 2007; Cheng and Belov 2014) have CD1 genes that can clearly be distinguished from classical MHC genes and form an interspecies group with the mammalian CD1 genes. CD1a, CD1b, CD1c, CD1d, and CD1e proteins most likely arose in a common ancestor of placental mammals from a primordial form of CD1. Gene duplications, deletions, and gene inactivations shaped the composition of the CD1 family of genes, possibly under selective pressure associated with an immune function. MR1 genes are absent in fish and reptiles, but in marsupials and mammals, there is typically a single functional MR1 gene. Of all the MHC Class I-like proteins, MR1 shows the highest conservation among species. Chromosomal location and gene structure of CD1 and MR1 In mammals, the CD1 and MR1 genes are not part of the MHC locus. In humans, the CD1 and MR1 genes are located on chromosome 1 and the MHC locus lies on chromosome 6. However, in chicken, the CD1 and MHC loci are linked (Miller et al. 2005; Salomonsen et al. 2005). The genes that encode the classical MHC molecules are highly polymorphic with hundreds to thousands allelic variants, and this is thought to be closely related to their function of presenting peptides to T cells. No allelic variants for CD1b, CD1c, and MR1 are known. Allelic variants in the form of non-synonymous single nucleotide polymorphisms in human CD1a and CD1d are known, but the affected amino acids are not located in the antigen-binding cleft or the TCR-binding surface (Han et al. 1999; Oteo et al. 2001). Together, this justifies the common description of CD1 and MR1 genes as non-polymorphic. Because of its non-polymorphic nature, shared structural features, and the fact that it presents small metabolite antigens rather than peptides, MR1 is often compared to CD1. However, MR1 genes are distinct from CD1 genes, and, based on sequence alignments, form their own interspecies group. Even though in humans MR1 and CD1 are located in the same MHC paralogous region on chromosome 1, the MR1 locus is separate from the CD1 cluster of genes, or CD1 locus (Hashimoto et al. 1995; Shiina et al. 2001). The CD1 locus is located between the KIRREL and olfactory receptor genes and consists of multiple genes, often in both orientations. The MR1 gene is embedded between STX6 and IER5 genes. In mice, a chromosomal rearrangement caused the separation of the CD1 and MR1 genes, which are located on chromosome 3 and chromosome 1, respectively (Dascher and Brenner 2003). In the available assembled primate data, MR1 and CD1 are located on the same chromosome, but in other mammals that were studied, like cow and pig, they are not (Goldfinch et al. 2010; Reinink and Van Rhijn 2016). The human genome encodes one MR1 protein and five different CD1 proteins, called CD1a through CD1e. These five CD1 proteins and their orthologs are called CD1 isoforms. Even though for other proteins this term is used to define RNA splice variants derived from the same gene, in the CD1 field, the word isoforms is used to indicate products of separate genes. The overall structure of CD1 and MR1 proteins resembles MHC class I molecules: a type I transmembrane protein, called the heavy chain, consisting of α1, α2, and α3 domains associated with β2 microglobulin. CD1 and MR1 genes have an intron-exon structure comparable to MHC class I genes: they contain 6 exons that encode 5′ UTR and leader signal peptide, α1 domain, α2 domain, α3 domain, transmembrane domain, and cytoplasmic tail and 3′ UTR combined (Yamaguchi et al. 1998). CD1 genes in mammals are named after the human CD1 isoform they group with, based on sequence comparison. Comparison, based on overall alignment of the full coding sequence or only the α1 and α2 domains that form the antigen binding cleft, gives identical results. This is caused by the fact that α3 domains are highly conserved among all CD1 isoforms within one species (Balk et al. 1989), and cytoplasmic domains, even though they show considerable differences among the isoforms, are very short. Thus, effectively, CD1 isoforms are grouped and named according to resemblance of the sequence encoding their antigen-binding cleft-forming α1 and α2 domains. CD1 in humans and common research, farm, and companion animals For nine mammalian species (human, rabbit, guinea pig, cow, pig, dog, horse, mouse, and rat), the CD1 genes have been studied extensively and, except for rabbit and guinea pig, their CD1 loci have been carefully mapped and curated. The functionality of many of these CD1 genes has been studied by cloning the transcripts from cDNA, sometimes followed by protein expression studies. Genomes from these nine mammalian species contain from one (rat (Katabami et al. 1998)) to thirteen (horse and dog (Dossa et al. 2014; Schjaerff et al. 2016)) CD1 genes. From rabbits, two CD1a, two CD1b, one CD1d, and one CD1e transcripts have been identified, but this study was not set up or intended to define the complete rabbit CD1 locus (Hayes and Knight 2001). Southern blots suggested the presence of at least eight CD1 genes in the rabbit (Calabi et al. 1989). Guinea pigs have been reported to contain four functional genes for CD1b, three for CD1c, one for CD1e gene, and at least five CD1 pseudogenes (Dascher et al. 1999). Later, one functional gene for CD1d was described (Looringh van Beeck et al. 2009), and one gene of unknown functionality for CD1a (Van Rhijn and Moody 2015). The guinea pig CD1 locus has not been mapped, but while the genome is being updated and assembled, even more CD1 genes have been identified in guinea pig (Reinink and Van Rhijn 2016). The first attempt to define the bovine CD1 locus was based on an early draft of the bovine genome and identified one CD1a gene, five CD1b genes, two CD1d genes, and a CD1e gene. Among these, the CD1a gene, three CD1b genes, and the CD1e were identified as functional genes (Van Rhijn et al. 2006). A later version of the genome brought the number of bovine CD1 genes to 12 (Nguyen et al. 2015). The porcine CD1 locus has been described based on BAC sequencing and contains six CD1 genes: two genes for CD1a, and one for each of the other isoforms (Eguchi-Ogawa et al. 2007). One of the CD1a genes is a pseudogene. Twelve CD1 genes were mapped in the canine locus based on the available genome at the time, and a 13th gene was not mapped in the locus and was thought to be an allele (Looringh van Beeck et al. 2008). Subsequently, BAC sequencing placed the 13th gene in the locus, which is now known to contain nine CD1a genes, and one gene or each of the other isoforms (Schjaerff et al. 2016). Four of the canine CD1a genes are thought to be functional. The horse genome contains 18 CD1 genes, among which five pseudogenes (Dossa et al. 2014). The functional equine CD1 genes are seven CD1a genes, two CD1b genes, one CD1c gene, one CD1d gene, and two CD1e genes. The mouse genome, with its small locus consisting of only two CD1d genes, has been shown to have undergone a rearrangement that caused the loss of the other CD1 genes (Bradbury et al. 1988; Dascher and Brenner 2003). Rats have only one CD1 gene, which encodes CD1d (Ichimiya et al. 1994). CD1 loci in less well studied mammalian genomes Only a very small number of CD1 loci of the total of almost 4000 extant species of placental mammals have been studied. For species other than human, rabbit, guinea pig, cow, pig, dog, horse, mouse, and rat, the study of CD1 genes is now facilitated by the availability of multiple genomes. It should be noted however that many genomes, especially from species other than primates and rodents, are not yet finalized and contain multiple gaps. Also, not all sequence materials may have been assigned to a chromosomal location yet, and artifacts like duplications may be present. Therefore, attempts to describe a CD1 locus based on genomic data available at a certain point in time may need to be adjusted later based on improved versions of the genome (Nguyen et al. 2015; Schjaerff et al. 2016). Because transcription and correct splicing of existing genes cannot reliably be predicted, mRNA- and cDNA-based sequences provide the most reliable data on functionality and the possible expression of individual CD1 proteins in vivo. However, RNA sequence-based data are unlikely to provide a complete overview of CD1 proteins in a species because expression of CD1 genes is often limited to a specific tissue and cell type. Therefore, to get an indication of the total number of CD1 genes in a species, searches using the basic local alignment search tool (BLAST) or collection of annotated genes in genomes provide more reliable information than expressed sequence tags or cDNA databases. Table 1 shows numbers of CD1 genes obtained from whole genomes of 15 mammals, including less well studied species like alpaca, dolphin, elephant, two bat species, panda, and sloth (Reinink and Van Rhijn 2016). Of note, BLAST-based searches reveal functional genes and pseudogenes. Therefore, differences between results of de novo BLAST searches, automated annotation of open reading frames, and published data with regard to the total number of CD1 genes per species are expected.Table 1 CD1 and MR1 gene numbers Common name Genome Binomial species name CD1a CD1b CD1c CD1d CD1e Total CD1 MR1 Alpaca vicPac2 Vicugna pacos 1 1 1 1 1 5 1 Bonobo panPan1 Pan paniscus 1 1 1 1 1 5 2 Chimpanzee panTro4 Pan troglodytes 1 1 1 1 1 5 2 Dog CanFam3 Canis lupus 9 1 1 1 1 13 1 Dolphina turTru2 Tursiops truncatus 0 1 0 0 0 1 0 Elephant loxAfr3 Loxodonta africana 1 2 1 1 1 6 1 Horse equCab2 Equus caballus 9 2 2 1 2 16 1 Human hg38 Homo sapiens 1 1 1 1 1 5 2 Megabat pteVam1 Pteropus vampyrus 3 1 1 0 1 6 1 Microbat myoLuc2 Myotis lucifugus 17 2 0 5 2 26 1 Mouse mm10 Mus musculus 0 0 0 2 0 2 1 Panda ailMel1 Ailuropoda melanoleuca 8 1 1 1 1 12 1 Pig susScr3 Sus scrofa 2 1 1 1 2 7 1 Rabbit oryCun2 Oryctolagus cuniculus 5 2 0 1 2 10 0 Rhesus macaque rheMac3 Macaca mulatta 2 1 1 1 1 6 2 Slotha choHof1 Choloepus hoffmanni 1 0 0 1 0 2 1 For each of the indicated mammalian genomes, a list of CD1 and MR1 genes as determined by BLAST-based searches was merged with a list of Ensembl-annotated CD1 and MR1 genes when available (adapted from (Reinink and Van Rhijn 2016)). Redundancies (genes with identical genomic location) were removed aThe dolphin and sloth genomes are not completely assembled and consist of relatively small contigs, which may have led to the fragmentation of CD1 or MR1 genes and subsequent failure of identification CD1 genes have undergone multiple duplications in most mammals, leading to extended multigene families. The general picture that emerges is that CD1a has undergone the most extensive multiplications and CD1e the least or none. Automated annotations of CD1 isoforms of species other than mouse or human often indicate a certain degree of uncertainty, and are sometimes annotated as “CD1a-like,” for example. However, the golden standard for isoform assignment is sequence alignment of the combined α1 and α2 domains. All CD1 genes from 15 mammals that were recovered group with one of the five known CD1 isoforms, but were not always correctly named by automated annotation (Fig. 1). The clear grouping with the five isoforms suggests that the isoform nomenclature that was based on the human CD1 locus appropriately describes all currently known mammalian CD1 isoforms. With more than 10 CD1 genes, the little brown bat, Myotis lucifugus, can be considered a species with high numbers of CD1 genes. Intermediate numbers are found primates, alpaca, and elephant. These data confirm once more that mouse and rat are atypical with the lowest absolute number of CD1 genes (two and one, respectively). Though not identical, the gene numbers obtained by BLAST-based searches and automated annotation are largely consistent (Reinink and Van Rhijn 2016). Because misassemblies of repeated sequences and unmerged overlaps due to polymorphisms resulting in artificial duplications can occur in incompletely assembled genomes, we expect that in some species, the exact numbers of CD1 genes will be adjusted in the future. However, it is clear that humans should be considered to have an intermediate number of CD1 genes, while many other mammals have more extensive families of CD1 genes. Large CD1 families are often dominated by multiplied CD1a genes and to a lesser degree, CD1b, CD1c, and CD1d genes.Fig. 1 CD1 and MR1 genes in mammals. From 16 mammals, known CD1 genes and predicted CD1 paralog open reading frames were obtained from Ensembl (www.ensembl.org). An alignment of these sequences and human MICA, MICB, HLA-A, HLA-B, and murine H2-M3 was generated by MUSCLE (Edgar 2004), clustered according to a neighbor joining algorithm, and shown as a radial cladogram. Groups were color-coded based on the clustering with human CD1 isoforms Mammalian MR1 genes Searches in whole genomes (Table 1), as well as other published data, show that primates have two MR1 genes. In humans and chimpanzees, one of these genes is known to be functional and the other one is a pseudogene (Parra-Cuadrado et al. 2001; Parra-Cuadrado et al. 2000). No MR1 gene could be identified in the rabbit genome, which is in line with recently published data on lagomorphs (rabbit and pika) (Boudinot et al. 2016). The panda and dog MR1 genes that resulted from BLAST-based searches (Table 1) are most likely pseudogenes because it was shown that members of the order of Carnivora, including cats, dogs, ferrets, and pandas, have a MR1 pseudogene and lack a functional MR1 gene (Boudinot et al. 2016). Other mammals (alpaca, elephant, horse, bat) also have one MR1 gene. While in dolphin no MR1 gene was found, this does not conclusively prove the absence of MR1 in this species because the dolphin genome is still incompletely assembled. The sloth MR1 gene has a gap in the current version of the sloth genome (choHof1). Functional specialization of CD1 isoforms Interspecies comparisons between individual CD1 proteins have been made based on sequence alignment, protein models (canine versus human CD1a (Looringh van Beeck et al. 2008)), or crystal structures (bovine versus human CD1b (Girardi et al. 2010), bovine versus human CD1d (Wang et al. 2012), and murine versus human CD1d (Koch et al. 2005)). CD1 proteins within one species differ from each other in ways that suggest functional specialization. One aspect determines the function of a CD1 protein is the size and shape of the antigen-binding cleft, which is formed by the α1 and α2 helices. Among the human CD1 proteins, the biggest size difference is observed between CD1a and CD1b. In addition, there are significant differences in shape with the CD1b cleft consisting of four pockets (Gadola et al. 2002) of which three are interconnected and the CD1a cleft consisting of one cavity (Zajonc et al. 2003). These differences translate into differences in size and shape of antigens that can be bound by the human CD1a and CD1b proteins. Because the CD1 isoforms form interspecies groups based on phylogenetic relationship based on the sequences of their α1 and α2 helices, functional specializations of CD1 molecules that relate to the antigen-binding cleft are generally conserved across species. Especially, the size of the cleft seems to be well conserved when human CD1 isoforms are compared with an ortholog. However, as seen in the bovine CD1d molecule, despite a very high homology between human and bovine CD1d, tryptophan 166 in the bovine protein, which is a cysteine in humans and mice, blocks part of the A’ pocket so that long fatty acyl chains will not fit (Wang et al. 2012). A comparable situation exists when human CD1b and bovine CD1b3 are compared: the tunnel that is present in the human CD1b protein is closed in bovine CD1b3 by valine 98, which prevents the binding of extremely long ligands (Girardi et al. 2010). Structural aspects of these CD1 molecules will be reviewed by Zajonc et al. in this special issue of Immunogenetics. Thus, despite very high sequence homology and overall structural resemblance between orthologs, a single amino acid difference can have profound functional impact. One specific feature of CD1 molecules that is difficult to predict from crude genomic data is the cytoplasmic tail, which determines the transport and subcellular location of individual CD1 isoforms (Moody and Porcelli 2003). The cytoplasmic tail is encoded by the very small exon 5 that is located between two longer introns of variable length and cannot reliably be predicted based on the genomic sequence of the CD1 gene. However, the many cDNA sequences that are available have provided a number of cytoplasmic tails that allows for comparative analysis and predictions on intracellular trafficking of these CD1 proteins (Table 2).Table 2 Cytoplasmic tails of CD1 proteins in mammals Gene name (alias) Cytoplasmic tail Motif Reference cDNA boCD1a2 RKSWSTYMSDA (Nguyen et al. 2015) boCD1a1 WKHWTHRESPSSVLPLE (Van Rhijn et al. 2006) boCD1a canCD1a2 KAHWRPQCMDFPSEREPSSPSSSTYLNPAQH (Schjaerff et al. 2016) canCD1a6 KRWKTHNRPQCTDFPLK (Looringh van Beeck et al. 2008) canCD1a9 KAHWRPQCTDFPSEQEPSSPGSSTYLNPAQH (Looringh van Beeck et al. 2008; Schjaerff et al. 2016) canCD1a8.1 canCD1a8 KRWKAH (Looringh van Beeck et al. 2008; Schjaerff et al. 2016) canCD1a8.2 eqCD1a1 THCEAPCTIVPLK (Dossa et al. 2014) eqCD1a2 IRHQLQRTLLPLD Dileucine (Dossa et al. 2014) eqCD1a3 IHSELPRTLLPLE Dileucine (Dossa et al. 2014) eqCD1a4 VISISVSILVRKPCATPRTPLPSQ (Dossa et al. 2014) eqCD1a5 RSCESASNLLWNEIPGAQDPGHI Dileucine (Dossa et al. 2014) eqCD1a6 WLRKRWTRCEPPSNLISLE (Dossa et al. 2014) eqCD1a7 WLRKRGTHCEFPRTCLPLE (Dossa et al. 2014) huCD1a RKRCFC (Calabi and Milstein 1986; Martin et al. 1987) pigCD1a1 WHRKHWKHCDPSSALHRLE (Chun et al. 1999; Eguchi-Ogawa et al. 2007) pCD1.1 rabCD1a1 RKCWIHHGPLETLLPLQ Dileucine (Hayes and Knight 2001) rabCD1a2 KKRWSHHGSPNSLLPLK Dileucine (Hayes and Knight 2001) boCD1b1 RFMGSHRVGHD (Nguyen et al. 2015; Van Rhijn et al. 2006) boCD1b3 RRWSYQNIL YXXZ, dileucine (Nguyen et al. 2015; Van Rhijn et al. 2006) boCD1b5 RRWSYQTIL YXXZ, dileucine (Nguyen et al. 2015) canCD1b RRWSYQSIS YXXZ (Looringh van Beeck et al. 2008) eqCD1b1 SYQNIS YXXZ (Dossa et al. 2014) eqCD1b2 SYLNIP YXXZ (Dossa et al. 2014) gpCD1b1 RRWSYEDIL YXXZ, dileucine (Dascher et al. 1999; Hiromatsu et al. 2002) gpCD1b2 KHWSYQDIL YXXZ, dileucine (Dascher et al. 1999; Dascher et al. 2002) gpCD1b3 RRLRCEGIF (Dascher et al. 1999; Dascher et al. 2002) gpCD1b4 RRWSYEDIF YXXZ (Dascher et al. 1999) huCD1b RRRSYQNIP YXXZ (Martin et al. 1987) ovCD1b RRWSYQNIL YXXZ, dileucine (Ferguson et al. 1996) scd1b42 ovCD1b RRWSHRNIL Dileucine (Ferguson et al. 1996) scd1b52 ovCD1b RRWSYLTIL YXXZ, dileucine (Ferguson et al. 1996) scd1a25 pigCD1b RRWSYQSVL YXXZ (Eguchi-Ogawa et al. 2007) rabCD1b RRRSYQNIL YXXZ, dileucine (Calabi et al. 1989; Hayes and Knight 2001) canCD1c RKCCSYQDIP YXXZ (Looringh van Beeck et al. 2008) eqCD1c SYQNIQRDSSPCFPHCNENCTAQEHRTTE YXXZ (Dossa et al. 2014) gpCD1c1 KRCTYQGIQ YXXZ (Dascher et al. 1999) gpCD1c2 KRCTYQGIP YXXZ (Dascher et al. 1999) gpCD1c3 KKCCTYQGIPa YXXZ (Dascher et al. 1999) huCD1c KKHCSYQDIL YXXZ, dileucine (Martin et al. 1987) eqCD1d KKRSSYQDIL YXXZ, dileucine (Dossa et al. 2014; Looringh van Beeck et al. 2009) gpcD1d RRGRSYQDIL YXXZ, dileucine (Looringh van Beeck et al. 2009) huCD1d RFKQTSYQGVL YXXZ, dileucine (Balk et al. 1989) lafCD1d KRHCS (Looringh van Beeck et al. 2009) moCD1d1 RRRSAYQDIR YXXZ (Bradbury et al. 1988) moCD1d2 RRRSAYQDIR YXXZ (Bradbury et al. 1988) ovCD1d RKHRRYQDIS YXXZ, dileucine (Rhind et al. 1999) scd1d pigCD1d RRRVYQNIQ YXXZ (Eguchi-Ogawa et al. 2007) rabCD1d RRRCSYQGIL YXXZ, dileucine (Calabi et al. 1989; Looringh van Beeck et al. 2009) ratCD1d RRRSYQDIM YXXZ (Ichimiya et al. 1994) The cytoplasmic tails of CD1 proteins, grouped by isoform. Only tail sequences that have been confirmed by cDNA sequences are included. Species from which the sequences were derived are indicated as bo: bovine; can: canine; eq: equine; gp: guinea pig; hu: human; laf: African elephant; mo: mouse; ov: ovine; rab: rabbit; dileucine: modified dileucine motif aThe sequence shown is based on GenBank sequence NM_001172855, but has been published as KKCCTYQGFP (Dascher et al. 1999) Intracellular trafficking is determined by the CD1 cytoplasmic tail Cellular lipids are bound in the antigen binding cleft during synthesis of the CD1 protein. When the CD1 protein recycles and encounters other lipids in an endosomal compartment, these may replace the initial endoplasmic reticulum-derived lipid. The type of endosomal compartment a lipid travels to is thus a determining factor in the kind of antigenic lipids it presents. The cytoplasmic tails of CD1 molecules may or may not contain short consensus sequences for adaptor proteins. CD1b, CD1c, or CD1d molecules often carry a modified dileucine motif and/or a tyrosine-based YXXZ motif, where X is any amino acid, and Z is a bulky hydrophobic amino acid. YXXZ motifs interact with adaptor proteins and are responsible for recycling from the cell surface to intermediate and late endosomal compartments (Briken et al. 2002; Sugita et al. 2002). CD1e does not appear at the cell surface and does not recycle and has a function in lipid antigen loading and processing (Angenieux et al. 2005; de la Salle et al. 2005). Comparing the cytoplasmic tails of all cell surface-expressed isoforms across species, the following general picture emerges: CD1a tails are highly variable in length. Most CD1a tails contain histidine residues, with unknown function, or cysteines, which can be palmitoylated. The human CD1a cytoplasmic tail contains a cysteine residue that is palmitoylated and involved in incorporation in lipid rafts (Barral et al. 2008; Sloma et al. 2008), but the mechanism of recycling to the sorting or early endosomal compartment, which is so typical for human CD1a, is unknown. None of the 17 known cytoplasmic tails of CD1a proteins has a modified dileucine- and/or a tyrosine-based YXXZ motif, consistent with the absence from intermediate and late endosomes of the human CD1a protein, and strongly suggesting that this is an evolutionary conserved feature (Table 2 ). CD1b, CD1c, and CD1d tails are of comparable length, typically contain a YXXZ motif, and sometimes an additional modified dileucine motif. In addition, all known CD1c cytoplasmic tails contain a cysteine of unknown function. There are exceptions to these general observations, like the bovine CD1b1 (Nguyen et al. 2015), guinea pig CD1b3 (Dascher et al. 1999), and African elephant CD1d (Looringh van Beeck et al. 2009) cytoplasmic tails, which lack any known motif. Indeed, guinea pig CD1b3 does not travel to late endosomes and rather has a CD1a-like subcellular distribution pattern (Dascher et al. 2002). However, the bovine CD1b3 and CD1b5 and guinea pig CD1b1, CD1b2, and CD1b4 have “typical” CD1b cytoplasmic tails. Thus, cows and guinea pigs have typical and atypical cytoplasmic tails among the proteins belonging to the CD1b isoform. For the African elephant, no other cytoplasmic tails have been sequenced from cDNA yet. Generally speaking, it seems that when there are multiple proteins of one CD1 isoform in a species, there is functional diversification, but when a species has just one gene for a certain CD1 isoform, this one gene usually shows the “typical” combination of cleft and tail. Expression patterns of CD1 isoforms Human cortical thymocytes express high levels of CD1a, CD1b, CD1c, and CD1d on their cell surface. Broad CD1 expression in the thymus was confirmed in all other species where CD1 expression was studied (Hiromatsu et al. 2002; Howard et al. 1993; Looringh van Beeck et al. 2008; Van Rhijn et al. 2006). Outside the thymus, the general picture that emerges is that CD1d is widely expressed at a low level, with some specific high CD1d-expressing cells in certain tissues, like hepatic stellate cells in the liver (Winau et al. 2007), and marginal zone B cells in the spleen (Makowska et al. 1999). In humans, CD1b expression is limited to dendritic cells; CD1a is expressed by dendritic cells and at high levels by Langerhans cells in the skin, and CD1c is expressed by dendritic cells and subsets of B cells. Like humans, dogs express CD1a in the skin and thymus (Looringh van Beeck et al. 2008). However, only two of the three studied canine CD1a genes (out of the nine CD1a genes in the dog CD1 locus) were preferentially expressed in the skin. In rabbits, one of the two CD1a genes is preferentially expressed in the skin, while the other one has a more general expression pattern (Hayes and Knight 2001). In guinea pigs, different CD1b genes have different expression patterns in peripheral blood and spleen (Hiromatsu et al. 2002). Mice that are transgenic for the part of the human locus that encodes CD1a, CD1b, CD1c, and CD1e with their endogenous promoters show a CD1 expression pattern that is surprisingly comparable to the human CD1 expression pattern in that all isoforms are expressed on thymocytes and lymph node dendritic cells, while CD1a stands out as highly expressed on Langerhans cells, and CD1c is expressed on B cells (Felio et al. 2009). Closing remarks A typical pattern of gene distribution has been described for immune function-related genes including genes that encode TCRs, immunoglobulins, classical MHC molecules, and NK receptors. This pattern is the result of gene family expansion, diversification, and contraction or “birth and death” evolution, and usually leads to families of genes that include a considerable number of pseudogenes (Nei and Rooney 2005). Furthermore, complete loss of certain family members and expansion of other family members are observed when contemporary animal species are compared. With its many duplications, deletions, and pseudogenes, the CD1 loci seem to follow this pattern. Although MR1 genes have been subject to inactivation in the carnivores, extensive gene family expansion and diversification were not observed in any of the species studied. The lack of functional diversification of MR1 genes is hard to explain but may be related to its seemingly exclusive interaction with the highly specialized MAIT cells. Even though mammals vary widely in the numbers of CD1 genes they have, some general observations can be made and help understand the function of the different CD1 isoforms. The skin seems to be a preferred site for CD1a expression across species. While CD1a genes have undergone extensive multiplication in some species, even more so than CD1b and CD1c, none of the currently described CD1a proteins contains any known signal for endosomal location. Multiplication and diversification of genes may have occurred in response to changes in the environment, including pathogenic and non-pathogenic microbes, to which evolving mammals were exposed. This article is published in the Special Issue CD1, MR1, NKT, and MAIT: Evolution and Origins of Non-peptidic Antigen Recognition by T lymphocytes with Guest Editor Dr. Dirk Zajonc This work was supported by The Netherlands Organisation for Scientific Research (NWO) grant 824.02.002. ==== Refs References Angenieux C Fraisier V Maitre B Racine V van der Wel N Fricker D Proamer F Sachse M Cazenave JP Peters P Goud B Hanau D Sibarita JB Salamero J de la Salle H The cellular pathway of CD1e in immature and maturing dendritic cells Traffic 2005 6 286 302 10.1111/j.1600-0854.2005.00272.x 15752135 Baker ML Miller RD Evolution of mammalian CD1: marsupial CD1 is not orthologous to the eutherian isoforms and is a pseudogene in the opossum Monodelphis domestica Immunology 2007 121 113 121 10.1111/j.1365-2567.2007.02545.x 17244156 Balk SP Bleicher PA Terhorst C Isolation and characterization of a cDNA and gene coding for a fourth CD1 molecule Proc Natl Acad Sci U S A 1989 86 252 256 10.1073/pnas.86.1.252 2463622 Barral DC Cavallari M McCormick PJ Garg S Magee AI Bonifacino JS De Libero G Brenner MB CD1a and MHC class I follow a similar endocytic recycling pathway Traffic 2008 9 1146 1157 10.1111/j.1600-0854.2008.00781.x 18419753 Boudinot P, Mondot S, Jouneau L, Teyton L, Lefranc MP, Lantz O (2016) Restricting nonclassical MHC genes coevolve with TRAV genes used by innate-like T cells in mammals. Proc Natl Acad Sci USA 113:E2983-E2992 Bradbury A Belt KT Neri TM Milstein C Calabi F Mouse CD1 is distinct from and co-exists with TL in the same thymus EMBO J 1988 7 3081 3086 2460336 Briken V Jackman RM Dasgupta S Hoening S Porcelli SA Intracellular trafficking pathway of newly synthesized CD1b molecules EMBO J 2002 21 825 834 10.1093/emboj/21.4.825 11847129 Calabi F Belt KT Yu CY Bradbury A Mandy WJ Milstein C The rabbit CD1 and the evolutionary conservation of the CD1 gene family Immunogenetics 1989 30 370 377 10.1007/BF02425277 2478462 Calabi F Milstein C A novel family of human major histocompatibility complex-related genes not mapping to chromosome 6 Nature 1986 323 540 543 10.1038/323540a0 3093894 Cheng Y Belov K Characterisation of non-classical MHC class I genes in the Tasmanian devil (Sarcophilus harrisii ) Immunogenetics 2014 66 727 735 10.1007/s00251-014-0804-3 25267059 Chun T Wang K Zuckermann FA Gaskins HR Molecular cloning and characterization of a novel CD1 gene from the pig J Immunol 1999 162 6562 6571 10352272 Dascher CC Evolutionary biology of CD1 Curr Top Microbiol Immunol 2007 314 3 26 17593655 Dascher CC Brenner MB Evolutionary constraints on CD1 structure: insights from comparative genomic analysis Trends Immunol 2003 24 412 418 10.1016/S1471-4906(03)00179-0 12909453 Dascher CC Hiromatsu K Naylor JW Brauer PP Brown KA Storey JR Behar SM Kawasaki ES Porcelli SA Brenner MB LeClair KP Conservation of a CD1 multigene family in the guinea pig J Immunol 1999 163 5478 5488 10553074 Dascher CC Hiromatsu K Xiong X Sugita M Buhlmann JE Dodge IL Lee SY Roura-Mir C Watts GF Roy CJ Behar SM Clemens DL Porcelli SA Brenner MB Conservation of CD1 intracellular trafficking patterns between mammalian species J Immunol 2002 169 6951 6958 10.4049/jimmunol.169.12.6951 12471129 de la Salle H Mariotti S Angenieux C Gilleron M Garcia-Alles LF Malm D Berg T Paoletti S Maitre B Mourey L Salamero J Cazenave JP Hanau D Mori L Puzo G De Libero G Assistance of microbial glycolipid antigen processing by CD1e Science 2005 310 1321 1324 10.1126/science.1115301 16311334 Dossa RG Alperin DC Hines MT Hines SA The equine CD1 gene family is the largest and most diverse yet identified Immunogenetics 2014 66 33 42 10.1007/s00251-013-0741-6 24196432 Edgar RC MUSCLE: multiple sequence alignment with high accuracy and high throughput Nucleic Acids Res 2004 32 1792 1797 10.1093/nar/gkh340 15034147 Eguchi-Ogawa T Morozumi T Tanaka M Shinkai H Okumura N Suzuki K Awata T Uenishi H Analysis of the genomic structure of the porcine CD1 gene cluster Genomics 2007 89 248 261 10.1016/j.ygeno.2006.10.003 17112699 Felio K Nguyen H Dascher CC Choi HJ Li S Zimmer MI Colmone A Moody DB Brenner MB Wang CR CD1-restricted adaptive immune responses to mycobacteria in human group 1 CD1 transgenic mice J Exp Med 2009 206 2497 2509 10.1084/jem.20090898 19808251 Ferguson ED Dutia BM Hein WR Hopkins J The sheep CD1 gene family contains at least four CD1B homologues Immunogenetics 1996 44 86 96 8662069 Gadola SD Zaccai NR Harlos K Shepherd D Castro-Palomino JC Ritter G Schmidt RR Jones EY Cerundolo V Structure of human CD1b with bound ligands at 2.3A, a maze for alkyl chains Nat Immunol 2002 3 721 726 10.1038/nrm947 12118248 Girardi E Wang J Mac TT Versluis C Bhowruth V Besra G Heck AJ Van Rhijn I Zajonc DM Crystal structure of bovine CD1b3 with endogenously bound ligands J Immunol 2010 185 376 386 10.4049/jimmunol.1000042 20519644 Goldfinch N Reinink P Connelley T Koets A Morrison I Van Rhijn I Conservation of mucosal associated invariant T (MAIT) cells and the MR1 restriction element in ruminants, and abundance of MAIT cells in spleen Vet Res 2010 41 62 10.1051/vetres/2010034 20507818 Han M Hannick LI DiBrino M Robinson MA Polymorphism of human CD1 genes Tissue Antigens 1999 54 122 127 10.1034/j.1399-0039.1999.540202.x 10488738 Hashimoto K Hirai M Kurosawa Y A gene outside the human MHC related to classical HLA class I genes Science 1995 269 693 695 10.1126/science.7624800 7624800 Hayes SM Knight KL Group 1 CD1 genes in rabbit J Immunol 2001 166 403 410 10.4049/jimmunol.166.1.403 11123318 Hiromatsu K Dascher CC Sugita M Gingrich-Baker C Behar SM LeClair KP Brenner MB Porcelli SA Characterization of guinea-pig group 1 CD1 proteins Immunology 2002 106 159 172 10.1046/j.1365-2567.2002.01422.x 12047745 Howard CJ Sopp P Bembridge G Young J Parsons KR Comparison of CD1 monoclonal antibodies on bovine cells and tissues Vet Immunol Immunopathol 1993 39 77 83 10.1016/0165-2427(93)90166-2 8310660 Ichimiya S Kikuchi K Matsuura A Structural analysis of the rat homologue of CD1. Evidence for evolutionary conservation of the CD1D class and widespread transcription by rat cells J Immunol 1994 153 1112 1123 7517972 Katabami S Matsuura A Chen HZ Imai K Kikuchi K Structural organization of rat CD1 typifies evolutionarily conserved CD1D class genes Immunogenetics 1998 48 22 31 10.1007/s002510050396 9601940 Koch M Stronge VS Shepherd D Gadola SD Mathew B Ritter G Fersht AR Besra GS Schmidt RR Jones EY Cerundolo V The crystal structure of human CD1d with and without alpha-galactosylceramide Nat Immunol 2005 6 819 826 10.1038/ni1225 16007090 Looringh van Beeck FA Reinink P Hermsen R Zajonc DM Laven MJ Fun A Troskie M Schoemaker NJ Morar D Lenstra JA Vervelde L Rutten VP van Eden W Van Rhijn I Functional CD1d and/or NKT cell invariant chain transcript in horse, pig, African elephant and guinea pig, but not in ruminants Mol Immunol 2009 46 1424 1431 10.1016/j.molimm.2008.12.009 19185921 Looringh van Beeck FA Zajonc DM Moore PF Schlotter YM Broere F Rutten VP Willemse T Van Rhijn I Two canine CD1a proteins are differentially expressed in skin Immunogenetics 2008 60 315 324 10.1007/s00251-008-0297-z 18488214 Makowska A Faizunnessa NN Anderson P Midtvedt T Cardell S CD1high B cells: a population of mixed origin Eur J Immunol 1999 29 3285 3294 10.1002/(SICI)1521-4141(199910)29:10<3285::AID-IMMU3285>3.0.CO;2-P 10540340 Martin LH Calabi F Lefebvre FA Bilsland CA Milstein C Structure and expression of the human thymocyte antigens CD1a, CD1b, and CD1c Proc Natl Acad Sci U S A 1987 84 9189 9193 10.1073/pnas.84.24.9189 2447586 Miller MM Wang C Parisini E Coletta RD Goto RM Lee SY Barral DC Townes M Roura-Mir C Ford HL Brenner MB Dascher CC Characterization of two avian MHC-like genes reveals an ancient origin of the CD1 family Proc Natl Acad Sci U S A 2005 102 8674 8679 10.1073/pnas.0500105102 15939884 Moody DB Porcelli SA Intracellular pathways of CD1 antigen presentation Nat Rev Immunol 2003 3 11 22 10.1038/nri979 12511872 Nei M Rooney AP Concerted and birth-and-death evolution of multigene families Annu Rev Genet 2005 39 121 152 10.1146/annurev.genet.39.073003.112240 16285855 Nguyen TK Reinink P El Messlaki C Im JS Ercan A Porcelli SA Van Rhijn I Expression patterns of bovine CD1 in vivo and assessment of the specificities of the anti-bovine CD1 antibodies PLoS One 2015 10 e0121923 10.1371/journal.pone.0121923 25815476 Oteo M Arribas P Setien F Parra JF Mirones I Gomez del Moral M Martinez-Naves E Structural characterization of two CD1A allelic variants Hum Immunol 2001 62 1137 1141 10.1016/S0198-8859(01)00314-7 11600221 Parra-Cuadrado JF Gomez del Moral M Garcia-Pavia P Setien F Martinez-Naves E Characterization of the MHC class I-related MR1 locus in nonhuman primates Immunogenetics 2001 53 643 648 10.1007/s00251-001-0380-1 11797097 Parra-Cuadrado JF Navarro P Mirones I Setien F Oteo M Martinez-Naves E A study on the polymorphism of human MHC class I-related MR1 gene and identification of an MR1-like pseudogene Tissue Antigens 2000 56 170 172 10.1034/j.1399-0039.2000.560211.x 11019920 Reinink P Van Rhijn I In silico method for identification of MHC class I-like molecules in whole genomes BioRxiv 2016 Rhind SM Hopkins J Dutia BM Amino-terminal sequencing of sheep CD1 antigens and identification of a sheep CD1D gene Immunogenetics 1999 49 225 230 10.1007/s002510050483 9914336 Salomonsen J Sorensen MR Marston DA Rogers SL Collen T van Hateren A Smith AL Beal RK Skjodt K Kaufman J Two CD1 genes map to the chicken MHC, indicating that CD1 genes are ancient and likely to have been present in the primordial MHC Proc Natl Acad Sci U S A 2005 102 8668 8673 10.1073/pnas.0409213102 15939887 Schjaerff M Keller SM Fass J Froenicke L Grahn RA Lyons L Affolter VK Kristensen AT Moore PF Refinement of the canine CD1 locus topology and investigation of antibody binding to recombinant canine CD1 isoforms Immunogenetics 2016 68 191 204 10.1007/s00251-015-0889-3 26687789 Shiina T Ando A Suto Y Kasai F Shigenari A Takishima N Kikkawa E Iwata K Kuwano Y Kitamura Y Matsuzawa Y Sano K Nogami M Kawata H Li S Fukuzumi Y Yamazaki M Tashiro H Tamiya G Kohda A Okumura K Ikemura T Soeda E Mizuki N Kimura M Bahram S Inoko H Genomic anatomy of a premier major histocompatibility complex paralogous region on chromosome 1q21-q22 Genome Res 2001 11 789 802 10.1101/gr.175801 11337475 Sloma I Zilber MT Vasselon T Setterblad N Cavallari M Mori L De Libero G Charron D Mooney N Gelin C Regulation of CD1a surface expression and antigen presentation by invariant chain and lipid rafts J Immunol 2008 180 980 987 10.4049/jimmunol.180.2.980 18178838 Sugita M Cao X Watts GF Rogers RA Bonifacino JS Brenner MB Failure of trafficking and antigen presentation by CD1 in AP-3-deficient cells Immunity 2002 16 697 706 10.1016/S1074-7613(02)00311-4 12049721 Van Rhijn I Koets AP Im JS Piebes D Reddington F Besra GS Porcelli SA van Eden W Rutten VP The bovine CD1 family contains group 1 CD1 proteins, but no functional CD1d J Immunol 2006 176 4888 4893 10.4049/jimmunol.176.8.4888 16585584 Van Rhijn I Moody DB CD1 and mycobacterial lipids activate human T cells Immunol Rev 2015 264 138 153 10.1111/imr.12253 25703557 Wang J Guillaume J Pauwels N Van Calenbergh S Van Rhijn I Zajonc DM Crystal structures of bovine CD1d reveal altered aGalCer presentation and a restricted A’ pocket unable to bind very long-chain lipids PLoS One 2012 7 1 10 Winau F Hegasy G Weiskirchen R Weber S Cassan C Sieling PA Modlin RL Liblau RS Gressner AM Kaufmann SH Ito cells are liver-resident antigen-presenting cells for activating T cell responses Immunity 2007 26 117 129 10.1016/j.immuni.2006.11.011 17239632 Yamaguchi H Kurosawa Y Hashimoto K Expanded genomic organization of conserved mammalian MHC class I-related genes, human MR1 and its murine ortholog Biochem Biophys Res Commun 1998 250 558 564 10.1006/bbrc.1998.9353 9784382 Yang Z Wang C Wang T Bai J Zhao Y Liu X Ma Q Wu X Guo Y Ren L Analysis of the reptile CD1 genes: evolutionary implications Immunogenetics 2015 67 337 346 10.1007/s00251-015-0837-2 25921705 Zajonc DM Elsliger MA Teyton L Wilson IA Crystal structure of CD1a in complex with a sulfatide self antigen at a resolution of 2.15 A Nat Immunol 2003 4 808 815 10.1038/ni948 12833155

PMC005xxxxxx/PMC5002278.txt

==== Front Virus GenesVirus GenesVirus Genes0920-85691572-994XSpringer US New York 134310.1007/s11262-016-1343-9ArticleComplete genome sequence of a sapovirus from a child in Zhejiang, China Zhou Xiaohong 1Sun Yi 2Shang Xiaochun 1Gao Jian 2Zhao Xueqin 1Shuai Huiqun 1Zhang Rui 1Zhang Yanjun +86-0571-87115198yjzhang@cdc.zj.cn 21 Xiacheng District Center for Disease Control and Prevention, Hangzhou, Zhejiang China 2 Zhejiang Provincial Center for Disease Control and Prevention, 3399 Binsheng Road, Hangzhou, 310051 Zhejiang China Edited by Hartmut Hengel. 28 4 2016 28 4 2016 2016 52 5 706 710 17 1 2016 16 4 2016 © The Author(s) 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.Although Sapovirus (Caliciviridae) has been accepted as one of the causes of acute gastroenteritis worldwide, little is known about the genetic characteristics of the whole genome of sapoviruses in China, especially those that infect humans. Here we report the complete genome sequence of a sapovirus strain, Human/Zhejiang1/2015/China, obtained from a child with acute gastroenteritis in Hangzhou, Zhejiang Province, China. Samples were collected and delivered to the CDC laboratories and were detected by RT-PCR. Sanger sequencing was used to obtain the full genome and molecular characterization of the genome was determined. A phylogenetic analysis of the genome was also performed. The results indicated that Human/Zhejiang1/2015/China belongs to Genogroup I. No recombination events were detected. This is the first complete sequence from a child to be reported in China. The sequence information is important for surveillance of this emerging gastrointestinal infection. Electronic supplementary material The online version of this article (doi:10.1007/s11262-016-1343-9) contains supplementary material, which is available to authorized users. Keywords SapovirusGenomePhylogenyMajor Science and Technology Special Projects of Zhejiang, China 2013C03045-1Zhang Yanjun Zhejiang Provincial Program for the Cultivation of High-level Innovative Health TalentsMonitor Technology Platform of Infectious Diseases of the State Major Science, Technology Special Projects during the 12th five-year plan of China2012ZX10004-210Zhang Yanjun Key Disciplinary of Health and family planning commission of Zhejiang ProvinceCX-9Zhang Yanjun Provincial Medical Research Fund of Zhejiang2015RCB010Sun Yi issue-copyright-statement© Springer Science+Business Media New York 2016 ==== Body Introduction Sapovirus (SAV; in the family Caliciviridae), causes acute gastroenteritis in both humans and animals. Major clinical symptoms include diarrhea and vomiting with constitutional symptoms such as nausea, chills, and myalgia. Compared to Norovirus, SAV infections are less common and are known to cause disease primarily in children, usually those under the age of 5 years [1]. SAV has received wide attention in public health studies because SAV has recently been a causative agent of gastroenteritis in people of all ages in both confined outbreaks and in sporadic cases [1, 2]. It has also been reported in outbreaks in long-term care facilities for the elderly [3]. Mortality is rare and the symptoms are generally mild. No licensed vaccines or antivirals are available for SAV infections. Many outbreaks caused by foodborne transmission of SAV have been reported since 1976 in England, where SAV particles were first detected in human diarrheic stool samples [4]. The largest foodborne SAV outbreak recorded, which involved 665 people, was in Japan in 2010 [5]. Many worldwide studies have reported that SAV can be divided into several genogroups, ranging from GI to GV [6–8]. Recent studies found that SAV genogroup VI in isolated fecal samples from diarrheic pigs [9]. Different genogroups result in the infection of different hosts; GI, GII, GIV, and GV infect humans and other primates, whereas GIII infects swine. Furthermore, there are at least eight genotypes of GI, five of GII, and one each of GIII, GIV, and GV. Most SAV detected in humans belong to GI and GII [10, 11]. Although SAV has been accepted as one of the causes of acute gastroenteritis worldwide, little is known about the genetic characteristics of sapoviruses in China based on whole genome analysis, especially about genotypes infecting humans, due to infrequent detection and difficulties in culturing and sequencing, especially whole genome sequencing. Here we report the complete genome sequence of SAV strain Human/Zhejiang1/2015/China obtained in late 2014 from a child with acute gastroenteritis in Hangzhou, Zhejiang Province, China. Phylogenetic analysis was performed for comparison with other genogroups/genotypes reported previously. The purpose of this study was to characterize and present the first full-length genome of SAV as fundamental information in mainland China. Materials and methods Samples were collected from a patient who had acute gastroenteritis in a kindergarten in the Xiacheng district of Hangzhou on December 25 and 26, 2014. Samples were delivered to the CDC laboratories and were made into 10 % (w/v) in phosphate-buffered saline suspensions (PBS) and centrifuged at 2000×g for 10 min. Viral RNA was extracted from 200 µl of the supernatant using a Qiagen RNeasy Mini kit (Qiagen) according to the manufacturer’s instructions. Eluted RNA was stored at −80 °C until use. RNA was detected by reverse transcription PCR with SAV universal primers (SV-F13, SV-F14, SV-R13, and SV-R14) using the following thermal cycling profile: 42 °C for 15 min; 95 °C for 3 min; 35 cycles of 95 °C for 30 s, 48 °C for 30 s, and 74 °C for 45 s according to the previous study of Okada et al. [14]. Full genome Sanger sequencing was performed according to the study by Liu et al. [15]. Geneious 4.8.3 (www.geneious.com) was used to assemble and check the sequences of this SAV, Human/Zhejiang1/2015/China. For the VP1 region and the whole genome, we performed multiple alignments with data matrixes of other sequences downloaded from GenBank for each segment, respectively. We blasted the homology sequences in NCBI for both the VP1 region and the genome, then checked for variant positions in amino acid sequences using MEGA 6.0 (www.mega.software.informer.com) for the whole genomic range. Dataset-specific models that were selected using the Akaike Information Criterion in Modeltest 3.7 were analyzed for each matrix [16]. Maximum likelihood (ML) analysis was processed in RAxML v7.2.8 using the VP1 region and genome, respectively (http://sco.h-its.org/exelixis/software.html). Optimal ML trees and bootstrap percentages (BP) were estimated in the same run. The ML BP values were obtained from 1000 BP replicates using the rapid BP algorithm. Both Simplot v3.5.1 and RDP 4.26 analysis with default set were used for detecting recombination event in this strain [17, 18]. Results and discussion This has been the first genome obtained from a child in China, in contrast to 47 complete human/porcine sapovirus sequences available in GenBank till the data published. The full genome sequence of the SAV from the patient was deposited in GenBank with accession number KT327081. The SAV obtained was a positive-sense, single-stranded RNA 7441 bp in size with a 3′-end poly (A) tail. Its 5′-UTR was 9 bp long; 3′-UTR was 94 bp long. The genome contained three open reading frames (ORFs). ORF1 encodes the nonstructural proteins (NS) from 10 to 6853 bp. ORF2 is a minor structural protein VP2 (from 6848 to 7347 bp). ORF3 is a hypothetical protein CDS from 5176 to 5662 bp, which has not been clearly defined [12]. The major capsid protein VP1, from 5164 to 6853 bp, was embraced in ORF1, which followed the seven nonstructural proteins from NS1 to NS7 in the 5′ to 3′ direction (supplementary materials). The NS pattern of ORF1 in our sequence was similar to the GI.1 Manchester strain (GenBank accession: X86560) except for one putative cleavage site between NS5 (VPg) and NS6–NS7 (Protease-RdRp), which was E/A rather than E/G, as mentioned in Oka’s review [13]. Typical amino acid motifs were also detected in our sequence, for example: GAPGIGKT from 480 to 487 amino acid (aa) in NS3, KGKTK, and DDEYDE from 941 to 945 aa and 962–967 aa in NS5, respectively. There were three PPG motifs in the VP1 region, which differed from the reference stain of GI.1 Manchester. The conserved GWS motif in the VP1 region was also detected in our sequence, which proved strictly characteristic of caliciviruses [13]. Chanthaburi-74/2004/Thailand (GenBank accession: AY646854) showed the highest query cover and percent nucleotide identity with our sequence based on the whole genome. We therefore compared amino acid substitutions in the coding regions (mainly including ORF1 and ORF2) in the two sequences (Table 1). Twenty-two variants were found in the nonstructural region while one and two substitutions were discovered in the VP1 and VP2 areas, respectively.Table 1 Unique amino acid changes observed in Human/Zhejiang1/2015/China compared to Chanthaburi-74/2004/Thailand (AY646854) Nonstructural proteins VP1 VP2 Polypeptide position 2 20 45 49 63 111 134 141 181 658 680 681 692 1082 1094 1179 1342 1388 1402 1484 1537 1590 1729 139 146 Chanthaburi-74/2004/Thailand A I F M E R T N V A R P A I V S Q K V V N D P P S Human/Zhejiang1/2015/China V V Y V K K A P G T K S T V I N H T I I S E S S L To better classify the sapovirus, we searched the NCBI database and found Hu/GI.1/8743/Maizuru/2008/JPN (HM030922) to be a close match in the VP1 region, while Chanthaburi-74/2004/Thailand (AY646854) was close to the whole genome. In these two maximum likelihood trees with Bootstrap percentages assigned, different genogroups, such as GI, GII, GIII, GIV, GV, and GVI, were reciprocally monophyletic with clades receiving 100 % BP support, with the exception of the GV clade in the genomic ML tree which had a BP value of 96 % (Fig. 1a, b). Our sequence, Human/Zhejiang1/2015/China, was in the GI clade and shared a relative high homology with Chanthaburi-74/2004/Thailand in both trees. The results of Blast from NCBI indicated sapovirus Human/Zhejiang1/2015/China to be in genogroup I/genotype 1 (G1.1). The phylogenetic structure in both ML trees was basically similar, although there was an inconsistency in that either GIV or GV could be the closest sister group of GI according to the VP1 region and the whole genome (Fig. 1a, b). Several possible scenarios may have led to this phenomenon, such as incomplete lineage sorting between genes and genome. We are further analyzing this hypothesis. This study, focused on the interior of the GI clade, indicated that Human/Zhejiang1/2015/China is not a recombinant, since both trees showed the strain we obtained to be clustered in a similar GI clade. This finding was subsequently substantiated by RDP analysis and Simplot (supplementary materials). Further studies are underway with additional viruses in the family Caliciviridae, including norovirus, based on surveillance programs of foodborne infections established by the national health and family planning commission of China.Fig. 1 Maximum likelihood trees for sapovirus based on VP1 region and whole genome, respectively. BP values based on 1000 replicates are indicated above branches. Solid circles in each tree indicate sequences obtained in our study. Triangles at nodes indicate monophyletic clades of different genogroups. a VP1 tree, b genome tree Conclusions This first genome sequence of SAV from a child in Zhejiang, China, provides genetic characteristics of SAV infections in humans in the eastern China region. Based on complete capsid gene sequences from the phylogenetic results we determined that this SAV, Human/Zhejiang1/2015/China, belongs to GI, which is consistent with previous reports that GI is known to infect humans. Recombination was not detected. Our SAV sequence may help in both surveillance and in understanding the characterization of emerging gastroenteritis infections in China. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary material 1 (TIFF 711 kb) Supplementary material 2 (TIFF 2352 kb) Supplementary material 3 (DOC 82 kb) Xiaohong Zhou, Yi Sun, and Xiaochun Shang contributed equally. This work was supported by the Major Science and Technology Special Projects of Zhejiang, China (2013C03045-1), Provincial Medical Research Fund of Zhejiang (2015RCB010), Zhejiang Provincial Program for the Cultivation of High-level Innovative Health Talents and Key Disciplinary of Health and Family Planning Commission of Zhejiang Province (CX-9). Dr. David E Boufford is gratefully acknowledged for editing the manuscript. Author contributions XHZ and YJZ conceived and designed the study. XCS and JG carried out the sequencing. YS analyzed the data and drafted the manuscript. XQS and RZ collected the sample and provided the reagents. All authors read and approved the final manuscript. Compliance with ethical standards Conflict of interest The authors declare that they have no competing interests. Ethics statement This study was approved by the ethics committee of the Zhejiang Provincial Center for Disease Control and Prevention (ZJCDC), China. This study was done in accordance with the Declaration of Helsinki, Good Clinical Practice, and Chinese regulatory requirements. ==== Refs References 1. Svraka S Vennema H van der Veer B Hedlund KO Thorhagen M Siebenga J Duizer E Koopmans M J. Clin. Virol. 2010 48 2191 2198 2. Oka T Wang Q Katayama K Saif LJ Clin. Microbiol. Rev. 2015 28 32 53 10.1128/CMR.00011-14 25567221 3. Lee LE Cebelinski EA Fuller C Keene WE Smith K Vinje J Besser JM Emerg. Infect. Dis. 2012 18 873 876 10.3201/eid1805.111843 22516204 4. Madeley CR Cosgrove BP Lancet 1976 307 199 200 10.1016/S0140-6736(76)91309-X 54714 5. Kobayashi S Fujiwara N Yasui Y Yamashita T Hiramatsu R Minagawa H Arch. Virol. 2012 157 1995 1997 10.1007/s00705-012-1394-8 22752792 6. Farkas T Zhong WM Jing Y Huang PW Espinosa SM Martinez N Morrow AL Ruiz-Palacios GM Pickering LK Jiang X Arch. Virol. 2004 149 1309 1323 10.1007/s00705-004-0296-9 15221533 7. Hansman GS Oka T Takeda N Virus Res. 2008 137 261 265 10.1016/j.virusres.2008.07.002 18674574 8. Dufkova L Scigalkova I Moutelikova R Malenovska H Prodelalova J Arch. Virol. 2013 158 549 558 10.1007/s00705-012-1528-z 23124843 9. Oka T Saif LF Yang ZZ Scheuer KA Stoltzfus GT Wang QH Genome Announc. 2014 2 e0027-14 10. Akihara S Phan TG Nguyen TA Yagyu F Okitsu S Muller WE Ushijima H J. Med. Virol. 2005 77 595 601 10.1002/jmv.20497 16254963 11. Oka T Mori K Iritani N Harada S Ueki Y Iizuka S Mise K Murakami K Wakita T Katayama K Arch. Virol. 2012 157 349 352 10.1007/s00705-011-1161-2 22075918 12. Hansman GS Takeda N Oka T Oseto M Hedund KO Katayama K Emerg. Infect. Dis. 2005 11 1916 1920 16485479 13. Chen R Neill JD Noel JS Hutson AM Glass RI Estes MK Prasad BV J. Virol. 2004 78 6469 6479 10.1128/JVI.78.12.6469-6479.2004 15163740 14. Okada M Yamashita Y Oseto M Shinozaki K Arch. Virol. 2006 151 2503 2509 10.1007/s00705-006-0820-1 16847552 15. Liu X Yamamoto D Saito M Imagawa T J. Clin. Virol. 2015 68 83 88 10.1016/j.jcv.2015.05.001 26071343 16. Posada D Crandall KA Bioinformatics 1998 14 817 818 10.1093/bioinformatics/14.9.817 9918953 17. Smith M J. Mol. Evol. 1992 34 126 129 1556748 18. Martin D Rybicki E Bioinformatics 2000 16 562 563 10.1093/bioinformatics/16.6.562 10980155

PMC005xxxxxx/PMC5002279.txt

==== Front Virus Genes Virus Genes Virus Genes 0920-8569 1572-994X Springer US New York 27314269 1361 10.1007/s11262-016-1361-7 Article First report of human salivirus/klassevirus in respiratory specimens of a child with fatal adenovirus infection Pei Na peina@genomics.cn 13 Zhang Jiaosheng wgpzjs@sina.com 2 Ma Jinmin majinmin@genomics.cn 1 Li Liqiang liliqiang@genomics.cn 1 Li Meng xueshenglimeng@163.com 14 Li Jiandong lijiandong@genomics.cn 1 Sun Yisuo sunyisuo@genomics.cn 1 Ji Jingkai jijingkai@genomics.cn 1 Jiang Hui jianghui@genomics.cn 1 Hou Yong houyong@genomics.cn 1 Xu Fengping xufengping@genomics.cn 1 Lu Haorong luhaorong@genomics.cn 1 Zhang Ruimu 734561445@qq.com 25 Wei Xuemei donnawei2007@hotmail.com 2 Xu Xun xuxun@genomics.cn 1 Deng Jikui jikui.deng@yahoo.com 2 1 grid.21155.320000000120341839BGI-Shenzhen, Shenzhen, 518083 China 2 grid.452787.b0000000418065224Shenzhen Children’s Hospital, Shenzhen, 518038 China 3 grid.21155.320000000120341839Shenzhen Key Laboratory of Transomics Biotechnologies, BGI-Shenzhen, Shenzhen, 518083 China 4 grid.410645.20000000104550905Qingdao University, Qingdao, 266071 Shandong China 5 grid.263451.7000000009927110XShantou University, Shantou, 515000 China Edited by Paul Schnitzler. 17 6 2016 17 6 2016 2016 52 5 620 624 6 4 2016 1 6 2016 © The Author(s) 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.Adenovirus is a leading cause of respiratory infection in children. Salivirus/klassevirus was first identified as an etiologic agent of gastroenteritis and was never reported in respiratory infection cases. The case being discussed here caught our attention because, although it is a common respiratory infection, it was fatal, while similar cases were mild. In order to find potential causes in the fatal case, we describe the clinical diagnosis and treatment, the sequencing analysis of the salivirus/klassevirus, and the co-infectious adenovirus. Metagenomics sequencing was conducted on the samples from a nasopharyngeal swab of the children with adenovirus infection. Sequences were assembled using IDBA-ud (1.1.1); phylogenetic analysis was performed using MEGA 5.2. RT-PCR and quantitative PCR were performed to verify the existence of the virus in the samples. A nearly full genome of this new virus strain was obtained with 7633 nt encoding a polyprotein of 2331 aa. Meanwhile, it was detected specifically in the nasopharyngeal swab by RT-PCR. Further, homology analysis indicated that the virus has a closer relationship with Salivirus A strain in Shanghai (GU245894). Our study reports the first case of Human salivirus/klassevirus in respiratory specimens of a child with fatal adenovirus infection in Shenzhen, China. The finding and investigation of the virus will provide more useful information for the clinical diagnosis of unexplained lethal infection and expand our knowledge of the new family, salivirus/klassevirus in picornavirus. Electronic supplementary material The online version of this article (doi:10.1007/s11262-016-1361-7) contains supplementary material, which is available to authorized users. Keywords Salivirus/klassevirusPicornavirusSequencingRespiratory infectionAdenovirusissue-copyright-statement© Springer Science+Business Media New York 2016 ==== Body Introduction Respiratory infection is a tremendous burden on humans. More specifically, it is a primary cause of death in children under 5 years old. Viral infections are the primary cause behind respiratory infections in children after bacterial infections. The most common respiratory viruses include respiratory syncytial virus (RSV), influenza virus, and adenovirus. Picornaviruses are non-enveloped viruses with a single-stranded positive-sense RNA genome that encodes a single polyprotein [1]. It consists of 29 genera, six of which were identified as potentially infectious to humans (Enterovirus, Hepatovirus, Parechovirus, Kobuvirus, Cosavirus, and Cardiovirus). Salivirus/klassevirus belong to the genus Salivirus, which is associated with human gastroenteritis [2]. Next-generation sequencing hastened the discovery of the new virus [3]. A sequence-based method [4] was used to dig the potential pathogen, which decreased the time and labor involved. More importantly, however, this study provides useful information for clinical diagnosis in cases of unexplained lethal infection [5]. In our analysis, we describe the first known case of salivirus/klassevirus in the respiratory specimens of a child infected with adenovirus and present the sequencing analysis of the salivirus/klassevirus and the co-infectious adenovirus, with PCR verification and quantification of the salivirus/klassevirus. Methods Sample source and verification One nasopharyngeal swab and a blood sample were collected from the fatal case. Additional nasopharyngeal swabs from another non-fatal case of adenovirus pneumonia were collected for parallel analysis. All clinical diagnosis and treatment were processed in accordance with the hospital standards. Total RNA was extracted from the samples using Qiagen RNeasykit (Qiagen, Inc., Germany) following the standard library construction process required for the next sequencing platform, Ion Proton (Life Inc., New York, USA) for metagenomics sequencing. RT-PCR was performed to verify the existence of the virus in both the nasopharyngeal and the blood samples of the fatal case. A quantitative PCR approach was used on the nasopharyngeal swab. The cDNA swab product of the first PCR with gel extraction was gradient diluted from 10−1, 10−2, to 10−7 to establish a standard (Fig. 1), and then real-time PCR was performed.Fig. 1 Relationship between tenfold serially diluted DNA and C t value. A linear range was observed for DNA concentrations from 0.075 ng/ul to 0.75 fg/ul Data analysis For genome assembly, we used the IDBA-ud (1.1.1) to assemble the data and then filtered the redundant sequences. Contigs were blasted against the reference genome and co-linearity analysis was made. Fourteen contigs were finally obtained, and specific primers were designed to fill the gaps to obtain the complete genome of Salivirus. Homology and Phylogenetic analysis was also performed between the partial gene of the new genome and the reference based on amino acid similarity. For the adenovirus sequences in the fatal case and the control sample, we used an in-house, reference-based method. First, non-human reads were mapped to the adenoviridae reference database. These were constructed first to obtain the adeno-associated reads and to select a candidate reference genome. Second, the candidate reference genome was reduced with the threshold of sequence identity greater than 90 %. All the adeno-associated reads were mapped to the reduced reference genome. Third, we calculated entropy based on the mapping result. Mapped reference positions with sequence depths greater than 0.5 were used as the consensus base. Unknown bases were noted as “N.” The reduced reference, which had the longest length and coverage rate, was the final sequence. Results Clinical diagnosis and treatment In July 2013, a previously healthy 7-month-old boy was admitted to Shenzhen Children’s Hospital because of a persistent cough of 3 days duration, with wheezing and a fever of 40.2 °C (104.36 °F). On the day of admission, he had mild cyanosis around his mouth, shortness of breath, and fine crackles and wheezing rale over both lungs could be heard. The white blood cell (WBC) count was 24.1 × 109/L; neutrophilic granulocyte count was 18.75 × 109/L, and C-reactive protein level was 51.9 mg/L. A chest radiograph showed an exudative lesion over the right lung. Bacterial cultures from blood and deep tracheal aspirate were negative. A virus test was performed using direct immunofluorescence antigen detection. Adenovirus was positive in both the nasopharyngeal swab, and a bronchoalveolar lavage fluid (BALF) procedure was performed to detect any other respiratory tract viruses present, including influenza A, influenza B, parainfluenza I, parainfluenza II, parainfluenza III, and RSV; the results were negative. Results were also negative for the following infectious disease testing: HIV antibody/antigen screen, syphilis enzyme immunoassay, serum IgM for mycoplasma pneumonia, nasopharyngeal swab and BALF PCRs for mycoplasma pneumonia and tubercle bacillus (TB). The child received aerosolized albuterol, supplemental oxygen, and an IV of vancomycin with cefoperazone/sulbactam without improvement. On day 5 of admission, he was transferred to pediatric intensive care unit (PICU) due to persistent anoxemia and was mechanically ventilated. In PICU, he deteriorated as his pneumonia continued until he died on day 9 of acute respiratory distresssyndrome (ARDS). Genome analysis and PCR verification Metagenomics scanning of the data was performed first to analyze the virus composition of the samples. In addition to the positive adenovirus in the samples, reads annotated as salivirus/klassevirus were found in the fatal case. No salivirus/klassevirus reads were identified in the control case. Fourteen contigs were finally obtained and covered 82.4 % (6581 bp) of the strain 02394-01 genome (GQ184145, 7989 bp). The complete genome of the virus was then determined by using five sets of specific primers (Table 1) designed on the contigs, and sanger sequencing was used to fill the gaps. The nearly full genome of this virus strain was 7633 nt encoding a polyprotein of 2331 aa. The genome of the salivirus/klassevirus has been submitted to GenBank under accession no. KT182636 (Supplement 3).Table 1 Primer used for getting the full genome and quantitative PCR Name Sequence Length (bp) Klas650F GATGGAGGGCTCTAACGGAT 979 Klas1608R AGTGTTGGGCTCAATGGAAGG Klas1987F CTGGCTCACCCACTTCAGTC 1337 Klas3304R GGTTCCCCATGTGTTGGAGT Klas3562F GCACTACTGCTCCCTACTATTCT 1495 Klas5045R GGACCGTCCCTTGTCGTTA Klas5049F GACAAGGGACGGTTCTACACC 737 Klas5766R ATGATCTTCATGACGGCGGG Klas6034F TTCGTTCTGCTTCCCCCAAG 1678 Klas7691R GACGGAGTAGGGAGTAAAGGC To verify the existence of this novel salivirus/klassevirus (Salivirus A SZ1), we have designed several pairs of specific primers based on the polyprotein regions. RT-PCR was performed to verify the existence of the virus in both the nasopharyngeal and the blood samples of the fatal case. All regions were detected specifically in the nasopharyngeal swab by RT-PCR. Blood samples of the fatal case were also analyzed to determine whether viremia was present; however, the PCR result was negative. Quantitative PCR was then performed using the nasopharyngeal swab. A linear range was observed for DNA concentrations from 0.075 ng/ul to 0.75 fg/ul, Ct values were observed with R 2 = 0.98, and the copy number of the virus was 1.53 × 104 copies/ul in the swab (Figs. 1, 2).Fig. 2 The distribution of C t values for 8 standard samples. Every sample required 3 repetitions Further homology analysis between the genome with all of the salivirus/klassevirus genomes from NCBI based on the 3D gene of the genome indicated a closer relationship. The new genome is closest to the Salivirus A strain in Shanghai (GU245894). A phylogenetic relationship between the new genome and selected species from the picornavirus genus based on amino acid similarity of VP1 genes was also established (Fig. 3).Fig. 3 Phylogenetic relationship of the new salivirus/klassevirus with selected species from picornavirus genus. It is based on amino acid similarity of VP1 region using neighbor-joining method with p-distance and 1000 bootstrap replications The adenovirus sequences were also analyzed in the fatal case and the control case using an in-house reference-based method. The results indicated that both the fatal case and the control case had a more coverage of the Human adenovirus 7, which is one of the most prevalent adenoviruses related to childhood respiratory diseases. The closest reference is the Human adenovirus type 7 strain NHRC 1315, complete genome with 91.6 % (the fatal case) and 89.6 % (the control case) coverage, respectively (Supplements 1 and 2), nearly all genes was covered (Fig. 4).Fig. 4 Genome structure schematic diagram of salivirus/klassevirus Discussion New virus findings in infectious diseases, especially respiratory infection diseases in children, are becoming increasingly significant. They can cause up to two million deaths in children each year in developing countries [6], and most had no defined etiology diagnosis. One reason is that there is no effective new virus detection method available for use in current clinical procedures. In addition, many infectious diseases are composed of more than one pathogen. High throughput sequencing in new virus findings may play a crucial role in the process of infectious disease. Salivirus/klassevirus is a new family in the picornavirus that is associated with diarrhea, especially in children, and is often found in feces [7]. However, the clinical significance of this virus is not clear. To the best of our knowledge, there have been no reports of infection with salivirus/klassevirus in respiratory samples, and the effect of klassevirus to humans remains unclear. Tae-Hee Han et al. [8] analyzed 142 nasopharyngeal samples in 2010, but did not detect any salivirus/klassevirus. In summary, in this study, we identified a new salivirus/klassevirus co-infection with Human adenovirus type 7 in a child with a fatal respiratory infection. This is the first salivirus/klassevirus case found that was associated with respiratory infection. The existence of the salivirus/klassevirus in the adenovirus infected child played a crucial role in the fatality and merits our attention to the need for additional studies. The first finding of salivirus/klassevirus in the respiratory sample may indicate that salivirus/klassevirus may be an etiologic agent of respiratory tract infections. It provides useful information for the clinical diagnosis of unexplained lethal infections, and expands our knowledge to the new family salivirus/klassevirus in the picornavirus. The government, disease control, and clinical workers should pay more attention to the study and control of this virus. Conclusion In conclusion, our finding of the salivirus/klassevirus in children is the first case in respiratory samples. Continued study and further investigation of the virus are needed. Electronic supplementary material Below is the link to the electronic supplementary material. Supplement 1 (TXT 34 kb). Adenovirus reads using the reference based method in the control case Supplement 2 (TXT 34 kb). Adenovirus reads using the reference based method in the fatal case Supplement 3 (TXT 7 kb). The complete genome of the new virus in FASTA format Na Pei, Jinmin Ma and Jiaosheng Zhang have contributed equally to this study. The authors gratefully acknowledge the grants from Shenzhen Key Laboratory of Transomics Biotechnologies 81 (NO. CXB201108250096A), and ShenZhen Engineering Laboratory for 83 Clinical molecular diagnostic, and China National GeneBank-Shenzhen, and International Science & Technology Cooperation Program of China (NO. 2011DFA33220). Author’s Contributions JD and XX designed and supervised the study; JZ, RZ and XW had roles in recruitment and looked after the patient. LL did laboratory testing and PCR verification. NP and JM did the genome sequencing and analysis. NP drafted the manuscript, and all authors contributed to review and revision. Compliance with ethical standards Conflict of interest The authors declare that they have no conflict of interest. Informed Consent Written informed consent was obtained from the patient for publication of this case report and any accompanying images. ==== Refs References 1. Greninger AL Runckel C Chiu CY Haggerty T Parsonnet J Ganem D DeRisi JL Virol. J. 2009 6 82 10.1186/1743-422X-6-82 19538752 2. Li L Victoria J Kapoor A Blinkova O Wang C Babrzadeh F Mason CJ Pandey P Triki H Bahri O Oderinde BS Baba MM Bukbuk DN Besser JM Bartkus JM Delwart EL J. Virol. 2009 83 22 12002 12006 10.1128/JVI.01241-09 19759142 3. Yang J Yang F Ren L Xiong Z Wu Z Dong J Sun L Zhang T Hu Y Du J Wang J Jin Q J. Clin. Microbiol. 2011 49 10 3463 3469 10.1128/JCM.00273-11 21813714 4. Lysholm F Wetterbom A Lindau C Darban H Bjerkner A Fahlander K Lindberg AM Persson B Allander T Andersson B PLoS One 2012 7 2 e30875 10.1371/journal.pone.0030875 22355331 5. Bibby K Trends Biotechnol. 2013 31 5 275 279 10.1016/j.tibtech.2013.01.016 23415279 6. Nakamura S Yang CS Sakon N Ueda M Tougan T Yamashita A Goto N Takahashi K Yasunaga T Ikuta K Mizutani T Okamoto Y Tagami M Morita R Maeda N Kawai J Hayashizaki Y Nagai Y Horii T Lida T Nakaya T PLoS One 2009 4 1 e4219 10.1371/journal.pone.0004219 19156205 7. Shan T Wang C Cui L Yu Y Delwart E Zhao W Zhu C Lan D Dai X Hua X Emerg. Infect. Dis. 2010 16 8 1303 1305 10.3201/eid1608.100087 20678331 8. Han TH Kim CH Chung JY Park SH Hwang ES Emerg. Infect. Dis. 2010 16 10 1623 1625 10.3201/eid1610.100539 20875296

PMC005xxxxxx/PMC5002280.txt

==== Front Virus GenesVirus GenesVirus Genes0920-85691572-994XSpringer US New York 136010.1007/s11262-016-1360-8ArticleMolecular epidemiological study of adenovirus infecting western lowland gorillas and humans in and around Moukalaba-Doudou National Park (Gabon) Nkogue Chimène Nze 12Horie Masayuki 13Fujita Shiho 4Ogino Michiko 5Kobayashi Yuki 6Mizukami Keijiro 7Masatani Tatsunori 13Ezzikouri Sayeh 38Matsuu Aya 13Mizutani Tetsuya 9Ozawa Makoto 13Yamato Osamu 7Ngomanda Alfred 2Yamagiwa Juichi 10Tsukiyama-Kohara Kyoko +81-99-285-3589kkohara@vet.kagoshima-u.ac.jp 131 Department of Pathological and Preventive Veterinary Science, The United Graduate School of Veterinary Science, Yamaguchi University, Yamaguchi, Japan 2 Institut de Recherche en Ecologie Tropicale (IRET), Centre National de Recherche Scientifique et Technologique (CENAREST), Libreville, Gabon 3 Transboundary Animal Diseases Research Centre, Joint Faculty of Veterinary Medicine, Kagoshima University, Kagoshima, Japan 4 Department of Behavioral Physiology and Ecology, Joint Faculty of Veterinary Medicine, Kagoshima University, Kagoshima, Japan 5 Faculty of Agriculture, Yamaguchi University, Yamaguchi, Japan 6 College of Bioresource Sciences, Nihon University, Fujisawa, Japan 7 Laboratory of Clinical Pathology, Joint Faculty of Veterinary Medicine, Kagoshima University, Kagoshima, Japan 8 Virology Unit, Viral Hepatitis Laboratory, Institut Pasteur du Maroc, Casablanca, Morocco 9 Research and Education Center for Prevention of Global Infectious Diseases of Animals, Tokyo University of Agriculture and Technology, Fuchu, Japan 10 Kyoto University, Kyoto, Japan Edited by Detlev H. Kruger. 11 6 2016 11 6 2016 2016 52 5 671 678 15 9 2015 1 6 2016 © The Author(s) 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.Adenoviruses are widespread in human population as well as in great apes, although the data about the naturally occurring adenovirus infections remain rare. We conducted the surveillance of adenovirus infection in wild western lowland gorillas in Moukalaba-Doudou National Park (Gabon), in order to investigate naturally occurring adenovirus in target gorillas and tested specifically a possible zoonotic transmission with local people inhabiting the vicinity of the park. Fecal samples were collected from western lowland gorillas and humans, and analyzed by PCR. We detected adenoviral genes in samples from both gorillas and the local people living around the national park, respectively: the overall prevalence rates of adenovirus were 24.1 and 35.0 % in gorillas and humans, respectively. Sequencing revealed that the adenoviruses detected in the gorillas were members of Human mastadenovirus B (HAdV-B), HAdV-C, or HAdV-E, and those in the humans belonged to HAdV-C or HAdV-D. Although HAdV-C members were detected in both gorillas and humans, phylogenetic analysis revealed that the virus detected in gorillas are genetically distinct from those detected in humans. The HAdV-C constitutes a single host lineage which is compatible with the host-pathogen divergence. However, HAdV-B and HAdV-E are constituted by multiple host lineages. Moreover, there is no evidence of zoonotic transmission thus far. Since the gorilla-to-human transmission of adenovirus has been shown before, the current monitoring should be continued in a broader scale for getting more insights in the natural history of naturally occurring adenoviruses and for the safe management of gorillas’ populations. Electronic supplementary material The online version of this article (doi:10.1007/s11262-016-1360-8) contains supplementary material, which is available to authorized users. Keywords AdenoviridaeGorillasGabonPhylogenyThe Ministry of Health, Labor and Welfare, JapanThe Ministry of Education, Culture, Sports, Science and Technology, Japanissue-copyright-statement© Springer Science+Business Media New York 2016 ==== Body Introduction Adenoviruses (AdVs) are non-enveloped icosahedral double-stranded DNA viruses. They belong to the family of Adenoviridae, which is divided into five genera: Mastadenovirus, Atadenovirus, Aviadenovirus, Siadenovirus, and Ichtadenovirus. Members of species belonging to genera Mastadenovirus and Atadenovirus are known to infect mammalian hosts [1, 2]. Mastadenoviruses infecting primates encompass seven Human mastadenovirus species (HAdV-A to G), the accepted species Simian mastadenovirus A and candidate species SAdV-B to G, and further not yet classified mastadenoviruses [2–4]. That classification into species is based on hemagglutination features, DNA (deoxyribonucleic acid) homology, and genomic organization [5]. There are currently over 60 HAdV types with HAdV-D containing the most members [5]. Adenoviruses were first isolated from humans and identified as the causative agent of epidemic febrile respiratory disease among military recruits in the 1950s [6, 7]. It is estimated that more than 90 % of the human population is seropositive for one or more serotypes of adenoviruses [8, 9]. The molecular biology of human-derived adenoviruses has been characterized extensively for species HAdV-C, for which human adenovirus 2 (HAdV-2) and HAdV-5 serve as prototypes [10]. Adenoviruses cause a variety of non-lethal infectious diseases in humans, and lethal disseminated adenovirus infection occurs in immunosuppressed patients [10]. The first description of a simian adenovirus in the literature was of a chimpanzee AdV [11], today known as SAdV-21 within the species H. mastadenovirus B. Later, when investigating chimpanzees suffering from kuru, four novel ape AdVs were discovered [12]. Ape AdVs have been detected or isolated from African apes including chimpanzees, bonobos, and gorillas [13–18]. Gorilla adenoviruses have been proposed to be members of HAdV-B, C, E, and F [13–18]. A recent report confirmed that the species HAdV-B which includes viruses from mixed host origin [14], originated from gorillas and have switched to humans and to chimpanzees during two different host switch events [18]. Serological surveys have found that anti-AdV antibodies were prevalent in 96 % of mountain gorillas, suggesting that AdVs are circulating among these animals [19]. In addition, Hoppe et al. recently reported high prevalence of AdV in wild apes including gorillas (45−100 %) [18]. Because AdVs are shed in the feces and saliva of infected animals [13], these viruses could possibly be transmitted among host animals via the fecal–oral route and inhalation of aerosols [20]. Comprehensive studies are still needed to clarify the origin and the diversity of adenoviruses spread in human and non-human primate populations. Thus, to fill the gap, understanding the evolution pattern of AdVs spread in non-human primates and in people frequently coming in contact with these animals is critical. In this study, we investigated AdV infection in two habituated western lowland gorilla groups in Moukalaba-Doudou national park (MDNP). In addition, we assessed AdV infection in the local people living around the national park to evaluate potential zoonotic transmissions. Materials and methods Sample collection and preparation The study site MDNP is located in the south-western part of Gabon (Fig. 1). MDNP has been reported to have a high gorilla density (more than three gorillas per square kilometer) [21], and the absence of hunting pressure from local villagers makes it a major habitat for western lowland gorillas in central Africa. From December 2010 to November 2011, during tracking, we collected 112 fresh fecal samples from 2 wild gorilla groups, which were named Group Gentil (GG) and Group 8 (G8). GG and G8 had been habituated to human observers since 2003 [21] and 2011, respectively. During the study period, GG consisted of 20−21 individuals, including 1 adult (expected age ≥13 years old) male, 6 adult (≥10 years old) females, 10 young (4–6 years old) males, and 3 young females, and all members were individually identified. In contrast, G8 was estimated to consist of 8–12 individuals, including 1 adult male, 2 adult females, and 5−8 young males and females. GG was mainly sampled near the village Doussala, in the ancient plantations, where the forest has been formerly used in various crop fields, while G8 was found far from the village in the primary forest (Fig. 1). In addition to the gorilla samples, 20 fecal samples were collected from villagers, including trackers working for the habituation of gorillas. Upon collection, each fecal specimen was immediately placed into a tube containing 2 ml of RNAlater (Ambion, Austin, TX, USA). The tubes were kept at room temperature for at most 20 days at the field camp until the samples were transported to the laboratory in Libreville, the capital city of Gabon. At the laboratory, the tubes were stored at –20 °C until DNA extraction.Fig. 1 Location features of the sampling area. a Map of Gabon, showing Moukalaba-Doudou National Park [21]. b The sampling area in the MDNP (blue line: rivers; black line: roads; red line: hunting area limitation; green line with black strips: national park limitation; dark green: primary forest; olive green: secondary forest; brown: savanna; spotted green: swamp; black circle: sampling points of G8 pointed by an arrow; gray circle pointed by an arrowhead: sampling points of GG; white circle: base camp; black rectangle with a black flag: village; white squares: habitations) (Color figure online) DNA extraction and PCR Total DNA was extracted from the sample using the QIAamp DNA Stool Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. We used the following primer sets for nested PCR: (1) 4431-s/4428-as and 4428-s/4429-as (Supplementary Table 1), targeting the HAdV DNA polymerase (DPOL) gene [14] and (2) AdhexF1/AdhexR1 and AdhexF2/AdhexR2, targeting loop 1—encompassing the hypervariable region (HVR1–6)—of the hexon gene of mastadenoviruses [22]. PCR for the DPOL gene was performed in a total volume of 20 µl containing 10 µl of 2× GoTaq Green Master Mix (Promega, Madison, WI, USA), 20 pmol of each primer, and 50 ng of DNA template. The following cycling conditions, slightly modified from Wevers et al. [14], were used: 95 °C for 2 min; 35 cycles of 95 °C for 30 s, 55 °C for 1 min, and 72 °C for 1 min; and a 7-min final extension step at 72 °C. PCR amplification of the hexon gene (HVR1–6) was performed in a total volume of 50 µl containing 200 µM of each dNTP, 20 pmol of each primer, 1.25 U of PrimeSTAR GXL polymerase (TaKaRa, Tokyo, Japan), and 50 ng of DNA template. The cycling conditions were as follows: 98 °C for 3 min; 35 cycles of 98 °C for 10 s, 45 °C for 1 min, and 72 °C for 2 min; and a final extension of 72 °C for 7 min. For the nested reaction, 2 µl of the first PCR product was amplified as above. Amplified products were separated on 1.5 % agarose gel and purified using the QIAquick Gel Extraction Kit (Qiagen) according to the manufacturer’s instructions; the amplicons were then directly sequenced with the primers for the second PCR. BLAST search BLAST searches were carried out in the NCBI database (http://blast.ncbi.nlm.nih.gov/Blast.cgi) using the determined nucleotide sequence as a query in the BLASTN program. The queries with at least 90 % identity with the deposited adenovirus gene sequences were considered for AdV species identification. Sequencing and phylogenetic analysis Twenty-four of the 27 positive samples (DNA quantity ≥5 ng/µl) were subjected to direct sequencing of DPOL gene fragments. Six samples were selected randomly for cloning and sequencing of DPOL and hexon HVR1–6 gene fragments. The PCR products were cloned into plasmid vector pCR-Blunt II-TOPO using the Zero Blunt TOPO PCR cloning kit (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Plasmid extraction was carried out using the Wizard Miniprep Kit (Promega), and the extracted plasmids were sequenced by Big Dye terminator cycle sequencing (Applied Biosystems, Foster City, CA, USA). The hexon HVR1–6 and DPOL gene sequences were edited and aligned using GENETYX software version 12.0 (Genetyx Co., Tokyo, Japan) and MEGA software version 5.05 [23]. The nucleotide sequences of DPOL (528-bp, corresponding to the position 29,200–29,727 in the reference simian adenovirus 21) and 782-bp fragments of the hexon gene (corresponding to the position 18,867–19,635 in the reference simian adenovirus 21) were aligned using MUSCLE, with the default parameters for gap opening and gap extension. These alignments were used for phylogenetic analyses. Phylogenetic trees were constructed using the neighbor-joining method in MEGA 5.05 [23]. A statistical test for the phylogeny was computed by means of bootstrapping. Percentages of 100 bootstrap replicates at the node were calculated to ensure the reliability of the trees. Nucleotide sequence accession numbers Preliminary names were given to candidate novel HAdVs following the method used by Wevers et al. [14]. The gorilla adenoviruses detected in this study were named as follows: Gorilla gorilla AdV B11-B23 (KM886307-KM886309, KM886311, KM886325-KM886328, KM886331-KM886335), Gorilla gorilla AdV C10-C18 (KM886310, KM886320-KM886324, KM886329), and Gorilla gorilla AdV E1 (KM886330). The sequences used as references for phylogenetic analysis are presented in Supplementary Table 2. Results Detection of AdV genes in western lowland gorillas in MDNP To survey AdV infection in gorillas in MDNP, we collected fecal samples from two gorilla groups (GG: well-habituated group, G8: newly habituated group) and analyzed them by nested PCR targeting the DPOL and hexon genes. The DPOL and hexon genes were detected in both groups (Table 1). The overall prevalence of AdV in the gorilla population was 24.1 % (27/112): of the 86 samples from GG, 21 were positive for both genes, 4 were positive only for the DPOL gene, and 1 sample was positive only for the hexon gene. In contrast, only 1 of the 26 samples was positive for both tested genes in G8 (Table 1). These data suggest that AdVs are naturally circulating among gorillas in MDNP. To confirm the detected AdV species, we further determined the nucleotide sequences of the amplicons and determined the species of the detected AdVs by BLAST searches. Of the tested samples, 16 belonged to HAdV-B; 10 to HAdV-C; and 1 to HAdV-E.Table 1 Detection of adenovirus DPOL and hexon genes in samples from gorilla groups in MDNP Gorilla groups No. of tested samples No. of positive samples in PCR (%) HAdV Species No. of samples B C E GG 86 26 (30.2 %) 16 9 1 G8 26 1 (3.8 %) 0 1 0 Total 112 27 (24.1 %) 16 10 1 Detection of AdV genes in local people living around the national park The prevalence of AdVs in well-habituated gorillas (30.2 % in GG group) was higher than that of newly habituated ones (3.8 % in G8 group), raising two possibilities either the AdVs in gorillas are derived from humans during the habituation process or AdVs are ubiquitous in the environment in and around the areas of human habitation. Therefore, we screened the local people (village Doussala in Fig. 1) for AdV infection. The prevalence in the local people was 35.0 % (7/20): two samples were positive for both DPOL and hexon genes, and five were positive only for the hexon gene (Table 2). These results revealed that the local people including trackers were also infected with AdVs. We sequenced the detected virus genes and identified the species of AdVs: one sample was infected with a HAdV-C type, and the others harbored HAdV-D members.Table 2 Adenovirus infection in humans Sample ID PCR DPOL PCR hexon H1 HAdV-C HAdV-C H2 HAdV-D H3 H4 H5 H6 HAdV-D H7 H8 HAdV-D H9 H10 H11 HAdV-D H12 HAdV-D H13 H14 H15 H16 H17 HAdV-D HAdV-D H18 H19 H20 Phylogenetic analysis HAdV-C genes were detected in both gorillas and humans in MDNP, suggesting zoonotic transmission of AdV between the human and gorilla populations. To investigate this possibility, as well as to gain insights into the genetic diversity of adenoviruses in MDNP, we performed phylogenetic analyses. In gorillas, on the tree based on the DPOL gene, 14 AdV genes identified in this study were divided into two groups; they clustered with SAdV-28.2, SAdV-46, SAdV-47, and gorilla AdV strains 6588 and 6575, which are representative strains of HAdV-B in gorillas, and unidentified simian adenoviruses recently described [18] (Fig. 2 and Supplementary Fig. 1). Nine AdV genes were clustered with simian AdV-45 and simian AdV-43, which are representative strains of HAdV-C in gorilla and new unidentified simian adenoviruses [18] (Fig. 2 and Supplementary Fig. 1). In contrast, one AdV gene clustered with SAdV-26 and chimpanzee AdV strain Y25, which are chimpanzee-specific strains belonging to HAdV-E (Fig. 2 and Supplementary Fig. 1). On the hexon gene-based trees, five HAdV-B (Supplementary Fig. 2a) and HAdV-E (Fig. 3) strains were identified among those isolated from gorillas. HADV-E is divided into four groups (Fig. 3): two groups of human origin and two of simian origin. The HAdV-E detected in gorillas in this study belonged to the H mastadenovirus E of simian origin (Fig. 3).Fig. 2 Phylogenetic tree of adenovirus (AdV) DPOL. The tree was constructed based on the alignment of AdV DPOL (539 bp) by using the neighbor-joining bootstrap-confirmed method in MEGA 5.05 software with 100 replicates. The names of simian isolates include the serotype nomenclature and the animal species of isolation (Ch chimpanzee, Go gorilla, Bo bonobo). Names of novel sequences obtained in this study are indicated with black dots. Bootstrap values <90 % are omitted. Scale bar, nucleotide substitutions per site Fig. 3 Phylogenetic tree of the adenovirus hexon gene loop 1 of HAdV-E. The tree was constructed based on the alignment of a 792-bp sequence of the hexon gene by using the neighbor-joining bootstrap-confirmed method in MEGA 5.05 software with 100 replicates. The names of simian isolates include the serotype nomenclature and the animal species of isolation (Ch chimpanzee, Go gorilla Bo bonobo). Names of novel sequences obtained in this study are indicated with black dots In the case of humans, the tree based on the DPOL gene showed one AdV gene clustered with HAdV-1 (HAdV-C), which is genetically different from the strains detected in gorillas (Fig. 2 and Supplementary Fig. 2b), and one clustered with the human AdV type 44 and human AdV type 47, which belong to the HAdV-D (Supplementary Fig. 1). The HAdV-D seems to be exclusively limited to the human population as reported earlier [18]. Discussion In this study, we detected several species of AdVs in western lowland gorillas in MDNP as well as in local people residing nearby. Interestingly, the positive rate in the well-habituated group (30.2 %) was higher than that of the newly habituated group (3.8 %). In addition, members of HAdV-C were detected in both gorillas and humans. However, the phylogenetic analyses revealed that the AdVs detected from gorillas are genetically distinct from those from local people living around the national park. Therefore, gorilla viruses and human viruses may have been separately circulating in each population in this region, and transmission between human and animals does not seem to happen easily in either direction, although we cannot exclude the possibility that we just missed zoonotically transmitted AdVs in this study. The difference in the prevalence between groups GG and G8 may be attributed to the quality of samples, because samples from GG might have been fresher than the ones from G8; GG was sampled while following animals, but G8 was sampled on trails, sometimes without observing the animals. In contrast, AdVs were reported to be transmitted between humans and non-human primates, indicating that AdVs have zoonotic potential [15, 18], despite the belief that AdVs have co-evolved with their hosts and are usually not transmitted to other species. Adenovirus infections have been reported in high prevalence in wild gorillas’ populations as well as in other great apes [15, 17, 18]. In this study, the overall prevalence of AdV infection in gorillas was 24.1 %, which is lower than the previously reported figure of 44.9 % in free-ranging gorillas in Congo Republic [17] or of 48 % in free-ranging gorillas in Loango National Park (Gabon) [18]. These differences might be due to the quality of the samples and/or sensitivity of the PCR. In addition, the PCR systems used in this study targeted the conserved DNA polymerase gene of mastadenovirus or the hypervariable region of the hexon gene, but in some samples, only one of the two genes was amplified. This shows that our PCR system might not be able to amplify all gene variants or that the samples could have been partially degraded [18]. Alternatively, the DNA amount of AdV in the gorillas included in this study was lower than detection limit. Further systematic studies are needed to assess these possibilities. We detected members of three species: HAdV-B, HAdV-C, and HAdV-E in western lowland gorillas in MDNP; these AdV species have been reported earlier [15–18] in western lowland gorillas as well as in other gorilla sub-species in sub-Saharan Africa. The gorilla adenoviruses of this study mainly belong to the HAdV-B (59 %). This confirms the gorilla as the major host of HAdV-B in sub-Saharan Africa. Based on the hexon tree (Supplementary Fig. 2a), the new virus named Gorilla gorilla adenovirus B19, together with the Human mastadenovirus B isolates 6560 and 6674 constitutes a single clade. The pattern observed within the species Human mastadenovirus C (Supplementary Fig. 1) is compatible with the host-pathogen divergence as previously reported [13, 15, 18]. All the lineages in HAdV-C are host specific [18]. The only member of HAdV-E detected in this study clusters with chimpanzee strains (Fig. 3). This finding supports previous report describing the non-human primate AdVs members of the HAdV-E to originate from chimpanzees [18]. We can suspect the Gorilla gorilla adenovirus E1 of this study to be the result of chimpanzee-to-gorilla transmission, as chimpanzees and gorillas are living sympatrically in MNDP. Broader screening would clarify the evolution of viruses belonging to HAdV-E. On the other hand, the adenoviruses detected in the human population around MDNP are mainly members of the HAdV-D (85.71 %) which confirms that the species HAdV-D originated in humans [18] and so far has been exclusively human specific. Four different serotypes were detected in this study, highlighting the diversity of adenoviruses circulating in the target human population. Further systematic studies should clarify the circulation of AdVs in human population. Taken together, our results show that AdVs are naturally present among gorillas and humans in MDNP in Gabon. Although there is no evidence of zoonotic transmission of AdVs in this region, our data show the feasibility of monitoring viral agents in wild habituated gorillas [24] and in local people living nearby for the safe management of wild gorilla populations and human health, as well as for understanding the evolution of virus. Since the zoonotic transmission of adenovirus already occurred during hominin evolution, assessing the zoonotic transmission of that virus in the context of habituation sites such as MNDP is recommended. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary material 1 (DOC 1146 kb) We are grateful to Dr. Bernhard Ehlers for providing the sequence alignment data of the last published AdV sequences in non-human primates, and to Professor Toshihiro Ito for his valuable comments on this paper. This study was part of a program supported by the Japanese Science and Technology Agency (JST) and Japan International Cooperation Agency (JICA) (PROCOBHA: Principal Investigator: JY). We would like to thank Centre National de la Recherche Scientifique et Technologique (CENAREST, Gabon), Institut de Recherche en Ecologie Tropicale (IRET/CENAREST, Gabon), and Agence Nationale des Parcs Nationaux (ANPN, Gabon) for permission to conduct study in Moukalaba-Doudou National Park. We also acknowledge the field assistants for help with data and sample collection, as well as the local people in Doussala, Konzi, and Mboungou for their full support during the field study. Author contributions CNN and SF designed the research. MH, YK, NO, YK, KM, TM, SE, AM, TM, MO, OY, AN, JY, and KTK helped with the experiments, data analysis, and discussion. CNN, MH, SF, and KTK have written the manuscript. ==== Refs References 1. Lehmkuhl HD Hobbs LA Serologic and hexon phylogenetic analysis of ruminant adenoviruses Arch. Virol. 2008 153 891 897 10.1007/s00705-008-0063-4 18340400 2. Pantó L Podgorski II Jánoska M Márkó O Harrach B Taxonomy proposal for Old World monkey adenoviruses: characterisation of several non-human, non-ape primate adenovirus lineages Arch. Virol. 2015 160 3165 3177 10.1007/s00705-015-2575-z 26370792 3. Chen EC Yagi S Kelly KR Mendoza SP Tarara RP Canfield DR Maninger N Rosenthal A Spinner A Bales KL Schnurr DP Lerche NW Chiu CY Cross-species transmission of a novel adenovirus associated with a fulminant pneumonia outbreak in a new world monkey colony PLoS Pathog. 2011 7 e1002155 10.1371/journal.ppat.1002155 21779173 4. Gal J Hornyak A Mandoki M Bakonyi T Balka G Szeredi L Marosan M Ludanyi T Forgach P Sos E Demeter Z Farkas SL Novel mastadenovirus infection and clinical disease in a pygmy marmoset (Callithrix [Cebuella] pygmaea ) Vet. Microbiol. 2013 167 695 699 10.1016/j.vetmic.2013.08.008 24012169 5. Robinson CM Singh G Lee JY Dehghan S Rajaiya J Liu EB Yousuf MA Betensky RA Jones MS Dyer DW Seto D Chodosh J Molecular evolution of human adenoviruses Sci. Rep. 2013 3 1812 10.1038/srep01812 23657240 6. Hilleman MR Werner JH Recovery of new agents from patients with acute respiratory illness Proc. Soc. Exp. Biol. Med. 1954 85 183 188 10.3181/00379727-85-20825 13134329 7. Rowe WP Huebner RJ Gilmore LK Parrot RH Ward TG Isolation of a cytopathic agent from human adenoids undergoing spontaneous degradation in tissue culture Proc. Soc. Exp. Biol. Med. 1953 84 570 573 10.3181/00379727-84-20714 13134217 8. D’Ambrosio E Del Grosso N Chicca A Midulla M Neutralizing antibodies against 33 human adenoviruses in normal children in Rome J. Hyg. (Lond) 1982 89 155 161 10.1017/S0022172400070650 7096999 9. Wadell GA Allard A Johansson M Svensson L Uhnoo I Enteric adenoviruses 1987 Chichester Wiley 73 91 10. Fields BN Knipe DM Howley PM Fields BN Adenoviruses Fields virology 1996 Philadelphia Lippincott-Raven Publishers 1504 11. Rowe WP Hartley JW Huebner RJ Additional serotypes of the APC virus group Proc. Soc. Exp. Biol. Med. 1956 91 260 262 10.3181/00379727-91-22231 13297771 12. Rogers NG Basnight M Gibbs CJ Gajdusek DC Latent viruses in chimpanzees with experimental kuru Nature 1967 216 446 449 10.1038/216446a0 4964401 13. Roy S Vandenberghe LH Kryazhimskiy S Grant R Calcedo R Yuan X Keough M Sandhu A Wang Q Medina-Jaszek CA Plotkin JB Wilson JM Isolation and characterization of adenoviruses persistently shed from the gastrointestinal tract of non-human primates PLoS Pathog. 2009 5 e1000503 10.1371/journal.ppat.1000503 19578438 14. Wevers D Leendertz FH Scuda N Boesch C Robbins MM Head J Ludwig C Kuhn J Ehlers B A novel adenovirus of Western lowland gorillas (Gorilla gorilla gorilla ) Virol. J. 2010 7 303 10.1186/1743-422X-7-303 21054831 15. Wevers D Metzger S Babweteera F Bieberbach M Boesch C Cameron K Couacy-Hymann E Cranfield M Gray M Harris LA Head J Jeffery K Knauf S Lankester F Leendertz SA Lonsdorf E Mugisha L Nitsche A Reed P Robbins M Travis DA Zommers Z Leendertz FH Ehlers B Novel adenoviruses in wild primates: a high level of genetic diversity and evidence of zoonotic transmissions J. Virol. 2011 85 10774 10784 10.1128/JVI.00810-11 21835802 16. Duncan M Cranfield MR Torano H Kuete HM Lee GP Glenn A Bruder JT Rangel D Brough DE Gall JG Adenoviruses isolated from wild gorillas are closely related to human species C viruses Virology 2013 444 119 123 10.1016/j.virol.2013.05.041 23806387 17. Seimon TA Olson SH Lee KJ Rosen G Ondzie A Cameron K Reed P Anthony SJ Joly DO McAloose D Lipkin WI Adenovirus and herpesvirus diversity in free-ranging great apes in the Sangha region of the Republic Of Congo PLoS One 2015 10 e0118543 10.1371/journal.pone.0118543 25781992 18. Hoppe E Pauly M Gillespie TR Akoua-Koffi C Hohmann G Fruth B Karhemere S Madinda NF Mugisha L Muyembe JJ Todd A Petrzelkova KJ Gray M Robbins M Bergi RA Wittig RM K.Zuberbuler Boesch C Schubert G Leendertz FH Ehlers B Calvignac-Spencer S Multiple cross-species transmission events of human adenoviruses (HAdV) during hominine evolution Mol. Biol. Evol. 2015 32 2072 2084 10.1093/molbev/msv090 25862141 19. C.A. Whittier, Ph.D. thesis, North Carolina State University College of Veterinary Medicine, 2009www.lib.ncsu.edu/resolver/1840.16/6215 20. Fong TT Phanikumar MS Xagoraraki I Rose JB Quantitative detection of human adenoviruses in wastewater and combined sewer overflows influencing a Michigan river Appl. Environ. Microbiol. 2010 76 715 723 10.1128/AEM.01316-09 19948848 21. Ando C Habituation and conservation of gorillas in Moukalaba-Doudou Gorilla J. 2009 38 28 29 22. Lu X Erdman DD Molecular typing of human adenoviruses by PCR and sequencing of a partial region of the hexon gene Arch. Virol. 2006 151 1587 1602 10.1007/s00705-005-0722-7 16502282 23. Tamura K Peterson D Peterson N Stecher G Nei M Kumar S MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods Mol. Biol. Evol. 2011 28 2731 2739 10.1093/molbev/msr121 21546353 24. K.V. Gilardi, T.R. Gillespie, F.H. Leendertz, E.J. Macfie, D.A. Travis, C.A. Whittier, and E.A. Williamson. Best practice guidelines for health monitoring and disease control in great ape populations. Occasional Papers of the IUCN Species Survival Commission (2015), pp. 56

PMC005xxxxxx/PMC5002281.txt

==== Front ImmunogeneticsImmunogeneticsImmunogenetics0093-77111432-1211Springer Berlin Heidelberg Berlin/Heidelberg 93810.1007/s00251-016-0938-6ReviewLocation, location, location: the evolutionary history of CD1 genes and the NKR-P1/ligand systems Rogers Sally L. 1Kaufman Jim +44-1223-766423jfk31@cam.ac.uk 231 Department of Biosciences, University of Gloucestershire, Cheltenham, GL50 4AZ UK 2 Department of Pathology, University of Cambridge, Cambridge, CB2 1QP UK 3 Department of Veterinary Medicine, University of Cambridge, Cambridge, CB3 0ES UK 25 7 2016 25 7 2016 2016 68 8 499 513 31 5 2016 4 7 2016 © The Author(s) 2016 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.CD1 genes encode cell surface molecules that present lipid antigens to various kinds of T lymphocytes of the immune system. The structures of CD1 genes and molecules are like the major histocompatibility complex (MHC) class I system, the loading of antigen and the tissue distribution for CD1 molecules are like those in the class II system, and phylogenetic analyses place CD1 between class I and class II sequences, altogether leading to the notion that CD1 is a third ancient system of antigen presentation molecules. However, thus far, CD1 genes have only been described in mammals, birds and reptiles, leaving major questions as to their origin and evolution. In this review, we recount a little history of the field so far and then consider what has been learned about the structure and functional attributes of CD1 genes and molecules in marsupials, birds and reptiles. We describe the central conundrum of CD1 evolution, the genomic location of CD1 genes in the MHC and/or MHC paralogous regions in different animals, considering the three models of evolutionary history that have been proposed. We describe the natural killer (NK) receptors NKR-P1 and ligands, also found in different genomic locations for different animals. We discuss the consequence of these three models, one of which includes the repudiation of a guiding principle for the last 20 years, that two rounds of genome-wide duplication at the base of the vertebrates provided the extra MHC genes necessary for the emergence of adaptive immune system of jawed vertebrates. Keywords 2RKLRBCLEC2http://dx.doi.org/http://dx.doi.org/10.13039/501100000384Higher Education Funding Council for EnglandHigher Education Funding Council for England (GB)issue-copyright-statement© Springer-Verlag Berlin Heidelberg 2016 ==== Body Introduction and some history Caesar Milstein, who along with Georges Koehler had discovered and developed monoclonal antibody (mAb) technology, was asked to choose the first antigen for the newly established cluster of differentiation antigen (CD) nomenclature, and he chose CD1 to be the cell surface molecule recognised by the mAb NA1/34 raised against human thymocytes (McMichael et al. 1979; Calabi and Milstein 1986; Martin et al. 1986). After decades of research, an enormous amount about CD1 genes, molecules and functions is now known (reviewed in Porcelli and Modlin 1999; Brigl and Brenner 2004; Salio et al. 2014; Mori et al. 2016, among many others). A most important point for this review is the oft-mentioned fact that the CD1 system has similarities to both the class I and class II systems. In short, the structures of CD1 genes and molecules are much like the class I system, the loading of antigen and the tissue distribution for CD1 molecules are much like the class II system, and phylogenetic analysis places CD1 between class I and class II sequences. This has led noted authorities to suggest that CD1 arose early in evolution, as a third kind of MHC molecule (Martin et al. 1986; Porcelli 1995). Just as in mammals, the story of CD1 outside of mammals began with a cell surface antigen identified by a mAb. Jim Pickel and Chen-lo Chen in the lab of Max Cooper showed that mAb CB3 raised to chicken bursal lymphocytes recognises on B cells a non-covalent complex of a 44 kDa membrane glycoprotein with chicken β2-microglobulin (β2m) with the help of mAb from Karsten Skjødt and Jan Salomonsen in the lab of Morten Simonsen (Pickel et al. 1990). At the Basel Institute for Immunology, Jan Salomonsen, convinced that this cell surface molecule would be the avian CD1, used the CB3 mAb to purify the “Pickel protein” from chicken bursal cells, but the work lurched forward with the discovery of a single expressed sequence tag (EST) in one of the first EST libraries. There followed much cloning of cDNA and genomic clones from CB chickens (with MHC haplotype B12), as well as several other inbred experimental lines. It took some years to show that there were two non-polymorphic chicken CD1 genes in tandem and in the same transcriptional orientation, both of which could be expressed as proteins on the cell surface, one of which (CD1.1) indeed is recognised by the mAb CB3 (Salomonsen et al. 2005). Using genetic mapping to find the CD1 region, Salomonsen and his colleagues were surprised to discover that the two CD1 genes mapped to the B locus, which contains the chicken MHC. At this point, a whole genome shotgun sequence for a red jungle fowl (one of the ancestors of the domestic chicken, Gallus gallus domesticus) became available (ICGSC 2004), which showed that the two CD1 genes were only 50 kb from the C4 gene at the end of the previously sequenced BF-BL region (Salomonsen et al. 2005). In the meanwhile, two other groups mined EST libraries and the red jungle fowl (chicken) genome to find CD1 genes that were located at one end of the B21 MHC region (Miller et al. 2005; Maruoka et al. 2005). Sadly, the three papers used slightly different nomenclatures (CD1.1 and CD1.2 in Salomonsen et al. are called chCD1-1 and chCD1-2 in Miller et al., but CD1.2 and CD1.1 in Maruoka et al.); we will use the Salomonsen designations throughout this review. Now, this region has been completely sequenced in one B haplotype (Shiina et al. 2007) and integrated into an overall view of the chicken MHC (Kaufman 2013; Miller and Taylor 2016). The latest additions to this story include genome mining that yields recognisable CD1 genes from several reptile species (Yang et al. 2015), including one sequence from the lizard green anole (AncaCD1; Anolis carolinensis), two expressed sequences from the Chinese alligator (AlsiCD1.1 and AlsiCD1.2, and several others by Southern blot; Alligator sinensis) and one expressed sequence and another intact sequence not found to be expressed from the Siamese crocodile (CrsiCD1.1 and CrsiCD1.2, as well as one clear pseudogene and many other sequences by Southern blot; Crocodylus siamensis). Similarly, CD1 sequences have been identified in three marsupials, a single gene each in the bandicoot (IsmaCD1; Isoodon macrourus) and the Tasmanian devil (SahaCD1; Sarcophilus harrisii), but a single pseudogene in the American opossum (ModoCD1; Monodelphis domestica) (Baker and Miller 2007; Cheng and Belov 2014). New opportunities for analysis are also provided by the low coverage genomes for many bird species that have recently been published (Jarvis et al. 2014; Zhang et al. 2014). CD1 genes are present in mammals, birds and reptiles, but isotypes, binding sites, recycling motifs and genomic locations are not well-conserved The genes reported in marsupials, chickens and reptiles were identified as CD1 genes based on a suite of characters, although each report differs in exactly what was analysed (Baker and Miller 2007; Cheng and Belov 2014; Salomonsen et al. 2005; Miller et al. 2005; Maruoka et al. 2005; Yang et al. 2015). The analyses included phylogenetic trees after sequence alignment, levels of polymorphism, hydrophobicity of putative binding grooves and presence of recycling motifs in cytoplasmic tails, and tissue distribution and cellular expression. While these genes had the broad characteristics consistent with the CD1 genes of placental mammals (that is to say, eutherians), some features were not conserved in detail or at all. Phylogenetic trees of either nucleotide or amino acid sequences show that all the CD1 genes from mammals, chickens and reptiles cluster together in a clade separate from the classical class I genes. Many other non-classical class I genes in vertebrates were found not to cluster with CD1 sequences, including the MIC, Mill and MR1 genes of mammals, the YF genes of chickens, the XNC genes of Xenopus and related frogs, and various fish genes. Interestingly, two analyses found that CD1 genes are a sister clade for the endothelial protein C receptor (ProCR) genes in humans and chickens (Maruoka et al. 2005; Papenfuss et al. 2015), an interesting finding of uncertain meaning. Anticipating the discussion below, ProCR genes are found on human chromosome 20, chicken chromosome 7 and anole chromosome 3 but not in Xenopus or zebra fish; some of the genes around ProCR are the same in chickens and anole but not in humans, and there is no obvious relationship to the MHC paralogous regions. All the CD1 genes in marsupials, chickens and reptiles have intron-exon structures consistent with eutherian mammals. The level of polymorphism has only been assessed for the chicken CD1.2 gene, for which it is extremely low. The eutherian CD1 genes are found as isotypes that differ in binding pockets and lipids bound, recycling motifs and presence in different intracellular vesicles, cell expression and tissue distribution, receptors on responding cells and function. Within placental mammals, genes of a particular isotype are more closely related between species than they are to other isotypes in the same species; for instance, human CD1D is more like mouse CD1d than like human CD1A, B, C and E. However, none of the CD1 sequences from marsupials, chickens and reptiles are equivalent to any of the five isotypes of placental mammals, as assessed by phylogenetic analysis for the whole nucleotide sequence or for the protein sequence of the whole protein or any of the domains. In essence, the whole nucleotide and protein sequences reported clustered by taxon and separate from placental mammals: the three marsupial isotypes together, the two chicken isotypes (CD1.1 and CD1.2) together and the three reptile isotypes together (with the crocodilian CD1.1 sequences together and CD1.2 sequences together, separate from lizard sequence). However, analysis of the extracellular domains placed chickens separate from reptiles for the α1 domain, but within the reptile cluster for the α2 and α3 domains (Yang et al. 2015). Such relationships are well-known for classical class I sequences, in which a birth and death model of evolution leads to expansion from a particular gene in each group of animals, so that isotypes can be less related between groups of animals than within a particular group (Nei and Rooney 2005; Eirín-López et al. 2012). The putative binding grooves formed by the α1 and α2 domains of all the CD1 genes were predicted from the sequences to be predominantly hydrophobic, more than those of mammalian classical class I molecules (although not more so than some chicken classical class I molecules, Salomonsen et al. 2005). Such hydrophobic regions are consistent with binding lipids as do eutherian CD1 molecules, and this was tested explicitly by X-ray structures for the extracellular regions of chicken CD1.1 and CD1.2 expressed in insect cells (Zajonc et al. 2008; Dvir et al. 2010). These structures show that both molecules have the same general structure as human CD1 molecules but differ in the details of the binding site. Based on the endogenous ligand from the insect cells as well as modelling and binding of exogenous ligands, the CD1.1 molecule has a large hydrophobic groove that binds lipids with two hydrophobic tails (like human CD1 molecules), with a 16 carbon chain in the F′ pocket and much longer (up to 50 carbons) chain in a long wrap-around A′ pocket that allows the end to extrude into an A′ cleft, along with the potential for substitutions on the main lipid to bind into three cavities along the A′ pocket. In contrast, CD1.2 has a single small A′ pocket (centrally located compared to the two CD1.1 pockets) which is restricted by the two α-helices being closer together as well as by various bulky amino acid side chains at the ends so that CD1.2 can only accommodate up to 16 carbons. This antigen-binding pocket that is so much smaller than those of humans was considered by the authors to be more primordial. All expressed CD1 sequences have cytoplasmic tails, and many have potential recycling motifs. All human CD1 molecules have very short cytoplasmic tails, with a four amino acid tyrosine-based motif just before the end of the sequence in CD1B, CD1C and CD1D (YQNI, YQDI, YQGV) but not for CD1A and CD1E. The same is true for the expressed marsupial sequences, IsmaCD1 and SahaCD1 which end with putative tyrosine-based motifs (YEGI, YEDM), with the ModoCD1 pseudogene sequence terminated before the transmembrane region. The two crocodilian CD1.1 sequences also have short tails with putative tyrosine-based motifs (both YQDI), but the two CD1.2 sequences are different, with two leucines at the C-terminus of a short cytoplasmic tail (a putative di-leucine motif) in AlsiCD1.2, but a potential tyrosine-based motif (YTRP) in the middle of a longer tail in CrisiCD1.2 (as well as a C-terminal glycine-leucine). The lizard CD1 also has a longer tail, with a potential tyrosine-based motif (YEDV) in the middle. The chicken molecules also differ, with a short tail ending in leucine-isoleucine (a putative di-leucine motif) for CD1.2, but a potential tyrosine-based motif (YGGC) in the middle of a much longer tail in CD1.1 (found in Salomonsen et al. 2005; Miller et al. 2005 but not in Maruoka et al. 2005). It seems likely that the marsupial, lizard and crocodilian CD1.1 cytoplasmic tails contain real tyrosine-based motifs as in human CD1 molecules, but it is not obvious for the other sequences with less canonical motifs located in the middle of longer tails. The consequences in terms of depth of recycling have not been determined for any CD1 molecule outside of mammals, except for chicken CD1.1 which co-localised with fluorescent ovalbumin but not transferrin, likely trafficking to late endosomes or lysosomes like human CD1B and CD1D (Ly et al. 2010). The expression of CD1 genes in tissues and cells was examined to differing extents in each study, although overall they were not inconsistent with mammals. Endpoint reverse-transcriptase PCR (RT-PCR) showed that RNA was very well expressed in spleen for chicken CD1.1 and CD1.2 (shown also by Northern blot), SahaCD1 (and CrsiCD1.1 and CD1.2 by quantitative RT-PCR), as well as lymph node for SahaCD1, thymus for IsmaCD1, and bursa for chicken CD1.1 and 1.2. Weaker expression was found for SahaCD1 in the lung and uterus, and much weaker for IsmaCD1 in the spleen. Good expression was found for both chicken isotypes in the bone marrow, thymus, lung and the ileum from intestine (and liver for chicken CD1.2), as well as in splenic B cells and CD8 (but not CD4) T cells. Weaker expression of CrsiCD1.1 and CD1.2 was found by quantitative RT-PCR for lung, kidney and small intestine. Expression at the protein level has only been determined for chicken CD1.1 molecules. The mAb CB3 (that recognises chicken CD1.1, Salomonsen et al. 2005) stained all B lymphocytes from bursa, a subpopulation of cells in spleen and blood, and a tiny proportion of cells in thymus, but not T cells (Pickel et al. 1990). A later study showed that the mAb NL1-1.A1 raised to chicken CD1.1 gives similar patterns of staining, with the CD1-bearing cells in the thymus located in class II-high CD3-low cells scattered in the medulla (dendritic cells or medullary epithelial cells) and putative Langerhans cells in the skin, as well as elsewhere (Ly et al. 2010). Thus, it appears that the CD1 molecules found in marsupials, chickens and reptiles do not correspond exactly to the mammalian CD1 isotypes, but that many of the features are similar, with differences particular to each species presumably reflecting the spectra of pathogens encountered (Dascher and Brenner 2003). The discovery of only a single CD1 molecule in the genomes of the bandicoot, the Tasmanian devil and the green anole suggests that the complex system of isotypes found in mammals is not essential for all vertebrates, and the presence of only a CD1 pseudogene in the opossum genome may mean that the CD1 system is dispensable. Of course, more CD1 genes may exist that have not been found in the genomes thus far, and/or other non-classical class I molecules may have evolved to fill the same purpose. However, unequivocally, one aspect of the CD1 molecules outside of mammals gave a big surprise: the genomic location. The genomic location of CD1 genes varies in different species The human CD1 genes are found together in (or near, see argument below) an MHC paralogous region on chromosome 1 (also known as the CD1 region), with the CD1 genes in other mammals in comparable (syntenic) regions. In addition to the human MHC on chromosome 6, there are (at least) three other MHC paralogous regions that have many genes (so-called paralogues) similar to those in the MHC, located on chromosomes 1, 9 and 19. The existence of these four paralogous regions was taken by Masanori Kasahara (Kasahara et al. 1997) (and independently by another group, Katsanis et al. 1996) as support for the 2R hypothesis, originally proposed by Susumo Ohno (Ohno 1970) and later given support by the group of Peter Holland (Garcia-Fernandez and Holland 1994), that there had been two rounds of genome-wide duplication near the base of the vertebrates, beginning some 600 million years (My) ago. Moreover, Kasahara proposed that class I genes were present in the primordial MHC before duplication so that class I genes would initially be distributed to the MHC paralogous regions, and at some later point in time, CD1 genes might arise by duplication in one of the paralogous regions, followed by differential loss so that eventually classical class I genes would be located in the MHC and CD1 genes in the CD1 region (and perhaps another non-classical class I gene, the Fc receptor of neonates, FcRn, in a third paralogous region, Kandil et al. 1996). The biggest surprise from the discovery of chicken CD1 genes was their genomic location at the edge of the BF-BL region, the chicken MHC, rather than in the MHC paralogous region equivalent to chromosome 1 in humans. Three models have been suggested to explain the location of CD1 genes (Fig. 1), the first being the model by Kasahara and colleagues just mentioned (Kasahara et al. 1997).Fig. 1 Three models have been proposed for the origin and subsequent evolution of CD1 genes. Lefthand column, the Kasahara model, in which classical MHC class I genes emerge in a primordial MHC before the two rounds of genome-wide duplication, then a duplication in one paralogous region gives rise to the gene that evolves into a CD1 gene, followed by differential silencing to give classical class I genes in the MHC and CD1 genes in the CD1 region of mammals, but requiring chromosomal translocations to put CD1 onto other chromosomes in birds and reptiles. Middle column, the Salomonsen-Kaufman model, in which classical MHC class I genes emerge in a primordial MHC and then one class I gene evolves into a CD1 gene, followed by two rounds of genome-wide duplication, with differential silencing to give both classical class I genes and CD1 genes in the MHC of the lineage that led to birds, classical class I genes in the MHC and CD1 genes in the CD1 region in the lineage leading to mammals, and classical class I genes in the MHC but CD1 in various regions in reptiles. Righthand column, the Hughes-Dascher model, in which the two rounds of genome-wide duplication are followed by the emergence of classical MHC class I genes in a primordial MHC next to one paralogous region, with a duplication in the MHC giving rise to the gene that evolves into a CD1 gene, with this arrangement persisting the lineage leading to birds, but with various translocations of CD1 genes out of the MHC near to a paralogous regions in the lineages leading to reptiles and mammals. Genes considered to part of the ancient paralogous region except for class I genes (green bars), classical class I genes (MHC-I) (red bars), and CD1 genes (pink bars). Paralogous regions are in order of the chromosome number found in humans: 1, chromosome 1; 6, chromosome 6 (MHC); 9, chromosome 9; 19, chromosome 19 (colour figure online) In a second model (Fig. 1), Salomonsen, Kaufman and colleagues (Salomonsen et al. 2005) interpreted this finding to mean that the CD1 genes had likely arisen by duplication from the classical class I genes already in the primordial MHC before duplication and that the differential silencing had occurred independently in the lineage leading to mammals and the lineage leading to birds, so that CD1 was present in the CD1 region of mammals but in the MHC of birds. If this scenario is true, then CD1 genes might have been expected in one or more of any of the MHC paralogous regions in other lineages. In fact, this is what has been found for the reptiles; in particular the CD1 genes are present in both the MHC and a region identified as a chromosome 19-paralogous region in alligators and crocodiles (Yang et al. 2015). However, this model cannot explain why no CD1 genes have been identified in the amphibian Xenopus or in the fish species examined. In a third model, Miller, Dascher and colleagues cited the phylogenetic trees reported by Hughes as well as their own unpublished phylogenetic trees to argue that CD1 molecules arose in a common ancestor of birds and mammals, around 310 My ago, although they did mention the Kasahara model as an alternative (Hughes 1991; Miller et al. 2005). The phylogenetic trees that formed the basis for this view can give a topology of evolution, but the timescale is dependent on the rate of substitution, which is notoriously difficult to determine in genes that are under strong selection for a new function. However, the eventual model (Fig. 1) proposed by Chris Dascher (Dascher 2007) was that the MHC with classical MHC genes only arose after the two rounds of genome-wide duplication at the base of the vertebrates, that a CD1 gene arose in this single MHC by duplication from classical class I genes followed by neofunctionalisation to bind lipids and that the CD1 gene(s) were then translocated to another chromosome during the fragmentation and reassembly of chromosomes early in the lineage leading to mammals. Thus, the presence of human CD1 genes near the MHC paralogous region on chromosome 1 was ascribed to chance, with the point being made that this paralogous region was broken into two halves anyway. In this model, the presence of the CD1 genes in the MHC is the ancestral situation, which would be found in reptiles (and perhaps in amphibians or even the ancestors of terrestrial tetrapod vertebrates). What evidence might allow us to decide between these three alternative models for the appearance of CD1 genes? There are three general points to be made. First, the Hughes-Dascher model predicts (and the Kasahara model can be consistent with) the lack of CD1 genes in most or all fish. In fact, no CD1 sequence has been found in any amphibian or fish by a variety of analyses. Among the most sensitive has been a search based on a hidden Markov model (HMM) that found all of the known class I genes (including CD1 genes in mammals and chicken) and predicted many other class I genes that have been verified but failed to find any identifiable CD1 genes in the genomes of the lizard green anole, the frog Xenopus tropicalis and fish including zebra fish Danio rerio, pufferfish Tetraodon and lamprey Petromyzon marinus (Papenfuss et al. 2015). However, there are at least four other explanations for the lack of CD1 genes identified in amphibians and fish. A trivial explanation is that the genomes are incomplete, as highlighted by the failure of the HMM to identify a CD1 gene in the green anole (Papenfuss et al. 2015), which was later found in a more complete genome. Correct long-range assembly is another issue, with a recent example being the fact that the organisation and chromosomal locations of so-called B30.2 genes only began to make sense with the most recent release of the zebra fish genome sequence (Howe et al. 2016). A third possibility is that CD1 did exist in fish but has been lost in favour of another nonclassical class I molecule which can perform the same function; in this regard, the chicken YF molecule appears to bind lipids much like CD1 (Afanassieff et al. 2001; Hee et al. 2010). A fourth explanation might be that CD1 genes arose early in evolution but simply have been lost in extant amphibians and fish. A precedent for such a loss in intervening animals is the antibody heavy chain isotype IgD, which was long thought to be specific to the mammalian lineage, until IgD genes were discovered first in fish (along with the related IgW genes), then in the frog Xenopus, and eventually in reptiles (Ohta and Flajnik 2006; Gambón-Deza and Espinel 2008; Fillatreau et al. 2013). Discovery of a CD1 gene in some fish could be accommodated by both the Kasahara model and the Salomonsen-Kaufman model. Second, the Hughes-Dascher model also predicts that the primordial MHC appeared after (not before!) the two rounds of genome-wide duplication at the base of the vertebrates. This prediction is consistent with the fact that no classical MHC genes, T cell receptor (TcR) genes, antibody genes or recombination activating genes (RAGs) have been discovered in jawless fish (Uinuk-Ool et al. 2002) (although genes that could be ancestors of TAP, TcR and CD4 have been described in lampreys, Uinuk-Ool et al. 2003; Pancer et al. 2004). However, in lampreys and hagfish, an analogous adaptive immune system based on genes encoding leucine-rich repeats (LRRs), the variable lymphocyte receptors (VLRs) have been described which are expressed by cells similar to B cells and T cells, the latter spending time in parts of the gills which express some genes like the vertebrate thymus, so-called thymoids (Pancer and Cooper 2006; Boehm et al. 2012). Again, it could be argued that the genes for the adaptive immune system were present in the common ancestor of the jawless fish and jawed vertebrates, and were then lost in the jawless fish, but that argument would require these genes to be lost from all MHC paralogous regions. In contrast, both the Kasahara model and the Salomonsen-Kaufman model predict that the classical MHC genes arose in a proto-MHC before the two rounds of genome-wide duplication. Indeed, the Salomonsen-Kaufman model is part of a larger model that was based on their discovery of an NK cell receptor gene called B-NK in the chicken MHC (Kaufman et al. 1999), which they used to suggest that not only the classical MHC genes as ligands but also their antigen processing/ peptide loading genes and their receptors were all present in primordial MHC, which would be the birthplace of the adaptive immune system (Rogers et al. 2005; Walker et al. 2011, Kaufman 2011). This view of the primordial MHC was echoed by Flajnik and Kasahara in their epic review (Flajnik and Kasahara 2010), citing the discovery of genes very similar to TAP transporter and tapasin genes in an MHC syntenic region of amphioxus, a key protochordate. However, no sequences recognisable as class I genes were found in amphioxus by the sensitive HMM method, as described above (Papenfuss et al. 2015). Third, the Hughes-Dascher model predicts that the CD1 gene(s) landed near an existing paralogous region during the fragmentation and rearrangement of chromosomes in the lineage leading to mammals, and therefore are not really a part of the MHC paralogous region on human chromosome 1 (nor would the FcRn be part of the paralogous region on human chromosome 19). It is not quite clear when this movement might have taken place, since the CD1 pseudogene on chromosome 2 in opossum is clearly located in a region with genes syntenic to the human CD1 region, while the CD1 gene of Tasmanian devils is reported to be on the same chromosome as the MHC (Baker and Miller 2007; Cheng and Belov 2014). The question of whether the FcRn gene is in or near an MHC paralogous region is even less clear. From ENSEMBL, the single FcRn gene (FCGRT) is found on human chromosome 19, opossum chromosome 4 and on a scaffold in Tasmanian devil, all very close to the genes NOSIP, RCN3 and RPS11. No such gene has been found by searching the chicken genome, but chickens do have a function for translocation of IgY to the egg. However, it depends on a molecule similar to phospholipase A2 receptor (West et al. 2004), encoded by a gene (PLA2R1) located on chromosome 7 in the chicken genome; of course this could be an example of another molecule assuming the function of a non-classical class I molecule (as mentioned above for YF and CD1). The increasingly broad understanding of potential paralogous regions coupled with constant refinement of the human genome sequence might help to determine whether the CD1 and FcRn genes were originally present in the region that was twice duplicated to become paralogous regions, or were translocated nearby after the appearance of the paralogous regions. The simplest (and most extreme) examples might be that there is a cluster of tightly linked paralogous genes with either the CD1/FcRn genes embedded in the middle and obviously part of the paralogous region, or located some way off and obviously not part of the paralogous region. Sadly, the situation is more complex than is usually presented (Fig. 2). The genes that were originally identified as being paralogous (Kasahara 1997) are located within a 3.2-Mb region in the MHC on human chromosome 6 but are spread over roughly 21 to 76 Mb in each of the paralogous regions, with these paralogous genes separated by many tens of other genes that are not obviously paralogous (as found by locating these genes in ENSEMBL). More recent analyses, taking into account other organisms (Kasahara 2000; Abi-Rached et al. 2002; Azumi et al. 2003; Danchin and Pontarotti 2004; Kasahara et al. 2004; Suurväli et al. 2014), improve but do not substantially change this view (with interesting details outside the scope of this review). The selective identification of genes considered to be paralogous in a sea of genes that are not obviously paralogous (particularly for genes that are part of a large multigene family with members throughout the genome) certainly led to skepticism for the early claims for two rounds of genome-wide duplication at the base of the vertebrates (for instance, Hughes 1998), which was largely dispelled once the amphioxus genome became available (Putnam et al. 2008).Fig. 2 The human MHC paralogous regions differ significantly in size, with the CD1 genes in the middle of a large region on chromosome 1 relative close to paralogous genes, the classical class I genes on the edge of a small region on chromosome 6 very close to paralogous genes, and the FcRn gene relatively far away from one end of paralogous genes on chromosome 19. Comparison of the four paralogous regions based on the genes originally proposed by Kasahara (along with the FcRn or FCGRT gene, but with those deemed later to be questionable excluded), roughly to scale with the chromosomal position according to the current assembly of the human genome (GRCh38.p5). Circles indicate centromeres, thin lines indicate gene(s) considered to be paralogous (many tens to thousands of other genes not shown), and positions are given in megabases (Mb) according to EMSEMBL (www.ensembl.org/Homo_sapiens). Class I genes (red) (colour figure online) The CD1 genes are found in the middle of the 76 Mb covering the original paralogous genes on chromosome 1, relatively close to some well-defined paralogous genes such as LMNA, SPTA1, PBX1 and RXR. This finding is most easily accommodated by the CD1 genes being part of the region before duplication, but it remains possible that the CD1 genes were inserted into the middle of an existing paralogous region. In contrast, the FcRn gene is located at one end of the 47-Mb region containing the paralogous genes on chromosome 19, relatively far away (and across the centromere) from the bulk of those paralogous genes (particularly as determined by later analyses, for instance Suurväli et al. 2014). Thus, examination of the human genome does not rule out any of the proposed models but it would seem most parsimonious for CD1 genes to be present before the genome-wide duplications and consistent with FcRn gene appearing later in evolution. Although it might be plausible that the CD1 genes moved from the MHC to a different chromosome in the lineages leading to mammals, another potential challenge to be explained by the Hughes-Dascher model is the presence of reptile CD1 genes in more than one chromosomal location (Yang et al. 2015), specifically in both the MHC (like some birds) and near to genes identified as similar to the paralogous region on human chromosome 19. One possibility is that the assemblies of these genomes are not accurate, and further work will show that all these genes are together in the MHC. Another possibility is that there have been multiple translocations out of the MHC, with the reptile CD1 genes landing near an MHC paralogous region on the equivalent of human chromosome 1 in mammals as well as near an MHC paralogous region on the equivalent of human chromosome 19. A third possibility is that the Kasahara and Salomonsen-Kaufman models are more parsimonious explanations for the data. Recently, low-coverage whole genome sequences have been reported for a wide variety of birds (Jarvis et al. 2014; Zhang et al. 2014). It must be kept in mind that many important genes will be absent in such partial genomes and that there will be many assembly errors for those genes that are present. However, cursory analysis finds CD1 genes in all bird genomes available (nearly 100 genes from over 40 species), ranging from the Palaeognathae (ostrich, kiwi and tinamou) to the Passeriformes (songbirds) of the Neoaves. Phylogenetic analysis (Fig. 3) of all available CD1 sequences (full length protein) gives four clades: one for mammals (both eutherian and marsupial, but surprisingly with one CD1 sequence from tinamou), one for reptiles, and two for birds, one with chicken CD1.1 and one with chicken CD1.2 (with most birds having at least one of each), suggesting that both avian isotypes were present in an early ancestor.Fig. 3 Neighbour-joining tree showing relationships for a selection of CD1 genes from eutherian and marsupial mammals, birds and reptiles. Alignment of full length amino acid sequences was done using MUSCLE. Confidence for major nodes indicated; other tree-building algorithms gave the same major clades of mammals (but including one tinamou sequence), reptiles and two avian clades, but joined with different topologies. Species indicated as follows: Alsi, Alligator sinensis (Chinese alligator); Anca, Anolis carolinensis (Carolina anole lizard); Ancy, Anser cygnoides domesticus (swan goose); Anpl, Anas platyrhynchos (mallard duck); Apau, Apteryx australis mantelli (North island brown kiwi); Apfo, Aptenodytes forsteri (emperor penguin); Apvi, Apaloderma vittatum (bar-tailed trogon); Aqch, Aquila chrysaetos Canadensis (golden eagle); Bare, Balearica regulorum gibbericeps (grey crowned crane); Caca, Caprimulgus carolinensis (chuck-will’s widow); Cacr, Cariama cristata (red-legged seriema); Capu, Calidris pugnax (ruff); Chma, Chlamydotis macqueenii (MacQueen’s bustard); Chpe, Chaetura pelagica (chimney swift); Chvo, Charadrius vociferous (killdeer); Cobr, Corvus brachyrhynchos (American crow); Coli, Columba livia (rock dove); Crsi, Crocodylus siamensis (Siamese crocodile); Egga, Egretta garzetta (little egret); Euhe, Eurypyga helias (sun bittern); Fach, Falco cherrug (Saker falcon); Fape, Falco peregrinus (peregrine falcon); Fial, Ficedula albicollis (collared flycatcher); Gaga, Gallus gallus (chicken); Gefo, Geospiza fortis (medium ground finch); Hale, Haliaeetus leucocephalus (bald eagle); Hosa, Homo sapiens (human); Isma, Isoodon macrourus (northern brown bandicoot); Mavi, Manacus vitellinus (green collared manakin); Neno, Nestor notabilis (kea); Nini, Nipponia Nippon (Japanese crested ibis); Pama, Parus major (great tit); Phca, Phalacrocorax carbo (great cormorant); Phle, Phaethon lepturus (white-tailed tropicbird); Picoides pubescens (downy woodpecker); Pshu, Pseudopodoces humilis (ground tit); Saha, Sarcophilus harrisii (Tasmanian devil); Seca, Serinus canaria (Atlantic canary); Stca, Struthio camelus australis (ostrich); Stvu, Sturnus vulgaris (common starling); Taer, Tauraco erythrolophus (red-crested turaco); Tagu, Taeniopygia guttata (zebra finch); Tigu, Tinamus guttatus (white-throated tinamou); Zoal, Zonotrichia albicollis (white-throated sparrow) Examination of the existing genome sequences reveals that avian CD1 genes are typically found on one to three contigs, each with one to three CD1 genes. To our surprise, only in the most primitive of bird species (Palaeognathae which include the flightless ratites, and Galloanseres of the Neoaves which include chickens and ducks) are the CD1 genes found next to recognisable MHC genes (Fig. 4). For instance, white throated tinamou (Tinamus guttatus), considered to be most primitive of all birds (Jarvis et al. 2014), has two putative CD1 genes next to each other with tenascin, with one of the CD1 genes midway in sequence between CD1 and classical class I genes. The ostrich (Struthio camelus australis) has a single CD1 gene together with TAP2 and class I genes, and the north island brown kiwi (Apteryx australis mantelli) has five CD1 genes, three together without other recognisable genes, and two together with tenascin. The mallard duck (Anas platyrhynchos) has four CD1 genes on three contigs, two together with a zinc finger, one with tenascin and one with nothing recognisable. The swan goose (Anser cygnoides domesticus) also has four CD1 genes on three contigs, two together with tenascin, TRIM7/butyrophilin-like, C4/α2-macroglobulin-like and two MHC class I genes, one with APOBEC4, and one with dynactin. This last constellation of genes is most intriguing, with class I genes typical of the mammalian class I region, tenascin typical of the mammalian class III region, TRIM7 and butyrophilin typical of the mammalian extended class I region, and α2-macroglobulin found in mammalian natural killer complex (NKC). As mentioned above, it has been suggested long ago that the NKC and the MHC were originally adjacent, as part of the primordial MHC (Rogers et al. 2005). For the bird lineage, it would appear that the ancestor had CD1 genes in the MHC (as well as other genomic locations), which were then lost from the MHC in the Neoaves.Fig. 4 The relative position of CD1 genes on scaffolds from selected avian genomes is shown in comparison to other genes. Genes were identified (S. Rogers, unpublished) using the ENSEMBL genome browser (www.ensembl.org), and most likely orthology of the gene confirmed by blasting back against the NCBI database (www.ncbi.nlm.nih.gov). All potential CD1 genes (red bold) are indicated, designated with letters going backwards from z. In addition, MHC (blue bold) genes as well as other genes (black) are shown, with uncharacterised predicted proteins labelled UC. The tinamou CD1 gene that groups with mammals in phylogenetic analyses is labelled CD1y, and the other CD1 gene on that contig may in fact be two genes, one of which is more closely related to an MHC class I gene. Lengths of contigs are shown with arrows, and where other genes are inserted is indicated by double forward slash (//), with the distance shown if significant. kb kilobases, Mb megabases (colour figure online) Overall, it would appear that the CD1 genes can be found in the MHC and more than one of the MHC paralogous regions, apparently even in the same species. One explanation would be that CD1 genes were present in all four paralogous regions in the last common ancestor of reptiles, birds and mammals, and have been differentially silenced in each lineage, indeed in each species. However, if CD1 genes were present in all four paralogous regions, were they originally present in amphibians and the many thousands of different fish species, and are now lost? Another explanation might be that CD1 genes can move around in the genome, perhaps preferentially between genomic regions with some level of sequence homology, as might be expected for MHC paralogous regions. This possibility leads us to consider another unexpected finding in the chicken MHC, the presence of the lectin-like NK receptor and potential ligand. The genomic location of lectin-like NK receptor and ligand genes vary in different species One of the many surprises that arose from the first genomic sequence of the chicken MHC was the presence of two genes that encoded lectin-like domains with apparent transmembrane exons located as in type II membrane proteins, located head-to-head in opposite transcriptional orientation. The one called B-NK was very clearly most closely related to a well-defined NK receptor in mammals, NKR-P1, and was found to be expressed in chicken NK cell clones (Kaufman et al. 1999). The other called B-lec was found to be closely-related to early activation antigens such as CD69, but most closely to human LLT1 and mouse clr genes, which were eventually shown to be ligands of NKR-P1 (Rogers et al. 2005). Subsequently, a sequencing facility renamed the genes, so in their view, B-lec became Blec1, B-NK became Blec2, and further genes with high identity to B-lec became Blec3, Blec4 and so on (Shiina et al. 2007). This nomenclature sadly removes both the relative relationships and the functional associations that are so critical to any deep understanding of the biology. A more appropriate nomenclature has been established in which NKR-P1 genes are part of the KLRB1 family, and genes related to LLT1/clr/B-lec are part of the CLEC2 family (Carlyle et al. 2008; Kirkham and Carlyle 2014), a nomenclature which we will adopt when appropriate for the rest of this review. In mammals, NKR-P1 (KLRB1) gene(s) and ligand (CLEC2) gene(s) are found to be located (typically in receptor-ligand pairs) in the natural killer complex (NKC), together with other lectin-like receptor genes such as KLRG1 (also known as MAFA) and other signature genes such as TMEM52b, GABARAPL1 and α2-macroglobulin. The NKC is located on chromosome 12 in humans, so not with the immunoglobulin-like NK receptors in the leukocyte receptor complex (LRC) on chromosome 19 or in the MHC on chromosome 6, or near to any of the MHC paralogous regions (Carlyle et al. 2008; Kirkham and Carlyle 2014). In fact, two chicken lectin-like genes are found in a large genomic region along with the genes for α2-macroglobulin, TMEM52b and GABARAPL1, but neither lectin-like gene is an NK receptor (Chiang et al. 2007; Neulen and Göbel 2012). As for the CD1 genes, we proposed that B-NK and B-lec were present in the primordial MHC before the two rounds of genome-wide duplication, and then these genes were differentially silenced in different paralogous regions of the lineage leading to mammals and to birds. We proposed that the NKC and the MHC were originally one region, in which both receptors and ligands were present in order to co-evolve (Rogers et al. 2005; 2008; Kaufman 2011); this idea has been picked up by others (Flajnik and Kasahara 2010). We found that the B-NK gene had high allelic polymorphism (with a different allele for each MHC haplotype), while B-lec was virtually monomorphic (Rogers and Kaufman 2008). Unexpectedly, no evidence was found that B-NK recognised B-lec (Viertlboeck et al. 2008), but multiple genes with high sequence identity to B-lec were found in the TRIM and BG regions next to the MHC and in the unlinked Rfp-Y region on the same chromosome (Rogers et al. 2003; Shiina et al. 2007; Salomonsen et al. 2014), any of which could be ligands for B-NK. Sequences for quail also showed a variety of B-lec genes (Shiina et al. 2004). It seemed possible to explain the appearance of a multigene family of B-lec genes with a single B-NK gene in and around the MHC of chickens and other galliforms, but it was a shock to find that the B-NK and B-lec gene pair in a passerine bird, the zebra finch (Taeniopygia guttata), is located on the Z sex chromosome (Ekblom et al. 2011). A further analysis using the most current zebra finch genome (taeGut3.2.4) reveals a CLEC2 gene and an unannotated KLRB1 gene next to SLC1A1 and CDC37 on the Z chromosome (Fig. 5), with no lectin-like genes near the TMEM52b and GABARAPL1 genes in chromosome 1, a region otherwise syntenic with the NKC. Similarly, the genome (Ellegren et al. 2012, FicAlb_1.4 version 84.1) of another passerine bird, the collared flycatcher (Ficedula albicollis), has both a KLRB1 gene and a CLEC2 gene in between SLC1a1 and CDC37L1.Fig. 5 The relative positions of CLEC2 and KLRB genes on chromosomes (or linkage groups, LG) in different species are shown in comparison to other genes. Genes were identified (S. Rogers, unpublished) using the ENSEMBL genome browser (www.ensembl.org), and most likely orthology of the gene confirmed by blasting back against the NCBI database (www.ncbi.nlm.nih.gov). In some species, the marker gene could not be identified or located and therefore it is not shown. All potential CLEC2 (green bold) and KLRB (blue bold) genes, as well as selected KLRG (purple bold), NKC (red bold) and MHC (black bold) genes identified in the various species are shown. Where other genes are inserted is indicated by double forward slash (//), with the distance shown if significant (Mb megabases) (colour figure online) Despite the fragmentary nature of the data (as discussed above), the low-coverage whole genome sequences for a wide variety of birds (Jarvis et al. 2014; Zhang et al. 2014) highlight the possibilities for a variety of locations for KLRB1 and CLEC2 genes. In the emperor penguin, for example, KLRB1-CLEC2 gene pairs are associated with a TRIM7 gene (which in mammals is located in the extended MHC) and with TMEM52b and GABARAPL1 (which in mammals are found in the NKC). Another eight CLEC genes are found in this putative NKC, as well as a KLRG1 gene next to α2-macroglobulin on a separate scaffold. Overall, KLRB1-CLEC2 gene pairs or singletons are found associated with genes that in mammals are found in the NKC, the extended MHC and/or the Z chromosome. Interestingly, the Z chromosome is believed to share a common ancestor with human chromosome 9 (Bellott et al. 2010), which also contains an MHC paralogous region. Thus, one possibility is that a piece of an MHC paralogous region became part of the Z chromosome in birds and was silenced in some bird lineages but not in passerine birds. In the genome sequence of the lizard, the green anole (Alföldi et al. 2011, AnoCar2.0 version 84.2), one KLRG1 gene, two KLRB1 genes and one CLEC2 gene are found on chromosome 2, along with many genes found in the mammalian NKC (including TMEM52b, GABARAPL1 and α2-macroglobulin), the MHC of chickens and/or mammals (including two class I genes, a CD1 gene and TRIM7.1) and the avian Z chromosome (including SLC1A1). Although this collection of genes is spread over 150 Mb, their presence all on the same chromosome is consistent with the idea that all three regions in birds were once together in a reptilian ancestor. The presence of KLRB1-CLEC2 gene pairs is less clear in amphibians and fish, although there is some evidence for an NKC. In the genome sequence of the frog, Xenopus tropicalis (Hellsten et al. 2010, JGI 4.2 version 84.42), a single CLEC2 gene is located next to α2-macroglobulin, with TMEM52b and GABARAPL1 located on another scaffold, suggestive of either incomplete assembly or a fragmented NKC. In a representative of the teleost fish, the zebra fish, the genome sequence (Howe et al. 2013, GRCz10 version 84.10) had no obvious KRLB1 or CLEC2 genes, and of the genes characteristic of the NKC, GABARAPL1 and α2-macroglobulin were found but on separate chromosomes. The genome sequence of a less derived fish, the spotted gar (Braasch et al. 2016, LepOcu1 version 84.1), also had no obvious KRLB1 or CLEC2 genes, but there were four lectin-like receptor genes very similar to KRLG1 in between GABARAPL1 and APOBEC genes, along with another KLRG1 gene close by, suggesting that an NKC was already present in basal fish (and possibly lost in teleost fish). Sadly, examination of the elephant shark genome (Venkatesh et al. 2014, accession number AAVX02000000) gave no indication of KLRG1, KRLB1 or CLEC2 genes, or of a syntenic region for the NKC (although this genome is particularly poorly assembled). Overall, the evidence is most consistent with the idea that the lectin-like genes of the NKC were present very early, likely in the primordial MHC before two rounds of genome-wide duplication, and that these genes were differentially silenced in different paralogous regions in different taxa. The alternative is that these genes move around selectively between particular regions of the genome. There is an enormous literature about recombination, translocation, chromosomal breakpoints and so on during evolution (as just one example in chickens, Romanov et al. 2014), far beyond the scope of this review. The important point to note is that these genes, ones that seem to be part of ancient MHC paralogous regions but in fact might not be, would have moved selectively near or into paralogous regions. Whether regional sequence identity or the presence of particular sequence features might contribute to such selective movement is again beyond the scope of this review. However, it is also possible that the appearance of such genes near ancient MHC paralogous regions is only due to our expectations and that when rigorous statistical analyses are carried out, the location of these genes will turn out to be random. The way backwards Our review of the evolution of CD1 and certain lectin-like NK cell receptors leaves more questions than answers. Even for the few marsupial, bird and reptile CD1 genes known, there is a great deal to learn through comparative immunology. In particular, the cells in which each of these molecules are expressed, the recycling patterns that the molecules undergo, the nature of their binding sites and the kinds of lipids that actually bind these molecules, and the cells that recognise the lipid-CD1 complexes all remain to be discovered. Some approaches should be relatively straightforward. Expression of CD1 molecules will allow structural studies (like those already done for chicken CD1) as well as the generation of monoclonal antibodies which can be used to determine tissue expression, study cell biology and isolate molecules to determine the lipids bound, which in turn will allow the production of lipid-CD1 multimers to discover the responding cells. A crucial but much more difficult step is to determine the role of each of these CD1 molecules in response to particular diseases (or for normal regulation of the immune response, if that is important). Once all of these tasks are accomplished, it may become possible to pick out the important patterns that have governed the details of CD1 evolution, in order to understand which features have been selected and which features are due to chance. The journey to understand the lectin-like NK receptors and their ligands is likely to be similar but much more complicated, given the enormous complexity of NK cell biology. Overshadowing the comparative biology of both the CD1 and lectin-like NK receptors is the question of their genomic histories. For nearly 20 years, a guiding principle for the evolution of adaptive immunity has been the concept of genome-wide duplications at the base of the vertebrates leading to the extra genes that provided the potential to evolve an adaptive immune system. A large literature has explored this concept (although there are as many reviews as actual data papers), and it remains a powerful way to organise and explain the disparate data that have been gathered. However, the Hughes-Dascher model completely repudiates this view, with the original adaptive immune system of jawed vertebrates envisaged to have appeared next to one of the paralogous regions after the two rounds of genome-wide duplication. Is it possible, after all of the sound and fury, that genome-wide duplication is in fact irrelevant to the origin and evolution of the adaptive immune system of the jawed vertebrates? Despite our own published models and beliefs, the existing data do not allow us to choose definitively between the dogma of Kasahara and the catastrophic antithesis of Hughes and Dasher. The analogous adaptive immune system of VLRs discovered in jawless fish does not provide clear evidence one way or the other, since the jawless fish have undergone at least one round of genome-wide duplication (Flajnik and Kasahara 2010). However, two other analogous systems have been reported: fibrinogen-related proteins (FREPs) in molluscs which gain somatic diversity by point mutation (Zhang et al. 2004; 2008) and Down syndrome cell adhesion molecule (DSCAMs) in arthropods which gain somatic diversity through alternative splicing (Schmucker and Chen 2009). If FREPs and DSCAMs are confirmed as adaptive immune systems among invertebrates (Ng et al. 2014; Armitage et al. 2015; Gordy et al. 2015), then genome-wide duplications are not necessary for the emergence of adaptive immunity. However, such duplications may still have played an important role in the emergence of our own adaptive immune system. Only time (and more research) will tell. This article is published in the Special Issue CD1, MR1, NKT, and MAIT: Evolution and Origins of Non-peptidic Antigen Recognition by T lymphocytes with Guest Editor Dr. Dirk Zajonc ==== Refs References Abi-Rached L Gilles A Shiina T Pontarotti P Inoko H Evidence of en bloc duplication in vertebrate genomes Nat Genet 2002 31 1 100 5 10.1038/ng855 11967531 Afanassieff M Goto RM Ha J Sherman MA Zhong L Auffray C Coudert F Zoorob R Miller MM At least one class I gene in restriction fragment pattern-Y (Rfp-Y), the second MHC gene cluster in the chicken, is transcribed, polymorphic, and shows divergent specialization in antigen binding region J Immunol 2001 166 5 3324 33 10.4049/jimmunol.166.5.3324 11207288 Alföldi J Di Palma F Grabherr M Williams C Kong L Mauceli E Russell P Lowe CB Glor RE Jaffe JD Ray DA Boissinot S Shedlock AM Botka C Castoe TA Colbourne JK Fujita MK Moreno RG ten Hallers BF Haussler D Heger A Heiman D Janes DE Johnson J de Jong PJ Koriabine MY Lara M Novick PA Organ CL Peach SE Poe S Pollock DD de Queiroz K Sanger T Searle S Smith JD Smith Z Swofford R Turner-Maier J Wade J Young S Zadissa A Edwards SV Glenn TC Schneider CJ Losos JB Lander ES Breen M Ponting CP Lindblad-Toh K The genome of the green anole lizard and a comparative analysis with birds and mammals Nature 2011 477 7366 587 91 10.1038/nature10390 21881562 Armitage SA Peuss R Kurtz J Dscam and pancrustacean immune memory – a review of the evidence Dev Comp Immunol 2015 48 2 315 23 10.1016/j.dci.2014.03.004 24657209 Azumi K De Santis R De Tomaso A Rigoutsos I Yoshizaki F Pinto MR Marino R Shida K Ikeda M Ikeda M Arai M Inoue Y Shimizu T Satoh N Rokhsar DS Du Pasquier L Kasahara M Satake M Nonaka M Genomic analysis of immunity in a Urochordate and the emergence of the vertebrate immune system: "waiting for Godot" Immunogenetics 2003 55 8 570 81 10.1007/s00251-003-0606-5 14530883 Baker ML Miller RD Evolution of mammalian CD1: marsupial CD1 is not orthologous to the eutherian isoforms and is a pseudogene in the opossum Monodelphis domestica Immunology 2007 121 1 113 21 10.1111/j.1365-2567.2007.02545.x 17244156 Bellott DW Skaletsky H Pyntikova T Mardis ER Graves T Kremitzki C Brown LG Rozen S Warren WC Wilson RK Page DC Convergent evolution of chicken Z and human X chromosomes by expansion and gene acquisition Nature 2010 466 7306 612 6 10.1038/nature09172 20622855 Boehm T McCurley N Sutoh Y Schorpp M Kasahara M Cooper MD VLR-based adaptive immunity Annu Rev Immunol 2012 30 203 20 10.1146/annurev-immunol-020711-075038 22224775 Braasch I Gehrke AR Smith JJ Kawasaki K Manousaki T Pasquier J Amores A Desvignes T Batzel P Catchen J Berlin AM Campbell MS Barrell D Martin KJ Mulley JF Ravi V Lee AP Nakamura T Chalopin D Fan S Wcisel D Cañestro C Sydes J Beaudry FE Sun Y Hertel J Beam MJ Fasold M Ishiyama M Johnson J Kehr S Lara M Letaw JH Litman GW Litman RT Mikami M Ota T Saha NR Williams L Stadler PF Wang H Taylor JS Fontenot Q Ferrara A Searle SM Aken B Yandell M Schneider I Yoder JA Volff JN Meyer A Amemiya CT Venkatesh B Holland PW Guiguen Y Bobe J Shubin NH Di Palma F Alföldi J Lindblad-Toh K Postlethwait JH The spotted gar genome illuminates vertebrate evolution and facilitates human-teleost comparisons Nat Genet 2016 48 4 427 37 10.1038/ng.3526 26950095 Brigl M Brenner MB CD1: antigen presentation and T cell function Annu Rev Immunol 2004 22 817 90 10.1146/annurev.immunol.22.012703.104608 15032598 Calabi F Milstein C A novel family of human major histocompatibility complex-related genes not mapping to chromosome 6 Nature 1986 323 6088 540 3 10.1038/323540a0 3093894 Carlyle JR Mesci A Fine JH Chen P Bélanger S Tai LH Makrigiannis AP Evolution of the Ly49 and Nkrp1 recognition systems Semin Immunol 2008 20 6 321 30 10.1016/j.smim.2008.05.004 18595730 Chiang HI Zhou H Raudsepp T Jesudhasan PR Zhu JJ Chicken CD69 and CD94/NKG2-like genes in a chromosomal region syntenic to mammalian natural killer gene complex Immunogenetics 2007 59 7 603 11 10.1007/s00251-007-0220-z 17505822 Cheng Y Belov K Characterisation of non-classical MHC class I genes in the Tasmanian devil (Sarcophilus harrisii) Immunogenetics 2014 66 12 727 35 10.1007/s00251-014-0804-3 25267059 Danchin EG Pontarotti P Towards the reconstruction of the bilaterian ancestral pre-MHC region Trends Genet 2004 20 12 587 91 10.1016/j.tig.2004.09.009 15522451 Dascher CC Evolutionary biology of CD1 Curr Top Microbiol Immunol 2007 314 3 26 17593655 Dascher CC Brenner MB Evolutionary constraints on CD1 structure: insights from comparative genomic analysis Trends Immunol 2003 24 8 412 8 10.1016/S1471-4906(03)00179-0 12909453 Dvir H Wang J Ly N Dascher CC Zajonc DM Structural basis for lipid-antigen recognition in avian immunity J Immunol 2010 184 5 2504 11 10.4049/jimmunol.0903509 20100930 Eirín-López JM Rebordinos L Rooney AP Rozas J The birth-and-death evolution of multigene families revisited Genome Dyn 2012 7 170 96 10.1159/000337119 22759819 Ekblom R Stapley J Ball AD Birkhead T Burke T Slate J Genetic mapping of the major histocompatibility complex in the zebra finch (Taeniopygia guttata) Immunogenetics 2011 63 8 523 30 10.1007/s00251-011-0525-9 21494955 Ellegren H Smeds L Burri R Olason PI Backström N Kawakami T Künstner A Mäkinen H Nadachowska-Brzyska K Qvarnström A Uebbing S Wolf JB The genomic landscape of species divergence in Ficedula flycatchers Nature 2012 491 7426 756 60 23103876 Fillatreau S Six A Magadan S Castro R Sunyer JO Boudinot P The astonishing diversity of Ig classes and B cell repertoires in teleost fish Front Immunol 2013 4 28 10.3389/fimmu.2013.00028 23408183 Flajnik MF Kasahara M Origin and evolution of the adaptive immune system: genetic events and selective pressures Nat Rev Genet 2010 11 1 47 59 10.1038/nrg2703 19997068 Gambón-Deza F Espinel CS IgD in the reptile leopard gecko Mol Immunol 2008 45 12 3470 6 10.1016/j.molimm.2008.02.027 18486212 Garcia-Fernandez J Holland PW Archetypal organization of the amphioxus Hox gene cluster Nature 1994 370 563 566 10.1038/370563a0 7914353 Gordy MA Pila EA Hanington PC The role of fibrinogen-related proteins in the gastropod immune response Fish Shellfish Immunol 2015 46 1 39 49 10.1016/j.fsi.2015.03.005 25765166 Hee CS Gao S Loll B Miller MM Uchanska-Ziegler B Daumke O Ziegler A Structure of a classical MHC class I molecule that binds "non-classical" ligands PLoS Biol 2010 8 12 e1000557 10.1371/journal.pbio.1000557 21151886 Hellsten U Harland RM Gilchrist MJ Hendrix D Jurka J Kapitonov V Ovcharenko I Putnam NH Shu S Taher L Blitz IL Blumberg B Dichmann DS Dubchak I Amaya E Detter JC Fletcher R Gerhard DS Goodstein D Graves T Grigoriev IV Grimwood J Kawashima T Lindquist E Lucas SM Mead PE Mitros T Ogino H Ohta Y Poliakov AV Pollet N Robert J Salamov A Sater AK Schmutz J Terry A Vize PD Warren WC Wells D Wills A Wilson RK Zimmerman LB Zorn AM Grainger R Grammer T Khokha MK Richardson PM Rokhsar DS The genome of the Western clawed frog Xenopus tropicalis Science 2010 328 5978 633 6 10.1126/science.1183670 20431018 Howe K Clark MD Torroja CF Torrance J Berthelot C Muffato M Collins JE Humphray S McLaren K Matthews L McLaren S Sealy I Caccamo M Churcher C Scott C Barrett JC Koch R Rauch GJ White S Chow W Kilian B Quintais LT Guerra-Assunção JA Zhou Y Gu Y Yen J Vogel JH Eyre T Redmond S Banerjee R Chi J Fu B Langley E Maguire SF Laird GK Lloyd D Kenyon E Donaldson S Sehra H Almeida-King J Loveland J Trevanion S Jones M Quail M Willey D Hunt A Burton J Sims S McLay K Plumb B Davis J Clee C Oliver K Clark R Riddle C Elliot D Threadgold G Harden G Ware D Begum S Mortimore B Kerry G Heath P Phillimore B Tracey A Corby N Dunn M Johnson C Wood J Clark S Pelan S Griffiths G Smith M Glithero R Howden P Barker N Lloyd C Stevens C Harley J Holt K Panagiotidis G Lovell J Beasley H Henderson C Gordon D Auger K Wright D Collins J Raisen C Dyer L Leung K Robertson L Ambridge K Leongamornlert D McGuire S Gilderthorp R Griffiths C Manthravadi D Nichol S Barker G Whitehead S Kay M Brown J Murnane C Gray E Humphries M Sycamore N Barker D Saunders D Wallis J Babbage A Hammond S Mashreghi-Mohammadi M Barr L Martin S Wray P Ellington A Matthews N Ellwood M Woodmansey R Clark G Cooper J Tromans A Grafham D Skuce C Pandian R Andrews R Harrison E Kimberley A Garnett J Fosker N Hall R Garner P Kelly D Bird C Palmer S Gehring I Berger A Dooley CM Ersan-Ürün Z Eser C Geiger H Geisler M Karotki L Kirn A Konantz J Konantz M Oberländer M Rudolph-Geiger S Teucke M Lanz C Raddatz G Osoegawa K Zhu B Rapp A Widaa S Langford C Yang F Schuster SC Carter NP Harrow J Ning Z Herrero J Searle SM Enright A Geisler R Plasterk RH Lee C Westerfield M de Jong PJ Zon LI Postlethwait JH Nüsslein-Volhard C Hubbard TJ Roest Crollius H Rogers J Stemple DL The zebrafish reference genome sequence and its relationship to the human genome Nature 2013 496 7446 498 503 10.1038/nature12111 23594743 Howe K, Schiffer PH, Zielinski J, Wiehe T, Laird GK, Marioni JC, Soylemez O, Kondrashov K, Leptin M (2016) Open Biol 6(4):160009 Hughes AL Evolutionary origin and diversification of the mammalian CD1 antigen genes Mol Biol Evol 1991 8 2 185 201 1710755 Hughes AL Phylogenetic tests of the hypothesis of block duplication of homologous genes on human chromosomes 6, 9, and 1 Mol Biol Evol 1998 15 7 854 70 10.1093/oxfordjournals.molbev.a025990 9656486 International Chicken Genome Sequencing Consortium Sequence and comparative analysis of the chicken genome provide unique perspectives on vertebrate evolution Nature 2004 432 7018 695 716 10.1038/nature03154 15592404 Jarvis ED Mirarab S Aberer AJ Li B Houde P Li C Ho SY Faircloth BC Nabholz B Howard JT Suh A Weber CC da Fonseca RR Li J Zhang F Li H Zhou L Narula N Liu L Ganapathy G Boussau B Bayzid MS Zavidovych V Subramanian S Gabaldón T Capella-Gutiérrez S Huerta-Cepas J Rekepalli B Munch K Schierup M Lindow B Warren WC Ray D Green RE Bruford MW Zhan X Dixon A Li S Li N Huang Y Derryberry EP Bertelsen MF Sheldon FH Brumfield RT Mello CV Lovell PV Wirthlin M Schneider MP Prosdocimi F Samaniego JA Vargas Velazquez AM Alfaro-Núñez A Campos PF Petersen B Sicheritz-Ponten T Pas A Bailey T Scofield P Bunce M Lambert DM Zhou Q Perelman P Driskell AC Shapiro B Xiong Z Zeng Y Liu S Li Z Liu B Wu K Xiao J Yinqi X Zheng Q Zhang Y Yang H Wang J Smeds L Rheindt FE Braun M Fjeldsa J Orlando L Barker FK Jønsson KA Johnson W Koepfli KP O’Brien S Haussler D Ryder OA Rahbek C Willerslev E Graves GR Glenn TC McCormack J Burt D Ellegren H Alström P Edwards SV Stamatakis A Mindell DP Cracraft J Braun EL Warnow T Jun W Gilbert MT Zhang G Whole-genome analyses resolve early branches in the tree of life of modern birds Science 2014 346 6215 1320 31 10.1126/science.1253451 25504713 Kandil E Egashira M Miyoshi O Niikawa N Ishibashi T Kasahara M The human gene encoding the heavy chain of the major histocompatibility complex class I-like Fc receptor (FCGRT) maps to 19q13.3. Cytogenet Cell Genet 1996 73 1–2 97 8 10.1159/000134316 8646894 Kasahara M Genome paralogy: a new perspective on the organization and origin of the major histocompatibility complex Curr Top Microbiol Immunol 2000 248 53 66 10793474 Kasahara M Nakaya J Satta Y Takahata N Chromosomal duplication and the emergence of the adaptive immune system Trends Genet 1997 13 3 90 2 10.1016/S0168-9525(97)01065-2 9066265 Kasahara M Suzuki T Pasquier LD On the origins of the adaptive immune system: novel insights from invertebrates and cold-blooded vertebrates Trends Immunol 2004 25 2 105 11 10.1016/j.it.2003.11.005 15102370 Katsanis N Fitzgibbon J Fisher EM Paralogy mapping: identification of a region in the human MHC triplicated onto human chromosomes 1 and 9 allows the prediction and isolation of novel PBX and NOTCH loci Genomics 1996 35 1 101 8 10.1006/geno.1996.0328 8661110 Kaufman J Kaufmann SHE Rouse BT Sachs DL The evolutionary origins of the adaptive immune system of jawed vertebrates (Chapter 3, pp 41–55) The Immune Response to Infection 2011 Washington DC American Society of Microbiology Press Kaufman J (2013) The Avian MHC (Chapter 8, pp 149–167). In: Avian Immunology, second edition; K. A. Schat, P. Kaiser and B. Kaspers, editors. Elsevier, Ltd Kaufman J Milne S Göbel TW Walker BA Jacob JP Auffray C Zoorob R Beck S The chicken B locus is a minimal essential major histocompatibility complex Nature 1999 401 6756 923 5 10.1038/44856 10553909 Kirkham CL Carlyle JR Complexity and Diversity of the NKR-P1:Clr (Klrb1:Clec2) Recognition Systems Front Immunol 2014 5 214 10.3389/fimmu.2014.00214 24917862 Ly N Danzl NM Wang J Zajonc DM Dascher CC Conservation of CD1 protein expression patterns in the chicken Dev Comp Immunol 2010 34 2 123 32 10.1016/j.dci.2009.08.015 19735672 Martin LH Calabi F Milstein C Isolation of CD1 genes: a family of major histocompatibility complex-related differentiation antigens Proc Natl Acad Sci U S A 1986 83 23 9154 8 10.1073/pnas.83.23.9154 3097645 Maruoka T Tanabe H Chiba M Kasahara M Chicken CD1 genes are located in the MHC: CD1 and endothelial protein C receptor genes constitute a distinct subfamily of class-I-like genes that predates the emergence of mammals Immunogenetics 2005 57 8 590 600 10.1007/s00251-005-0016-y 16133451 McMichael AJ Pilch JR Galfré G Mason DY Fabre JW Milstein C A human thymocyte antigen defined by a hybrid myeloma monoclonal antibody Eur J Immunol 1979 9 3 205 10 10.1002/eji.1830090307 376318 Miller MM Taylor RL Jr Brief review of the chicken Major Histocompatibility Complex: the genes, their distribution on chromosome 16, and their contributions to disease resistance Poult Sci 2016 95 2 375 92 10.3382/ps/pev379 26740135 Miller MM Wang C Parisini E Coletta RD Goto RM Lee SY Barral DC Townes M Roura-Mir C Ford HL Brenner MB Dascher CC Characterization of two avian MHC-like genes reveals an ancient origin of the CD1 family Proc Natl Acad Sci U S A 2005 102 24 8674 9 10.1073/pnas.0500105102 15939884 Mori L, Lepore M, De Libero G (2016) The Immunology of CD1- and MR1-Restricted T Cells. Annu Rev Immunol Nei M Rooney AP Concerted and birth-and-death evolution of multigene families Annu Rev Genet 2005 39 121 52 10.1146/annurev.genet.39.073003.112240 16285855 Ng TH Chiang YA Yeh YC Wang HC Review of Dscam-mediated immunity in shrimp and other arthropods Dev Comp Immunol 2014 46 2 129 38 10.1016/j.dci.2014.04.002 24727482 Neulen ML Göbel TW Identification of a chicken CLEC-2 homologue, an activating C-type lectin expressed by thrombocytes Immunogenetics 2012 64 5 389 97 10.1007/s00251-011-0591-z 22205394 Ohno S Evolution by gene duplication 1970 New York Springer-Verlag Ohta Y Flajnik M IgD, like IgM, is a primordial immunoglobulin class perpetuated in most jawed vertebrates Proc Natl Acad Sci U S A 2006 103 28 10723 8 10.1073/pnas.0601407103 16818885 Pancer Z Cooper MD The evolution of adaptive immunity Annu Rev Immunol 2006 24 497 518 10.1146/annurev.immunol.24.021605.090542 16551257 Pancer Z Mayer WE Klein J Cooper MD Prototypic T cell receptor and CD4-like coreceptor are expressed by lymphocytes in the agnathan sea lamprey Proc Natl Acad Sci U S A 2004 101 36 13273 8 10.1073/pnas.0405529101 15328402 Papenfuss AT Feng ZP Krasnec K Deakin JE Baker ML Miller RD Marsupials and monotremes possess a novel family of MHC class I genes that is lost from the eutherian lineage BMC Genomics 2015 16 535 10.1186/s12864-015-1745-4 26194104 Pickel JM Chen CL Cooper MD An avian B-lymphocyte protein associated with beta 2-microglobulin Immunogenetics 1990 32 1 1 7 10.1007/BF01787321 2197221 Porcelli SA The CD1 family: a third lineage of antigen-presenting molecules Adv Immunol 1995 59 1 98 10.1016/S0065-2776(08)60629-X 7484459 Porcelli SA Modlin RL The CD1 system: antigen-presenting molecules for T cell recognition of lipids and glycolipids Annu Rev Immunol 1999 17 297 329 10.1146/annurev.immunol.17.1.297 10358761 Putnam NH Butts T Ferrier DE Furlong RF Hellsten U Kawashima T Robinson-Rechavi M Shoguchi E Terry A Yu JK Benito-Gutiérrez EL Dubchak I Garcia-Fernàndez J Gibson-Brown JJ Grigoriev IV Horton AC de Jong PJ Jurka J Kapitonov VV Kohara Y Kuroki Y Lindquist E Lucas S Osoegawa K Pennacchio LA Salamov AA Satou Y Sauka-Spengler T Schmutz J Shin-I T Toyoda A Bronner-Fraser M Fujiyama A Holland LZ Holland PW Satoh N Rokhsar DS The amphioxus genome and the evolution of the chordate karyotype Nature 2008 453 7198 1064 71 10.1038/nature06967 18563158 Romanov MN Farré M Lithgow PE Fowler KE Skinner BM O’Connor R Fonseka G Backström N Matsuda Y Nishida C Houde P Jarvis ED Ellegren H Burt DW Larkin DM Griffin DK Reconstruction of gross avian genome structure, organization and evolution suggests that the chicken lineage most closely resembles the dinosaur avian ancestor BMC Genomics 2014 15 1060 10.1186/1471-2164-15-1060 25496766 Rogers SL Göbel TW Viertlboeck BC Milne S Beck S Kaufman J Characterization of the chicken C-type lectin-like receptors B-NK and B-lec suggests that the NK complex and the MHC share a common ancestral region J Immunol 2005 174 6 3475 83 10.4049/jimmunol.174.6.3475 15749883 Rogers SL Kaufman J High allelic polymorphism, moderate sequence diversity and diversifying selection for B-NK but not B-lec, the pair of lectin-like receptor genes in the chicken MHC Immunogenetics 2008 60 8 461 75 10.1007/s00251-008-0307-1 18574582 Rogers S Shaw I Ross N Nair V Rothwell L Kaufman J Kaiser P Analysis of part of the chicken Rfp-Y region reveals two novel lectin genes, the first complete genomic sequence of a class I alpha-chain gene, a truncated class II beta-chain gene, and a large CR1 repeat Immunogenetics 2003 55 2 100 8 12692693 Rogers SL Viertlboeck BC Göbel TW Kaufman J Avian NK activities, cells and receptors Semin Immunol 2008 20 6 353 60 10.1016/j.smim.2008.09.005 18948017 Salio M Silk JD Jones EY Cerundolo V Biology of CD1- and MR1-restricted T cells Annu Rev Immunol 2014 32 323 66 10.1146/annurev-immunol-032713-120243 24499274 Salomonsen J Chattaway JA Chan AC Parker A Huguet S Marston DA Rogers SL Wu Z Smith AL Staines K Butter C Riegert P Vainio O Nielsen L Kaspers B Griffin DK Yang F Zoorob R Guillemot F Auffray C Beck S Skjødt K Kaufman J Sequence of a complete chicken BG haplotype shows dynamic expansion and contraction of two gene lineages with particular expression patterns PLoS Genet 2014 10 6 e1004417 10.1371/journal.pgen.1004417 24901252 Salomonsen J Sørensen MR Marston DA Rogers SL Collen T van Hateren A Smith AL Beal RK Skjødt K Kaufman J Two CD1 genes map to the chicken MHC, indicating that CD1 genes are ancient and likely to have been present in the primordial MHC Proc Natl Acad Sci U S A 2005 102 24 8668 73 10.1073/pnas.0409213102 15939887 Schmucker D Chen B Dscam and DSCAM: complex genes in simple animals, complex animals yet simple genes Genes Dev 2009 23 2 147 56 10.1101/gad.1752909 19171779 Shiina T Briles WE Goto RM Hosomichi K Yanagiya K Shimizu S Inoko H Miller MM Extended gene map reveals tripartite motif, C-type lectin, and Ig superfamily type genes within a subregion of the chicken MHC-B affecting infectious disease J Immunol 2007 178 11 7162 72 10.4049/jimmunol.178.11.7162 17513765 Shiina T Shimizu S Hosomichi K Kohara S Watanabe S Hanzawa K Beck S Kulski JK Inoko H Comparative genomic analysis of two avian (quail and chicken) MHC regions J Immunol 2004 172 11 6751 63 10.4049/jimmunol.172.11.6751 15153492 Suurväli J Jouneau L Thépot D Grusea S Pontarotti P Du Pasquier L Rüütel Boudinot S Boudinot P The proto-MHC of placozoans, a region specialized in cellular stress and ubiquitination/proteasome pathways J Immunol 2014 193 6 2891 901 10.4049/jimmunol.1401177 25114105 Uinuk-Ool T Mayer WE Sato A Dongak R Cooper MD Klein J Lamprey lymphocyte-like cells express homologs of genes involved in immunologically relevant activities of mammalian lymphocytes Proc Natl Acad Sci U S A 2002 99 22 14356 61 10.1073/pnas.212527699 12391333 Uinuk-ool TS Mayer WE Sato A Takezaki N Benyon L Cooper MD Klein J Identification and characterization of a TAP-family gene in the lamprey Immunogenetics 2003 55 1 38 48 12679854 Venkatesh B Lee AP Ravi V Maurya AK Lian MM Swann JB Ohta Y Flajnik MF Sutoh Y Kasahara M Hoon S Gangu V Roy SW Irimia M Korzh V Kondrychyn I Lim ZW Tay BH Tohari S Kong KW Ho S Lorente-Galdos B Quilez J Marques Bonet T Raney BJ Ingham PW Tay A Hillier LW Minx P Boehm T Wilson RK Brenner S Warren WC Elephant shark genome provides unique insights into gnathostome evolution Nature 2014 505 7482 174 9 10.1038/nature12826 24402279 Viertlboeck BC Wortmann A Schmitt R Plachý J Göbel TW Chicken C-type lectin-like receptor B-NK, expressed on NK and T cell subsets, binds to a ligand on activated splenocytes Mol Immunol 2008 45 5 1398 404 10.1016/j.molimm.2007.08.024 17950458 Walker BA Hunt LG Sowa AK Skjødt K Göbel TW Lehner PJ Kaufman J The dominantly expressed class I molecule of the chicken MHC is explained by coevolution with the polymorphic peptide transporter (TAP) genes Proc Natl Acad Sci U S A 2011 108 20 8396 401 10.1073/pnas.1019496108 21536896 West AP Jr Herr AB Bjorkman PJ The chicken yolk sac IgY receptor, a functional equivalent of the mammalian MHC-related Fc receptor, is a phospholipase A2 receptor homolog Immunity 2004 20 5 601 10 10.1016/S1074-7613(04)00113-X 15142528 Yang Z Wang C Wang T Bai J Zhao Y Liu X Ma Q Wu X Guo Y Zhao Y Ren L Analysis of the reptile CD1 genes: evolutionary implications Immunogenetics 2015 67 5–6 337 46 10.1007/s00251-015-0837-2 25921705 Zajonc DM Striegl H Dascher CC Wilson IA The crystal structure of avian CD1 reveals a smaller, more primordial antigen-binding pocket compared to mammalian CD1 Proc Natl Acad Sci U S A 2008 105 46 17925 30 10.1073/pnas.0809814105 19004781 Zhang G Li C Li Q Li B Larkin DM Lee C Storz JF Antunes A Greenwold MJ Meredith RW Ödeen A Cui J Zhou Q Xu L Pan H Wang Z Jin L Zhang P Hu H Yang W Hu J Xiao J Yang Z Liu Y Xie Q Yu H Lian J Wen P Zhang F Li H Zeng Y Xiong Z Liu S Zhou L Huang Z An N Wang J Zheng Q Xiong Y Wang G Wang B Wang J Fan Y da Fonseca RR Alfaro-Núñez A Schubert M Orlando L Mourier T Howard JT Ganapathy G Pfenning A Whitney O Rivas MV Hara E Smith J Farré M Narayan J Slavov G Romanov MN Borges R Machado JP Khan I Springer MS Gatesy J Hoffmann FG Opazo JC Håstad O Sawyer RH Kim H Kim KW Kim HJ Cho S Li N Huang Y Bruford MW Zhan X Dixon A Bertelsen MF Derryberry E Warren W Wilson RK Li S Ray DA Green RE O’Brien SJ Griffin D Johnson WE Haussler D Ryder OA Willerslev E Graves GR Alström P Fjeldså J Mindell DP Edwards SV Braun EL Rahbek C Burt DW Houde P Zhang Y Yang H Wang J Consortium AG Jarvis ED Gilbert MT Wang J Comparative genomics reveals insights into avian genome evolution and adaptation Science 2014 346 6215 1311 20 10.1126/science.1251385 25504712 Zhang SM Adema CM Kepler TB Loker ES Diversification of Ig superfamily genes in an invertebrate Science 2004 305 5681 251 4 10.1126/science.1088069 15247481 Zhang SM Zeng Y Loker ES Expression profiling and binding properties of fibrinogen-related proteins (FREPs), plasma proteins from the schistosome snail host Biomphalaria glabrata Innate Immun 2008 14 3 175 89 10.1177/1753425908093800 18562576

PMC005xxxxxx/PMC5002282.txt

==== Front J Struct Funct GenomicsJ. Struct. Funct. GenomicsJournal of Structural and Functional Genomics1345-711X1570-0267Springer Netherlands Dordrecht 920410.1007/s10969-016-9204-2ArticleDevelopment of a protein–ligand-binding site prediction method based on interaction energy and sequence conservation Tsujikawa Hiroto Sato Kenta Wei Cao Saad Gul Sumikoshi Kazuya Nakamura Shugo Terada Tohru http://orcid.org/0000-0002-3245-0052Shimizu Kentaro +81-3-5841-5455shimizu@bi.a.u-tokyo.ac.jp Department of Biotechnology, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-Ku, Tokyo 113-8657 Japan 11 7 2016 11 7 2016 2016 17 2 39 49 23 1 2016 20 6 2016 © The Author(s) 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.We present a new method for predicting protein–ligand-binding sites based on protein three-dimensional structure and amino acid conservation. This method involves calculation of the van der Waals interaction energy between a protein and many probes placed on the protein surface and subsequent clustering of the probes with low interaction energies to identify the most energetically favorable locus. In addition, it uses amino acid conservation among homologous proteins. Ligand-binding sites were predicted by combining the interaction energy and the amino acid conservation score. The performance of our prediction method was evaluated using a non-redundant dataset of 348 ligand-bound and ligand-unbound protein structure pairs, constructed by filtering entries in a ligand-binding site structure database, LigASite. Ligand-bound structure prediction (bound prediction) indicated that 74.0 % of predicted ligand-binding sites overlapped with real ligand-binding sites by over 25 % of their volume. Ligand-unbound structure prediction (unbound prediction) indicated that 73.9 % of predicted ligand-binding residues overlapped with real ligand-binding residues. The amino acid conservation score improved the average prediction accuracy by 17.0 and 17.6 points for the bound and unbound predictions, respectively. These results demonstrate the effectiveness of the combined use of the interaction energy and amino acid conservation in the ligand-binding site prediction. Electronic supplementary material The online version of this article (doi:10.1007/s10969-016-9204-2) contains supplementary material, which is available to authorized users. Keywords Protein structureProtein–ligand bindingBinding site predictionInteraction energySequence conservationthe Platform Project for Supporting in Drug Discovery and Life Science Research from Japan Agency for Medical Research and Development (AMED).issue-copyright-statement© Springer Science+Business Media Dordrecht 2016 ==== Body Introduction It is well known that the biological function of many proteins depends on binding to small molecules, termed ligands. Therefore, the function of a protein can be inferred by determining what kinds of ligands it binds. In addition, in recent years, the three-dimensional (3D) structures of proteins have been used in structure-based drug design. Because ligands bind to specific sites on the surfaces of proteins, identification of the ligand-binding sites is an essential step in these studies. Various methods have been developed for ligand-binding site prediction, and because ligand-binding sites are often located in large depressions (pockets) on protein surfaces, many of these prediction methods use 3D protein structures to predict ligand-binding sites. These structure-based methods can be largely classified into two groups: (a) purely geometric methods [1–4] and (b) energetic methods [5, 6]. In the purely geometric approaches, the ligand-binding site is presumed to be located within the largest pocket on the protein surface. However, when the size of the pocket is larger than that of the ligand, the exact binding site cannot be easily predicted. Furthermore, the spatial range of the detected pocket varies between the prediction methods, reflecting differences in the definition of the pocket among the methods. The energetic methods are based on the concept that a ligand binds the site where the interaction energy with the protein is minimal. To search for such a site, ligands are virtually placed on the protein surface, and the interaction energy with protein atoms is calculated at each position to estimate the stability of the binding site. Laurie and Jackson’s Q-SiteFinder [6] is one of the most successful energetic methods for predicting ligand-binding sites. Q-SiteFinder first places methyl probes (−CH3) in a grid around a protein molecule and calculates van der Waals interaction energy between the atoms of the protein and probes. Probes with low energy are then clustered. The clustering is repeated until a cluster with the total interaction energy of the probes being lower than a defined threshold is obtained. Clusters thus obtained are ranked according to the total energy, and the cluster with the lowest total energy is expected to be the most appropriate for the ligand-binding site. However, energy-based methods are not always superior to purely geometric methods, and adequate precision cannot be achieved by only defining the “energetically stable site.” Therefore, improvement of the precision of the prediction has been attempted by combining new information that indicates ligand-binding site-like features with the conventional methods. Two sets of information have often been used in practice. One is amino acid frequency around ligand-binding sites. Amino acids on a protein surface are more likely to be in the ligand-binding sites than those buried in the protein. Accordingly, the likelihood of a site being a ligand-binding site can be quantified by evaluating the frequencies of the 20 amino acids for the site and comparing them with those for the protein surface and protein interior [7]. The other set of information is amino acid conservation. Because ligand-binding sites are the most important sites for expressing protein function, there is a strong tendency for amino acids around binding sites to be conserved among homologues [8]. Several approaches have been attempted to improve the precision of prediction by applying one of these two sets of information to conventional methods. Kulharia et al. [7] reported improvement in prediction precision by the inclusion of amino acid frequency in the prediction by Q-SiteFinder. In contrast, sequence conservation (amino acid conservation) has been employed in purely structure-based methods. In LIGSITEcsc [4], multiple pockets obtained from a grid search are re-ranked based on the degree of amino acid conservation in the proteins. Concavity [9] is a method that employs amino acid conservation in pocket searching; this method differs from LIGSITEcsc in incorporating conservation information directly into the search for pockets rather than using conservation information to postprocess predicted pockets. Capra et al. [9] states that concavity outperforms LIGSITEcsc because of this difference. In the present study, we developed a prediction method that combined amino acid conservation with an energy-based method. The energy-based pocket search is similar to Q-SiteFinder and our previous work [18], and the amino acid conservation is directly incorporated into the ranking of ligand-binding sites, as in concavity. As for the energy calculation, we use van der Waals energy as an interaction energy between a carbon atom probe and the protein. This energy is not related to a physical protein–ligand-binding energy and is used only as a tool to identify and rank protein cavities. Such an abstraction can be useful for coping with various types of ligands, particularly when the ligands are not known in advance. We collected a wide range of test data and performed a general assessment of prediction precision. Our method successfully predicted binding sites with a higher precision than a conventional energy-based method, Q-SiteFinder, clarifying the effect of amino acid conservation on the prediction precision. Materials and methods Dataset construction Two sets of protein structures corresponding to each other are developed in the present study: ligand-bound structures in which ligands are bound to proteins and ligand-unbound structures that are ligand-free structures of proteins in the ligand-bound structure set. The list of each protein set was obtained from LigASite [10], a protein–ligand-binding site database. A non-redundant version of the list including 391 protein pairs was used. Among them, two proteins were omitted because ligand-bound coordinates could not be found. Also, as we only consider homo-multimer proteins, seven hetero-multimer proteins were omitted. Hence, in the present study, 382 protein pairs were used in total. For some pairs, there are multiple ligand-bound structures in the list. In those cases, one structure was randomly selected from the list for each pair. Biological oligomeric assemblies were obtained from the predicted quaternary structures using Proteins, Interfaces, Structures and Assemblies (PISA) [11]. The protein structures used in this study were determined by X-ray crystallography with a resolution of ≤2.4 Å and R-factor ≤0.25. According to LigASite [10], the dataset is limited to binding sites for the clusters of non-water molecules appearing in the HETATM records of Protein Data Bank (PDB) entries within 4 Å that consist of at least 10 heavy atoms and that make at least 70 interatomic contacts with protein atoms. This method was adopted to select biologically relevant ligands in this study. As for the quality of ligands and ligand-binding site structures of PDB, we performed the analysis of B-factor and Local Ligand Density Fit (LLDF). The results of these analyses and a list of PDBIDs of 382 ligand-bound and ligand-unbound structures used in the present study are shown in Online Supplementary Material. Prediction method outline Our prediction method includes the following steps, illustrated for the example of biotin binding of streptavidin in Fig. 1a.Fig. 1 Ligand-binding site prediction. a An example using the protein (streptavidin) and the ligand (biotin), PDBID: 1stp. b Placement of probes around the protein and calculation of interaction energies between probes and the protein. c Forming clusters of probes. d Ranking clusters. e Clusters of the top three ranking. Steps related to amino acid conservation are not shown Addition of hydrogen atoms and construction of missing side chains of proteins. Placement of carbon atom probes around the protein (Fig. 1b). Calculation of van der Waals energy. After calculation of interaction energy of all the probes, if a probe with lower interaction energy is found or a probe with similarly low interaction energy is found within 1 Å of another probe, cluster them together. Expand the size of a cluster until it reaches a defined size (Fig. 1c). Calculation of amino acid conservation. Weighting of interaction energy with amino acid conservation. Ranking of clusters in ascending order according to total probe interaction energy within a cluster (Fig. 1d, e). Addition of hydrogen atoms and construction of missing side chains of proteins Normally, 3D coordinates obtained from PDB and PISA does not include hydrogen atoms if they were determined by X-ray crystallography. In the proposed method, a hydrogen addition tool, protonate, in the molecular dynamics software AMBER10 was used to add hydrogen atoms to proteins for calculating van der Waals energy between all atoms and probes of a protein. With regard to proteins with side chains whose complete atomic coordinates were not registered, a side chain modeling tool, SCWRL3 [12], was used to reconstruct the missing side chain before the addition of hydrogen atoms. Placement of carbon atom probes around the protein Carbon atom probes were placed in a grid around a protein with 0.5 Å intervals. The van der Waals interactions commonly appear in the protein–ligand binding, and we calculated the van der Waals interaction energy between a carbon atom and the protein. This interaction energy is not directly related to real ligand-binding energy; it is only used for obtaining a position with minimum energies. Carbon atoms are simple and have lower precision than all-atom methane or methyl probes; however, in this model, coordinates of hydrogen atoms that change because of atomic rotation need not be considered, indicating an advantage of reduced computational complexity. Calculation of van der Waals energy Van der Waals energy between probe–protein atoms was calculated using the equation, Lennard–Jones 6–12 potential Evan: Ei,jvan=εjεj×Ri+Rjrij12-2×Ri+Rjrij6 In this equation, i and j represent probe and protein atom, respectively, whereas rij indicates the distance between two atoms, R the van der Waals distance (Å), and ε the van der Waals depth (kcal/mol). Each atom has its unique R and ε. In the present study, van der Waals energy was calculated based on a force field parameter, parm94 [13]. To reduce computation cost, van der Waals energy between a probe and protein atoms located within 10 Å of the probe was calculated for each probe. Clustering of probes Interaction energy with protein atoms was calculated for each probe, and probes were clustered using the value of the interaction energy. First, the probe with the lowest energy was found, and its energy was used as the energy threshold. Next, the energy threshold was increased by 0.1 kcal/mol, and probes with energy lower than the energy threshold were searched for. If such probes were found, probes with energy lower than the energy threshold and located within 1 Å of that probe were clustered together. The cluster can be merged during this step. The distance between clusters was defined as the distance between two probes in the clusters nearest to each other. When no more probes were found for clustering, the energy threshold of the search was broadened by 0.1 kcal/mol. The maximum number of probes that could be included in a cluster was defined as Pnum, and when the total number of probes within a cluster reached Pnum, the clustering process was completed. Calculation of amino acid conservation PSI-BLAST, with up to three iterations against the NCBI non-redundant (nr) database, was used to compare the sequence similarity between the amino acid sequence of the target protein and the sequences of its homologues, and the resulting position-specific scoring matrix was used to calculate amino acid conservation. When P indicates amino acid frequency in a protein and Q indicates the amino acid frequency in the background, the Kullback–Leibler divergence can be obtained [14] as the difference between the mean frequency and both P and Q. Furthermore, the Jensen–Shannon divergence may be derived [15] by calculating the mean of Kullback–Leiber divergences and was defined as the degree of amino acid conservation, Cscore. Cscore and Evan [in Eq. (1)] obtained as described above were used for calculating the weighted score. The weighted score of probe i is defined as follows: 1 Ei=∑j=1nw1Ei,jvan+w2Cscorej+w3Ei,jvanCscorej where n is the number of the protein atoms that interact with probe i and j denotes their index. Cscore is defined for each amino acid residue and is applied to all the atoms of that residue. In this equation, w1, w2, and w3 indicate the weights on the respective terms. We changed the values of w2 and w3 with the value of w1 fixed to 0 or 1. Only Evan of protein atoms within 10 Å of a probe and Cscore were weighted. This is because the amino acids distant from a ligand were considered to have low influence on the stable binding of the ligand. Accordingly, for the second term of Eq. (1), amino acids found ≤6 Å from a probe was used by calculating Cscore, and amino acids located >6 Å away from the ligand were not used. Ranking of clusters All clusters obtained were arranged in the ascending order of the sum of the probe scores Ei in them. For multimeric proteins, similar clusters with similar energy values were obtained for corresponding sites of all chains. In these cases, we consider only one cluster with the lowest energy value among them. Methods for assessment and comparison of prediction results Two measures were used to assess the performance: the proportion of predicted clusters that spatially overlap with actual ligands above chosen threshold (hereafter referred to as ligand-binding space prediction) and the degree of agreement between amino acids around a cluster and the exact amino acid around the ligand (hereafter referred to as ligand-binding residue prediction). Assessment of the results of ligand-binding space prediction For ligand-bound structures, a method of calculating the proportion of clusters that spatially cover ligands [6] was used for performance assessment. We determine that a probe in a cluster overlaps with the ligand if the probe is located within 1.6 Å from one of the heavy atoms of the ligand. We adopted the same value 1.6 Å as that of Q-SiteFinder to compare the performance directly. The term “precision” used here defines the proportion of probes in a cluster that overlap with ligands. When the value is ≥25 %, the cluster is regarded as a success. A precision of 100 % means that all probes in the cluster are located within 1.6 Å of one of the heavy atoms of the ligand, and when the cluster is larger, the precision decreases. In some cases, the prediction results even with small precision values are useful; however, if we regard a cluster with a precision >0 % as a success, a very large cluster that accidentally includes a small number of probes within 1.6 Å from the ligand will also be a success. To avoid such cases, we referred to Laurie et al. and defined a precision threshold of ≥25 %. As for the ligand-unbound structures, the ligand coordinates are not available, so ligand coordinates should be copied to the ligand-unbound structure from its pair (i.e., ligand-bound structure) to obtain a pseudo-binding site. However, superposition of proteins is difficult when large structural changes occur upon ligand binding. Therefore, we instead used a method that compares residues around a ligand-binding site that is generally applicable to ligand-unbound structures. Assessment of the results of ligand-binding residue prediction The assessment method of the results of ligand-binding residue prediction employed in the present study, which is similar to the assessment score calculation used in the ligand-binding site prediction category [16] of the Critical Assessment of Structure Prediction (CASP, http://www.predictioncenter.org/), is as follows:Residues located within 5 Å of each probe in a cluster are regarded as the ligand-binding residues of the cluster. Calculate the following values: the number of residues predicted correctly as ligand-binding residues (true positive), the number of residues correctly predicted as nonligand-binding residues (true negative), the number of residues incorrectly predicted as ligand-binding residues (false positive), and the number of residues incorrectly predicted as nonligand-binding residues (false negative). Finally, calculate scores, Sresidue, defined as the proportion of correctly predicted ligand-binding residues among all the positively predicted ligand-binding residues and Matthew’s correlation coefficient (MCC) based on the above values. Information about residues around ligands (i.e., correct ligand-binding residues) was obtained from LigASite. In addition to MCC, we used Sresidue as an assessment score to indicate the proportion of correctly predicted residues around ligands. Parameter selection using cross-validation In ligand-binding space prediction, fivefold cross-validation was performed using the 382 ligand-bound structures; these structures were randomly divided into five subsets of equal size, and four of them were used for training parameters as described in the Results section and then one was used for testing. This process is repeated five times, each time using a different subset for testing and the other four subsets for training. The performance was assessed as the average of each success rate and average precision. Ligand-binding residue prediction can be used to assess the results of prediction of both ligand-bound and ligand-unbound structures. Therefore, fivefold cross-validation was first performed using ligand-bound structures, and the parameters showing the best performance were further used to predict residues for ligand-unbound structures. The procedure was evaluated with respect to the successful prediction of correct ligand-binding sites using ligand-unbound proteins. Results The parameters for the proposed method were the maximum number of probes within a cluster (Pnum) and the weights for the amino acid conservation (w1, w2, w3) in the probe energy [Eq. (1)]. In ligand-binding space prediction, the optimum value Pnum was searched first at intervals of 100 in the range 100–600, and fivefold cross-validation was performed for the 382 ligand-bound structures. Pnum = 500 was selected because it yielded the highest success rate. Next, with Pnum = 500, the weights (w1, w2, w3) were determined by fivefold cross-validation, and the mean values of the highest success rate and the corresponding mean precision in each cross-validation were calculated. Among the five parameter sets with a highest success rate at each test step in the cross-validation, the most frequently observed combination (w1, w2, w3) = (1, −0.05, 9) was used as the optimal weights. Table 1 shows the results of ligand-binding space prediction. Excluding proteins containing over 10,000 atoms, which are not accepted by the Q-SiteFinder server, 342 ligand-binding sites were used for the assessment as their ligand-bound and ligand-unbound structures were both available. The predicted clusters are ranked according to the score defined in Eq. (1). The method denoted as “Our method (without Cscore)” uses the score with w2 and w3 zero. In Table 1, “Top1” indicates the best cluster per structure and “Top3” indicates a cluster in the three best clusters per structure.Table 1 Comparison of prediction results from our proposed method and Q-SiteFinder (for ligand-binding space prediction, 342 ligand-bound structures) Top1, precision ≥25 % Top1, precision >0 % Top3, precision ≥25 % Ratio (%) Average precision (%) Ratio (%) Average precision (%) Ratio (%) Average precision (%) Our method 74.0 64.4 80.1 60.5 88.0 66.4 Our method (without C score) 57.0 59.7 66.1 53.5 73.7 62.7 Q-SiteFinder 52.0 66.0 58.5 60.2 71.9 66.9 The results of “our method” are highlightened in bold “Precision” indicates the proportion overlapping with ligands, and the overlap indicates that a probe in a cluster is located within 1.6 Å from one of the ligand heavy atoms. The predicted clusters are ranked according to the score defined in Eq. (1). The condition “Top1, precision ≥25 %” indicates the best cluster with precision ≥25 %, “Top1, precision >0 %” indicates the best cluster with precision >0 %, and “Top3, precision ≥25 %” indicates that at least one cluster in the best three clusters has precision ≥25 %. “Ratio” is the number of proteins with the specified condition divided by the number of proteins in the dataset. “Average precision” is the average precision of the proteins with the specified condition With regard to the ligand-binding space prediction, the success rate was confirmed to be higher than that of Q-SiteFinder under each criterion of success. More specifically, the success rate for the clusters in the first prediction rank with precision ≥25 % increased from 52.0 to 74.0 % and that for the clusters within the first three prediction ranks with precision ≥25 % increased from 71.9 to 88.0 %. With regard to average precision, sometimes Q-SiteFinder was superior depending on the criteria of success; however, even in those cases, the values from the two methods were similar. When our method with amino acid conservation [Cscore in Eq. (1)] was compared with that without conservation, the success rate for the cluster in the first prediction rank with precision ≥25 % increased from 57.0 to 74.0 % and that for the clusters within the first three prediction ranks with precision ≥25 % increased from 73.7 to 88.0 %. These results show that the effect of using amino acid conservation was pronounced. Table 2 shows the results of ligand-binding residue prediction. Here 348 ligand-unbound structures were used for the assessment as their ligand-bound and ligand-unbound structures were both available. As in ligand-binding space prediction, the predicted clusters are ranked according to the score defined in Eq. (1). The method denoted as “Our method (without Cscore)” uses the score with w2 and w3 zero. With regard to ligand-binding residue prediction, success rate, average precision, and average MCC were confirmed to be higher than those of Q-SiteFinder under all the criteria of success.Table 2 Comparison of prediction results from our proposed method and Q-SiteFinder (for ligand-binding residue prediction, 348 ligand-unbound structures) Top1, S residue ≥25 % Top1, S residue ≥0 % Top3, S residue ≥25 % Top1 Ratio (%) Average S residue (%) Ratio (%) Average S residue (%) Ratio (%) Average S residue (%) Average MCC Our method 73.9 61.3 86.2 54.3 85.6 61.6 0.51 Our method (without C score) 56.3 58.2 74.7 47.0 76.1 58.6 0.39 Q-SiteFinder 53.4 60.4 69.3 49.2 74.4 60.1 0.33 The results of “our method” are highlightened in bold For ligand-binding residue prediction, S residue is used as the measure of precision. The condition “Top1, S residue ≥25 %” indicates the best cluster with S residue ≥25 %, “Top1, S residue >0 %” indicates the best cluster with S residue >0 %, and “Top3, S residue ≥25 %” indicates that at least one cluster in the best three clusters has the S residue ≥25 %. “Ratio” is the number of proteins with the specified condition divided by the number of proteins in the dataset. “Average precision” is the average precision of the proteins with the specified condition More specifically, the success rate for the cluster in the first prediction rank with Sresidue ≥ 25 % increased from 56.3 to 73.9 % and that for the clusters within the first three prediction ranks with Sresidue ≥ 25 % increased from 74.4 to 85.6 %. Average Sresidue was found to be greater than Q-SiteFinder for all criteria. When our methods with amino acid conservation were compared with that without conservation, the success rate for the cluster in the first prediction rank with Sresidue ≥ 25 % was improved from 53.4 to 73.9 % and that for the clusters within the first three prediction ranks with Sresidue ≥ 25 % increased from 76.1 to 85.6 %. These results show that, again, the effect of using amino acid conservation was pronounced. Figures 2 and 3 show examples of prediction results from both our proposed method and Q-SiteFinder.Fig. 2 Comparison of prediction results from our proposed method and Q-SiteFinder (for ligand-binding space prediction). From left, crystal structures of protein–ligand complex, prediction results from the proposed method, and prediction results from Q-SiteFinder. Green proteins; gray ligands; orange clusters in the first prediction rank; and blue clusters in the lower than second prediction ranks. a Prediction results of PDBID: 1mka (ligand: 2-decenal n-acetyl cysteine). Precision of the proposed method: 93.2 %, precision of Q-SiteFinder: 86.3 %. Ligand-binding site was predicted with higher precision by the proposed method. b Prediction results of PDBID: 1fcv (n-acetyl-D-glucosamine). Precision of the proposed method: 59.0 %, precision of Q-SiteFinder (blue): 100 %. The cluster correctly predicted by Q-SiteFinder was in the sixth prediction rank, whereas that by our proposed method was in the first prediction rank. c Prediction results of PDBID: 2wn7 (ligand: nicotinamide adenine dinucleotide). Precision of the proposed method (blue): 38.3 % and precision of Q-SiteFinder: 34.7 %. The cluster correctly predicted by our proposed method was in the second prediction rank Fig. 3 Comparison of prediction results from our proposed method and Q-SiteFinder (for ligand-binding residue prediction). From left, crystal structures of protein–ligand complex, prediction results from the proposed method, and prediction results from Q-SiteFinder. Orange cluster predicted in the first prediction rank, green correctly predicted residues (true positive, TP), red incorrectly predicted residues (false positive, FP), and yellow unpredicted residues (false negative, FP). a Prediction results of PDBID: 1y2q. Proposed method: S residue, 93.3 % and MCC, 0.701. Q-SiteFinder: S residue, 86.3 % and MCC, 0.859. The results show that many residues were not predicted by the proposed method, whereas Q-SiteFinder incorrectly predicted many residues. b Prediction results of PDBID: 1cwy. Proposed method: S residue, 68.2 % and MCC, 0.667. Q-SiteFinder: S residue, 10.0 % and MCC, 0.078. Q-SiteFinder formed a cluster deviating to the left of the correct binding site. c Prediction results of PDBID: 2b78. Proposed method: S residue, 15.4 % and MCC, 0.096. Q-SiteFinder: S residue, 61.3 % and MCC, 0.662. Our proposed method formed a cluster deviating to the left of the correct binding site. MCC, Matthew’s correlation coefficient Discussion The main differences between the proposed method and Q-SiteFinder include force field parameters, energy threshold of clustering, and the use of amino acid conservation. With respect to the force field parameter, for ligand-binding site prediction based on interaction energy calculation, we found that prediction with AMBER parm94 yields a higher success rate than GRUB [17], employed by Q-SiteFinder [18]. The energy threshold of clustering is an important parameter governing the size of a cluster. If it is too low, clusters are unlikely to be formed and sensitivity will be degraded; conversely, if it is too high, more probes with small interaction energies can be formed. Q-SiteFinder uses 1.4 kcal/mol as the fixed energy threshold. However, the interaction energy between a protein and a probe should be different for each protein. Therefore, we instead used the upper limit of the number of probes in a cluster, Pnum, to restrict the size of the cluster. Of course, this solution is not perfect: for proteins that bind small ligands, our method tends to generate clusters larger than the size of the ligands. This tendency affects the precision. As seen in Tables 1 and 2, the average precision and average Sresidue of our method without the amino acid conservation score (Cscore) were lower than those of Q-SiteFinder. This is ascribed to the use of the fixed Pnum value larger than the size of the small ligands. In Fig. 4, the mean values of amino acid conservation (Cscore) around ligand-binding sites are plotted against the mean value of amino acid conservation in entire protein sequences. In proteins with multiple ligand-binding sites, residues around all binding sites were treated as ligand-binding residues. Points above the lines indicate proteins for which the mean values of amino acid conservation around ligand-binding sites are greater than those of the entire sequence. In contrast, points below lines indicate proteins for which mean values of amino acid conservation around ligand-binding sites are smaller than those of the complete sequence. This figure shows that in both ligand-bound proteins (Fig. 4a) and ligand-unbound proteins (Fig. 4b), there are more proteins lying above than below the lines, irrespective of the feasibility of prediction, that is, most of the proteins used for evaluation in the present study showed higher amino acid conservation around ligand-binding sites than in entire sequences. The numbers of such proteins were 309 (=244 + 65) for ligand-bound structures and 315 (=247 + 68) for ligand-unbound structures.Fig. 4 Correlation of the mean values of amino acid conservation (C score) around ligand-binding sites with the mean value of amino acid conservation in entire protein sequences. Blue points represent successfully predicted proteins, whereas red points represent proteins for which prediction failed. Lines correspond to equal mean values of amino acid conservation between the entire sequence and around ligand-binding sites. a Prediction results for 342 ligand-bound structures (for ligand-binding space prediction). b Prediction results for 348 ligand-unbound structures (for ligand-binding residue prediction) Comparison of the prediction results of our proposed method and Q-SiteFinder indicates that several red points lying above the lines in Q-SiteFinder graphs have changed to blue points in the graphs produced by our proposed method. In particular, when the difference between amino acid conservation between entire sequences and ligand-binding sites is high (proteins located further from the line in the upper left graph), more proteins were successfully predicted in the present study. These results show that our proposed prediction method is likely to succeed for proteins in which the amino acids around the ligand-binding sites are highly conserved. However, these results do not indicate that higher amino acid conservation around the ligand-binding site will always result in higher prediction precision. The optimal weights, which determine the contribution of the van der Waals energy term (Evan) and the amino acid conservation term (Cscore), depend on the training dataset but the results show that the contribution of van der Waals energy is larger than the amino acid conservation; the amino acid conservation alone does not yield a better result than the van der Waals energy alone. Proteins for which prediction failed Figure 5 indicates the number of proteins for which prediction by our proposed method, Q-SiteFinder, or both methods failed. Both the proposed method and Q-SiteFinder failed to predict 62 ligand-bound and 63 ligand-unbound structures. In these proteins, the prediction of ligand-binding site based on interaction energy calculation was indicated to be difficult. In addition, prediction of almost 30 proteins of either type failed only when the proposed method was used.Fig. 5 Numbers of proteins for which prediction by Q-SiteFinder and our proposed method failed. (Left) Prediction results for 342 ligand-bound structures (for ligand-binding space prediction). (Right) Prediction results for 348 ligand-unbound structures (for ligand-binding residue prediction). For both methods, the numbers of failed proteins are similar for ligand-bound and ligand-unbound proteins Prediction failure can be attributed to two major causes. First, prediction failed when there was no ligand at the predicted site, but one of its homologue proteins has a ligand at the corresponding site (i.e., the predicted site was one of the potential ligand-binding sites). Among failed structures, 19 ligand-bound structures and 23 ligand-unbound structures were categorized as this type. Accordingly, 5–7 % of proteins in the dataset for which prediction of our method failed were practically successful. Second, the conservation of ligand-binding sites was low, and another site with higher conservation was incorrectly predicted. Among failed structures, 11 proteins were categorized as this type. Summary Our method predicts ligand-binding sites using van der Waals energy and amino acid conservation calculated from alignment with homologous sequences. In a comparison of our proposed method with Q-SiteFinder, a ligand-binding site prediction method based on interaction energy calculation, a much higher success rate (proportion of successfully predicted proteins) was obtained by our method than that by Q-SiteFinder. The present study indicates that amino acid conservation is an important factor in the success of ligand-binding site prediction, and that its combination with interaction energy calculation enables more precise site prediction. One of the binding sites of proteins with multiple ligand-binding sites was correctly predicted; however, applying the proposed method to a protein with less-conserved ligand-binding sites sometimes resulted in failure. This result suggests that in some cases, when the conservation of ligand-binding sites is low, the weighting of amino acid conservation can result in prediction failure. In this study, we did not explicitly filter out the suspicious ligands. The selection of ligands using the results of B-factor and LLDF described in Supplementary Material would be a future work. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary material 1 (DOCX 53 kb) Abbreviations PDBProtein data bank PSSMPosition-specific scoring matrix SBDDStructure-based drug design 3DThree-dimensional This research is supported by the Platform Project for Supporting in Drug Discovery and Life Science Research from Japan Agency for Medical Research and Development (AMED). ==== Refs References 1. Kuntz ID Blaney JM Oatley SJ Langridge R Ferrin TE A geometric approach to macromolecule-ligand interactions J Mol Biol 1982 161 269 288 10.1016/0022-2836(82)90153-X 7154081 2. Levitt DG Banaszak LJ POCKET: a computer graphics method for identifying and displaying protein cavities and their surrounding amino acids J Mol Graph 1992 10 229 234 10.1016/0263-7855(92)80074-N 1476996 3. Hendlich M Rippmann F Barnickel G LIGSITE: automatic and efficient detection of potential small molecule-binding sites in proteins J Mol Graph Model 1997 15 359 363 10.1016/S1093-3263(98)00002-3 9704298 4. Huang B Schroeder M LIGSITEcsc: predicting ligand binding sites using the Connolly surface and degree of conservation BMC Struct Biol 2006 16995956 5. Goodford PJ A computational procedure for determining energetically favorable binding sites on biologically important macromolecules J Med Chem 1985 28 849 857 10.1021/jm00145a002 3892003 6. Laurie ATR Jackson RM Q-SiteFinder: an energy-based method for the prediction of protein-ligand binding sites Bioinformatics 2005 21 1908 1916 10.1093/bioinformatics/bti315 15701681 7. Kulharia M Bridgett SJ Goody RS Jackson RM InCa-SiteFinder: a method for structure-based prediction of inositol and carbohydrate binding sites on proteins J Mol Graph Model 2009 28 297 303 10.1016/j.jmgm.2009.08.009 19762259 8. Magliery TJ Regan L Sequence variation in ligand binding sites in proteins BMC Bioinform 2005 9. Capra JA Laskowski RA Thornton JM Singh M Funkhouser TA Predicting protein ligand binding sites by combining evolutionary sequence conservation and 3D structure PLoS Comput Biol 2009 19997483 10. Dessailly BH Lensink MF Orengo CA Wodak SJ LigASite–a database of biologically relevant binding sites in proteins with known apo-structures Nucleic Acids Res 2008 36 D667 D673 10.1093/nar/gkm839 17933762 11. Krissinel E Henrick K Inference of macromolecular assemblies from crystalline state J Mol Biol 2007 372 774 797 10.1016/j.jmb.2007.05.022 17681537 12. Canutescu AA Shelenkov AA Dunbrack RL Jr A graph-theory algorithm for rapid protein side-chain prediction Protein Sci 2003 12 2001 2014 10.1110/ps.03154503 12930999 13. Cornell WD Cieplak P Bayly CI Gould IR Merz KM Ferguson DM Jr Spellmeyer DC Fox T Caldwell JW Kollman PA A second generation force field for the simulation of proteins, nucleic acids, and organic molecules J Am Chem Soc 1995 117 5179 5197 10.1021/ja00124a002 14. Wang K Samudrala R Incorporating background frequency improves entropy-based residue conservation measures BMC Bioinform 2006 15. Capra J Singh M Predicting functionally important residues from sequence conservation Bioinformatics 2007 23 1875 1882 10.1093/bioinformatics/btm270 17519246 16. Schmidt T Haas J Gallo Cassarino T Schwede T Assessment of ligand-binding residue predictions in CASP9 Proteins 2011 79 126 136 10.1002/prot.23174 21987472 17. GRID (2016) http://www.moldiscovery.com/software/grid/. Accessed 21 Jan 2016 18. Morita M Nakamura S Shimizu K Highly accurate method for ligand-binding site prediction in unbound state (apo) protein structures Proteins 2008 73 468 479 10.1002/prot.22067 18452211

PMC005xxxxxx/PMC5002283.txt

==== Front J Comput Neurosci J Comput Neurosci Journal of Computational Neuroscience 0929-5313 1573-6873 Springer US New York 615 10.1007/s10827-016-0615-7 Article Interaction between synaptic inhibition and glial-potassium dynamics leads to diverse seizure transition modes in biophysical models of human focal seizures Y. Ho E. C. ernest_ho@brown.edu 12 Truccolo Wilson wilson_truccolo@brown.edu 12 1 Department of Neuroscience & Institute for Brain Science, Brown University, Providence, RI USA 2 U.S. Department of Veterans Affairs, Center for Neurorestoration and Neurotechnology, Providence, RI USA Action Editor: Steven J. Schiff 3 8 2016 3 8 2016 2016 41 2 225244 7 9 2015 18 6 2016 6 7 2016 © The Author(s) 2016 How focal seizures initiate and evolve in human neocortex remains a fundamental problem in neuroscience. Here, we use biophysical neuronal network models of neocortical patches to study how the interaction between inhibition and extracellular potassium ([K+]o) dynamics may contribute to different types of focal seizures. Three main types of propagated focal seizures observed in recent intracortical microelectrode recordings in humans were modelled: seizures characterized by sustained (∼30−60 Hz) gamma local field potential (LFP) oscillations; seizures where the onset in the propagated site consisted of LFP spikes that later evolved into rhythmic (∼2−3 Hz) spike-wave complexes (SWCs); and seizures where a brief stage of low-amplitude fast-oscillation (∼10−20 Hz) LFPs preceded the SWC activity. Our findings are fourfold: (1) The interaction between elevated [K+]o (due to abnormal potassium buffering by glial cells) and the strength of synaptic inhibition plays a predominant role in shaping these three types of seizures. (2) Strengthening of inhibition leads to the onset of sustained narrowband gamma seizures. (3) Transition into SWC seizures is obtained either by the weakening of inhibitory synapses, or by a transient strengthening followed by an inhibitory breakdown (e.g. GABA depletion). This reduction or breakdown of inhibition among fast-spiking (FS) inhibitory interneurons increases their spiking activity and leads them eventually into depolarization block. Ictal spike-wave discharges in the model are then sustained solely by pyramidal neurons. (4) FS cell dynamics are also critical for seizures where the evolution into SWC activity is preceded by low-amplitude fast oscillations. Different levels of elevated [K+]o were important for transitions into and maintenance of sustained gamma oscillations and SWC discharges. Overall, our modelling study predicts that the interaction between inhibitory interneurons and [K+]o glial buffering under abnormal conditions may explain different types of ictal transitions and dynamics during propagated seizures in human focal epilepsy. Keywords Focal epilepsy Seizure dynamics Spike-wave discharges Gamma oscillations http://dx.doi.org/10.13039/100000065 National Institute of Neurological Disorders and Stroke (US) R01NS079533 Y. Ho E. C. http://dx.doi.org/10.13039/100000738 U.S. Department of Veterans Affairs (US) RX000668-01A2 Y. Ho E. C. issue-copyright-statement© Springer Science+Business Media New York 2016 ==== Body Introduction Epilepsy is a neurological disorder characterized by recurrent seizures affecting an estimated 65 million people worldwide (Thurman et al. 2011). Its apparent simple characterization belies the often complex and varied cellular and synaptic basis underlying the abnormal neuronal activity during epileptic seizures. As a result, despite significant advances in understanding epilepsy over the past several decades, key knowledge gaps remain with respect to the disorder. One example is the mechanism(s) concerning the transition of neural dynamics from normal to ictal as focal seizures propagate to distal regions in the brain, especially in neocortex. Seizures have been hypothesized to originate from a wide variety of cellular and network mechanisms (Richardson et al. 2008). It is possible that, even in the same patient, different seizures might be triggered by different factors. Facing the diversity of possible triggers of seizure transition, recent theoretical studies have proposed canonical models that focus on general unifying dynamical principles (Kramer et al. 2012; Jirsa et al. 2014; Sritharan and Sarma 2014; Wang et al. 2014; Wei et al. 2014b; Naze et al. 2015). Recent animal experiments (Trevelyan et al. 2006; Zhang et al. 2012; Cammarota et al. 2013; žiburkus et al. 2013) and human studies (Truccolo et al. 2011; Schevon et al. 2012; Ahmed et al. 2014) have also suggested that complex dynamics between different types of neurons can be in play as cortical networks transition into seizures. In particular, these studies have indicated the potential critical role of fast-spiking (FS) inhibitory interneurons (see Paz and Huguenard 2015 for a review) and their interactions with changes in extracellular potassium concentration [K+]o (e.g. žiburkus et al. 2006). Elucidating such interplay between interneuron activity and ionic concentrations during seizure initiation and propagation can provide the much-needed clues for the development of seizure prediction, early detection and novel therapeutic approaches. Furthermore, studies by Truccolo et al. (2011, 2014) and Wagner et al. (2015) have characterized spiking in ensembles of single neurons and local field potential (LFP) activity recorded via 96-microelectrode arrays in humans during focal seizures. Three main types of seizure activity were recorded in neocortical patches distal (∼2−4 cm) to the identified seizure onset zones: seizures consisting of sustained (∼30−60 Hz) gamma-band oscillations, seizures where the onset consisted of a direct transition into LFP spike discharges which evolved later into (∼2−3 Hz) spike-wave complexes (SWCs), and seizures where the onset included a brief transient of low voltage fast oscillations (10−20 Hz) before the evolution into SWC discharges. (For a more comprehensive classification of seizures in terms of onset patterns and relationships to pathology in larger datasets, see Perucca et al. 2014). In the gamma seizures, despite the narrowband nature of the oscillations, neuronal spiking activity remained highly irregular and asynchronous (Truccolo et al. 2014). Classification of recorded single neurons into putative principal and interneuron cells suggested that inhibition was preserved throughout the various ictal stages in these gamma seizures (Truccolo et al. 2011). On the other hand, a recent investigation of SWC seizures in these datasets indicated that inhibitory activity seems to shut off just before the emergence of large amplitude spike-wave discharges (Ahmed et al. 2014). Ahmed et al. (2014) hypothesized the role of depolarization block in interneurons and examined their sensitivity to [K+]o in a single cell model of a FS inhibitory interneuron. Here, we examine how the interaction between FS inhibitory interneuron activity and extracellular potassium concentration in biophysical neuronal network models can capture the different types of propagated seizure dynamics observed in these human neocortical recordings. We used neuronal network models of neocortical patches that include pyramidal and FS interneurons (Wang and Buzsáki 1996), and [K+]o diffusion coupled with glial cell activity. The neuronal network models spanned 9×9 minicolumns, each 25 μm wide and containing 12 pyramidal neurons and 4 FS interneurons (for a total of 9×9×16=1296 model neurons). Coupled [K+]o and glial buffering dynamics were adapted from previous models in Fröhlich et al. (2010). Several previous computational models (Kager at al. 2000; Bazhenov et al. 2004; Traub et al. 2005; Anderson et al. 2007; Cressman Jr. et al. 2009; Ullah et al. 2009; Fröhlich et al. 2010; Krishnan and Bazhenov, 2011; Krishnan et al. 2013, Wei et al. 2014a, 2014b) have explored the role of intracellular and extracellular concentrations of various types of ions (e.g. potassium, sodium, chloride, calcium, etc.), as well as inhibitory/excitatory conductances and metabolic factors, on seizure activity. Our model specification was inspired by several of these previous studies, including, for example, the formulation of glial-buffering processes (see Section 2). Nevertheless, we emphasize that, to our knowledge, the work presented here is the first to examine how the interaction between inhibitory activity and abnormal levels of extracellular potassium concentration can lead to three different types of seizures (gamma, SWC, low-voltage fast-oscillations followed by SWCs) observed in intracortical recordings from patients with focal epilepsy. Overall, we demonstrate that transition into and maintenance of both gamma-band and SWC seizures in the model can be explained by variations in inhibition strength (and therefore changes in the excitation/inhibition balance) and [K+]o levels. We consider this to be the main novel contribution reported in our study. Methods Model description and parameter choices Single neuron model We used a single-compartment conductance based model to represent both the fast-spiking inhibitory interneurons and pyramidal cells (Wang and Buzsáki 1996; Skinner 2006; Anderson et al. 2007). The differential equation for the membrane potential of a given neuron corresponds to: 1 C×dVldt=Iintl(t)+Isynl(t)+Iextl(t), for l=1,2,…,1296 neurons. Iintl denotes the intrinsic current of the neuron in question, Isynl denotes synaptic currents and Iextldenotes external input. The specific capacitance C is assumed to be 1 μF/cm2. Iintl has the following general form: 2 Iintl=−INa−IK−IL−IKCa=−gNa×m∞3×h×(V−ENa)−gK×n4×(V−EK([K+]o))−gL×(V−EL([K+]o))−IKCa, which represent sodium, delayed-rectifier potassium, leak and calcium-mediated potassium currents, respectively. The leak reversal potential is dependent on various ionic concentrations via the Goldman-Hodgkin-Katz voltage equation: 3 EL([K+]o)=26.64mV×loge1×[K+]o+0.085×[Na+]o+0.1×[Cl−]i1×[K+]i+0.085×[Na+]i+0.1×[Cl−]o, where [K+]o,[K+]i,[Na+]o,[Na+]i,[Cl−]o,[Cl−]i denote ionic concentrations, with subscript i and o representing intracellular and extracellular, respectively. The coefficient before each type of ionic concentration is the relative membrane permeability of that ion. The “fast” activation variable m∞ of the sodium current is assumed to have reached a steady state and has the form: 4 m∞=αmαm+βm,αm(V)=−0.1mV−1×(V+35mV)exp(−0.1mV−1×(V+35mV))−1,βm(V)=4×exp−V+60mV18mV, and the inactivation variable h of the sodium and the activation variable n of delayed-rectifier potassium current obey the following differential equation: 5 dXdt=5ms−1×(αX×(1−X)−βX×X), where X can be h or n, with 6 αh(V)=0.07×exp−V+58mV20mV,βh(V)=1exp(−0.1mV−1×(V+28mV))+1,αn(V)=−0.01mV−1×(V+34mV)exp(−0.1mV−1×(V+34mV))−1,βn(V)=0.125×exp−V+44mV80mV. The purpose of the calcium-mediated potassium current (IKCa) is to provide the spike-frequency adaptation of the model pyramidal cells. It has the following form: 7 IKCa=gKCa×(V−EK)×[Ca2+]i1mM+[Ca2+]i, where gKCa and [Ca2+]i denote the conductance for calcium-mediated potassium currents and the intracellular calcium concentration, respectively. The [Ca2+]i dynamics are given according to Cressman et al. (2009) and Ullah et al. (2009): 8 d[Ca2+]idt=−0.002molcm⋅mC×gCa×(V−ECa)1+exp−(V+25mV)2.5mV−[Ca2+]i80ms, where gCa and ECa denote the conductance and reversal potential related to Ca2+, respectively. Network structure and connectivity The structure of our network of model neurons is based on the arrangement of neurons in minicolumns in the neocortical area (Buxhoeveden and Casanova 2002). In this network model, each minicolumn consists of 16 single-compartment model neurons, of which 12 are excitatory pyramidal cells and 4 are inhibitory interneurons. The neurons are arranged into four layers within each minicolumn (Fig. 1). Each layer consists of four neurons in a square arrangement, with each vertex of the square being the location of one neuron. Pyramidal cells occupy the top three layer and the bottom layer is for the four inhibitory interneurons. (We do not attempt here to replicate the actual 6-layer neocortical structure). The entire cortical network consists of 9 ×9 such minicolumns for a total of 1296 model neurons. Details of the structure configuration of the cortical network model and various distance parameters between model neurons are given in Fig. 1 and Table 1. Unless otherwise stated, we chose a global random connectivity to avoid boundary effects associated with the small size of the simulated cortical network. Fig. 1 Model cortical network: spatial configuration. Open circles represent pyramidal cells and black circles represent inhibitory interneurons. Two minicolumns are shown. Table 1 specifies the values for the distance parameters in the figure Table 1 Network geometry parameters. Please refer to Fig. 1 for the description of parameters Parameter Units Values Δd ee1 cm 5×10−4 Δd ee2 cm 5×10−4 Δd ie2 cm 30×10−4 Δd ii1 cm 5×10−4 Δd mm cm 15×10−4 Glial and potassium ionic dynamics Each neuron in the network is endowed with a potassium pump. An associated potassium uptake dynamics related to nearby glial cells are included in the model to regulate the extracellular potassium concentration [K+]o. The potassium pump is represented by: 9 IKpump=IKmax1+[K+]o(eq)[K+]o2, where IKmax and [K+]o(eq) are defined in Table 3. The glial potassium uptake mechanism is modelled in terms of a reversible binding process between extracellular potassium ions and the glial buffer (Kager et al. 2000; Bazhenov et al. 2004; Park and Durand 2006; Fröhlich et al. 2008, 2010). The binding/unbinding process can be represented as 10 K++B⇌kbkfKB, where K+ and B denote the extracellular potassium ions and the free buffer in the glia, respectively, and kb and kf correspond to the backward and forward glial binding rate, respectively (Tables 3 and 4). K+ and B combine reversibly to form the bound form of the buffer KB. We fixed the maximal amount of free glial buffer [B]max to 500mM (Table 3, see also Kager et al. 2000). The above process (10) can be written as a set of differential equations: 11 ∂[K+]o∂t|glia=kb×([B]max−[B])1.1−kf×[K+]o×[B],d[B]dt=kb×([B]max−[B])−kf×[K+]o×[B],kf=kb/1mM1+exp[K+]o−[K+]o(th)𝜃, where [K+]o(th) and 𝜃 are specified in Table 3. We vary [K+]o(th)and 𝜃 to simulate a range of physiological and pathological states for the potassium glial-buffering system. [K+]o(th) corresponds to the equilibrium level (set point) for the extracellular potassium concentration being targeted by the glial buffering system. How tightly the glial buffering system tracks this equilibrium set-point depends on the parameter 𝜃. For example, the lower the absolute values of 𝜃, the closer the glial buffer system attempts to keep the [K+]o concentration to the target [K+]o(th) level. Under this condition, the system may fail if there is a big enough difference between [K+]o and [K+]o(th), leading to bi-stability and abnormal functioning. In more physiological conditions, the absolute value of 𝜃 is kept at higher values so that the glial buffering system tracks [K+]o(th) in a softer fashion, allowing a broader range of [K+]o levels and more effectively avoiding bi-stability. We note that while we vary [K+]o(th) and 𝜃 in order to change the glial buffering dynamics and to induce abnormal states, we do not yet have a hypothesis about specific causes that would lead to these changes in first place. The diffusion of extracellular potassium ions across the network was modelled according to the diffusion equation: 12 ∂[K+]o∂tdiff=D×∇2[K+]o, where ∇2 is the Laplace operator ∂2∂x2+∂2∂y2+∂2∂z2 and D is the diffusion constant. In the simulations, ∇2 is approximated by a set of finite difference equations, since the model neurons are discretely located (Fig. 1). We impose the “no-flux” or Neumann boundary condition (i.e. ∂[K+]o∂x=0 on the y−z planes at both the x boundaries and so on) in the network simulations. Therefore, the total time derivative of [K+]o of a particular model neuron, d[K+]odt (unit in mM ⋅ms−1), is the sum of all the potassium, pump, glia and diffusion currents 13 d[K+]odt=5096489molcm⋅mC×(IKCa+IK−IKpump+gL|[K+]×(V−EK([K+]o)))+∂[K+]o∂tglia+∂[K+]o∂tdiff, where gL|[K+] (≥0) is the value of the leak conductance attributable to potassium. The value can be estimated by fitting the equation, gL×(V−EL([K+]o))=gL|[K+]×(V−EK([K+]o))+gL|[Cl−]×(V−ECl)+gL|[Na+]×(V−ENa), with the constraint that 14 gL=gL|[K+]+gL|[Cl−]+gL|[Na+], where gL|[Cl−] (≥0) and gL|[Na+] (≥0) are the values of leak conductance due to Cl− and Na+ respectively, while ECl and ENa are their respective reversal potentials as shown in equations (16) and (17). gL|[K+] values at ∼ 6–13 % of total leak give reasonable fits over a wide range of physiological [K+]o and membrane potential values. (Various fitting scenarios indicate that >70 % of the total leak conductance value is due to chloride). Based on the above procedure, we assume a gL|[K+] value at 6 % of the gL value. The potassium reversal potential EK is affected by [K+]o via the Nernst equation: 15 EK=26.64mV×loge[K+]o[K+]i, where the intracellular potassium concentration [K+]i is set at 133 mM (Table 3). Similarly, ECl is given by 16 ECl=26.64mV×loge[Cl−]i[Cl−]o∼−74mV, and ENa by 17 ENa=26.64mV×loge[Na+]o[Na+]i∼54mV. See also Table 3 for values of [Na+]o, [Na+]i, [Cl−]o and [Cl−]i. Synaptic model Unless otherwise stated, the synaptic current Isynl corresponds to Isynl=∑m≠lΓm→l×gsynm→l×sm(t−τdelay)×(E{e,i}−Vl), with s⋅(t) representing the synaptic gating variable of each model neuron, such that 18 sl→sl+smax(whenVlcrosses 0 from below),dsldt=−slτ{e,i}(otherwise). In the above, Γm→l denotes the binary connectivity matrix element in which a value of 1 represents the existence of a synaptic connection in neuron l from neuron m, and a value 0 represents the absence of such connection. Table 5 lists the probabilities of connection between pyramidal neurons (P(e→e)), from pyramidal neurons to interneurons (P(e→i)), from interneurons to pyramidal neurons (P(i→e)) and between interneurons (P(i→i)). The term E{e,i} denotes the excitatory or inhibitory reversal potential value, respectively. (Whether the reversal potential is excitatory or inhibitory depends on the particular neuron–i.e. neuron m–from which neuron l receives synaptic current). As done in Ho et al. (2012) for computational efficiency, we use a discontinuous model to represent the opening of the synapses (equation set (18)) whenever there is a spike. Parameters τ{e,i} and smax (for sl) denote the excitatory (inhibitory) synaptic decay time constant and the value of the maximal opening of the synaptic gates per spike, respectively (whether the decay time constant should be excitatory or inhibitory depends on whether neuron l itself is a pyramidal cell or an inhibitory interneuron). Table 5 provides more details on the terms of the above equations. Background synaptic activity Unless otherwise stated, each model neuron is driven by an external input Iext with the form (Rudolph et al. 2004) to mimic the activities of the fluctuating synaptic background: 19 Iextl=gel(t)×(Ee−Vl)+gil(t)×(Ei−Vl),dg{e,i}l(t)dt=g{e,i}0−g{e,i}l(t)τ{e,i}+2×σ{e,i}2τ{e,i}×χ{e,i}l(t), where E{e,i} denotes the excitatory/inhibitory reversal potential and τ{e,i} is the excitatory/inhibitory time constant. The excitatory/inhibitory conductance values g{e,i} are stochastic variables following an Ornstein-Uhlenbeck process (equation set (19)), with mean g{e,i}0 and SD σ{e,i} (excitatory/inhibitory fluctuations). The stochastic term χ{e,i} is such that ∫0𝜖χ{e,i}(s)×ds is a Gaussian distributed random variable with mean zero and variance 𝜖>0. Further details on the numerical implementation of the stochastic elements of equation set (19) can be found in Ho et al. (2012). Table 5 provides more details on the terms in the above equation. Implementing the simulations The code for the model cortical network simulation was written in C++ (gcc, version 4.4.7, http://gcc.gnu.org). We used open MPI (version 1.8, http://www.open-mpi.org) for parallel processing of each simulation. Each typical 1296-neuron simulation used eight CPUs. A 10-hour simulation on the Brown University computer cluster (https://www.ccv.brown.edu/) usually yielded ∼ 2-3 minutes of data. Parameter scanning was automated via the use of custom Perl (version 5.18.4) and Linux shell scripts. Post simulation analysis of data was carried out using Matlab (version R2013b, http://www.mathworks.com), Octave (version 3.8.1, http://www.gnu.org/software/octave/) and customized Perl scripts. Single neuron simulations (Fig. 2) were performed using xppaut (Ermentrout 2002). Graphical rendering of data was performed by gnuplot (version 4.6.3, http://www.gnuplot.info), xfig (version 3.2) and gimp (version 2.8.10, http://www.gimp.org). The parula palette of the time-frequency plots is from https://github.com/Gnuplotting/gnuplot-palettes. The C++ code is available in the NEURON ModelDB database (https://senselab.med.yale.edu/ModelDB/showModel.cshtml?model=190306). The ModelDB accession number for the network model reported in this paper is 190306. Fig. 2 Four single neuron model simulations with physiological and abnormal glial parameters. Bi-stability is observed with abnormal glial parameters. In each of a, b, c and d: Top panel shows the membrane potential (red) and potassium reversal potential E K (blue). Second panel shows the extracellular potassium concentration [K +]o (red) and the glial bound buffer concentration [K B] (blue). Third panel shows the rate of change of extracellular potassium concentration due to the glial action ∂[K+]o∂t|glia. Bottom panel shows the potassium pump current I Kpump. (a)“Physiological” parameters. Initial conditions [K +]o(t=0)=1mM, [K B](t=0)=3mM. (b)“Physiological” parameters. Initial conditions [K +]o(t=0)=10mM, [K B](t=0)=10mM. (c)“Pathological” parameters. Initial conditions [K +]o(t=0)=1mM, [K B](t=0)=3mM. (d)“Pathological” parameters. Initial conditions [K +]o(t=0)=10mM, [K B](t=0)=10mM Results We build model neocortical networks (Fig. 1) using conductance-based model neurons endowed with glial dynamics for potassium uptake. Based on the observations of human focal seizure data (LFPs and spiking activity in ensembles of single neurons), our aim is to use these model networks to reproduce the neural dynamics during the transition into ictal states of propagated focal seizures. Our presentation is structured as follows. First, based on both single neuron and network simulations (Section 2), we demonstrate that pathological glial uptake of potassium can lead to bi-stable dynamics in the model cortical network. The bi-stable dynamics consist of a “high-activity” state characterized by an elevated [K+]o level, and a state of “low-activity” with a low [K+]o level. We relate the “high-activity” state to “ictal” events and the “low-activity” state to non-epileptic, normal activity. Although network bistability has been implicated in computational studies as an avenue for seizure transition (Fröhlich et al. 2010), here we demonstrate that abnormal glial buffering can increase the propensity of single neurons (and thus the cortical network) to exhibit bi-stability (Fig. 2). The underlying mechanism for the bistabilty is the imbalance in the reversible reaction of the potassium buffer between the bound and unbound forms (10). Second, with the seizure transition mechanism in place for the model cortical network, we then examine how certain types of synaptic changes involving mostly inhibitory interneurons and their interaction with potassium extracellular concentration may lead to various types of seizure transitions as observed in the human data. Our strategy is to explore inhibitory synaptic and intrinsic parameters relevant to the occurrence of depolarization block in model inhibitory interneurons. We also note that, in our model, pyramidal neurons are not as prone to depolarization block (in comparison to inhibitory interneurons) because of their Ca2+-mediated potassium conductances (gKCa) and the corresponding after-spike hyperpolarization effects (7). Variations of parameters related to synaptic inhibition and glial potassium buffering allow us to replicate in detail three types of seizure patterns observed in the human data (gamma seizures and two types of spike-and-wave seizures; Truccolo et al. 2014). Thus, all three types of seizures can each be linked, via the same computational network model, to specific temporal patterns of synaptic activities and glial buffering dynamics during the evolution of propagated ictal states (Figs. 3, 4 and 5). Table 2 summarizes the results of our network simulations in terms of the paths leading to the three types of seizure transitions. All of the primary variables and parameters used in the simulations and analyses below are introduced and defined in Tables 1, 3, 4 and 5. Fig. 3 An example gamma seizure simulation with pathological glial and pump parameters. (a) Depiction of neural activity and time course of various biophysical parameters during a 90-second simulation. The network starts with a “low-activity” non-seizure state (first 40 seconds of simulation). A DC stimulation is applied to every neuron in the network between 40 and 42.5 seconds, after which the system is “kick-started” into a high activity state with an elevated level of extracellular potassium (after 42.5 seconds). Topmost panel: average spike rate of pyramidal cells (red) and interneurons (blue) during the simulation. Second panel: raster plot of action potentials of neurons in 2 of the 81 simulated minicolumns. These 2 minicolumns are located in the centre of the network. Third panel: average membrane potential values of interneurons (blue) and pyramidal cells (red) during simulation. Fourth panel: average extracellular potassium level values for interneurons (blue) and pyramidal cells (red) during simulation. Fifth panel: Time-frequency plot of the average membrane potential values of the pyramidal cells (third panel-red). Clear sustained gamma band oscillations (∼40 Hz) at the population level emerge with the DC stimulation and persist even after the stimulation is terminated. (b) A 2-second segment from the 90-second simulation in a. Same conventions as in a, except that the bottom panel is a power spectral density (PSD) plot of the 2-second average membrane potential values of the pyramidal cells. The maximum of the PSD is located in the gamma range. (c) Histograms of interspike intervals (ISIs) of all the pyramidal cells (red) and interneurons (blue) during gamma population activity (t≳45 seconds). Insets: Histograms of the ISI coefficients of variation (CV) for spike trains of individual pyramidal cells (red) and interneurons (blue), also during gamma population activity (t≳45 seconds) Fig. 4 An example simulation showing a transition into a spike-wave-complex (SWC) seizure via low inhibitory synaptic strength. (a) Depiction of a 90-second simulation showing several biophysical variables. As in Fig. 3, the network starts with a “low-activity” non-seizure state (first 40 seconds of simulation). DC stimulation is applied to every neuron between 40 and 42.5 seconds. Topmost panel: average spike rate of pyramidal cells (red) and interneurons (blue) during the simulation. As rhythmic SWC discharges emerge, the firing rate of pyramidal neurons approaches ∼5 Hz. Second panel from top: raster plot of action potentials for neurons in 2 of the simulated 81 minicolumns. The location of these 2 minicolumns is the same as in Fig. 3. Third panel: average membrane potential values of interneurons (blue) and pyramidal cells (red) during simulation. Fourth panel: average extracellular potassium level values for interneurons (blue) and pyramidal cells (red). Fifth panel: Time-frequency plot of the average membrane potential values for the pyramidal neurons (third panel-red). The power peak at ∼5 Hz (after the DC stimulation) corresponds to the emergence and maintenance of rhythmic SWC discharges, even after the DC stimulation is terminated. (b) A 10-second selection from the 90-second simulation in (a) showing the transition (after the DC stimulation) into SWCs. Same convention as in (a), except the bottom panel is a power spectral density (PSD) plot of the 10-second average membrane potential values of the pyramidal cells (third panel-red). The cessation of spiking of interneurons at around t=46 seconds (first and second panels) as they enter depolarization block (third panel-blue) is concomitant with the emergence of SWCs of increasing amplitude (third panel-red) Fig. 5 An example simulation showing transition into low voltage fast (∼20 Hz) oscillation followed by SWC discharges, obtained via initial high inhibitory synaptic strength and subsequent depletion of inhibition. (a) Depiction of a 90-second simulation showing several biophysical variables. The network spends its first 40 seconds in a low-activity “non-epileptic” state. A DC stimulation is applied to every neuron in the network between 40 and 42.5 seconds. During this DC stimulation period, the effective inhibition of the network is also drastically increased to mimic the build up of inhibitory strength prior to seizure (unlike in Figs. 3 and 4 where the effective inhibition is kept constant throughout the entire simulation). This increased effective inhibition is held constant up to t=49.5 seconds, after which inhibition wears off to zero (starting point of this depletion phase is marked by the red triangle). Topmost panel: average spike rate of pyramidal cells (red) and interneurons (blue) during the simulation. Second panel: raster plot of action potentials of neurons in 2 of the 81 minicolumns simulated. The location of these 2 minicolumns is the same as in Figs. 3 and 4. Third panel: average membrane potential values of interneurons (blue) and pyramidal cells (red). Fourth panel: average extracellular potassium level values for interneurons (blue) and pyramidal cells (red). Fifth panel: time-frequency plot of the average membrane potential values of the pyramidal cells (third panel). (b) A 30-second selection from the 90-second simulation in a showing both the transient low voltage fast (∼20 Hz) oscillations and the subsequent emergence of rhythmic SWC discharges. Same convention as in a, except the bottom panel is a power spectral density (PSD) plot of the 30-second average membrane potential values of the pyramidal cells. Transient (∼ 20 Hz) oscillations are observed after the DC stimulation but before inhibition wears off starting at t=49.5 seconds (red triangle). As inhibition begins to wear off, the firing rate of interneurons shows an initial transient increase, after which it decreases to zero when the interneurons enter depolarization block (third panel-blue, after around t=50 seconds). The reduced inhibition (due to both active wearing off of inhibition and the cessation of interneuron spiking) and increased [K +]o (fourth panel) also lead to the increased firing of pyramidal cells and the emergence of rhythmic SWC discharges (third panel-red, towards the end) Table 2 Pathways to three main seizure types in the focal seizure model Seizure type Synaptic inhibition Interneuron depolarization block Representative simulation Gamma Strong synaptic inhibition No Fig. 3 Spike-wave complex (SWC) Weak synaptic inhibition Yes Fig. 4 Low-voltage fast oscillations followed by SWCs Strong synaptic inhibition at seizure onset, followed by a weakening of inhibition (e.g. GABA depletion). Yes Fig. 5 Table 3 Intrinsic, glial and ionic parameters used for network simulations Parameter Description Units Values Pyramidal Interneurons C Specific neuron membrane capacitance μF/cm2 1 1 g Na Sodium conductance mS/cm2 35 35 g K Potassium conductance mS/cm2 9 S g L Leak conductance mS/cm2 0.135 0.1 g KCa Ca 2+-activated potassium conductance mS/cm2 S 0 g Ca Calcium conductance mS/cm2 0.1 0 E Na Sodium reversal potential mV 54 54 E Ca Calcium reversal potential mV 120 120 I ext External applied current to model neurons μA/cm2 S S [K +]i Intracellular potassium concentration mM 133 133 [N a +]o Extracellular sodium concentration mM 130 130 [N a +]i Intracellular sodium concentration mM 17 17 [C l −]o Extracellular chloride concentration mM 130 130 [C l −]i Intracellular chloride concentration mM 8 8 D Diffusion constant of K + in extracellular space (eqn. (12)) cm 2/ms 2.5×10−9 2.5×10−9 I Kmax Maximal potassium pump current μA/cm2 S S [K +]o(eq) Equilibrium extracellular potassium concentration of I Kpump mM S S [B]max Maximal glial free buffer concentration (eqn. set (11)) mM 500 500 k b Backward glial unbinding rate (eqn. set (11)) ms −1 0.0008 0.0008 [K +]o(th) Threshold extracellular potassium concentration of glial buffer (eqn. set (11)) mM S S 𝜃 see eqn. set (11) mM S S An “S” on the value column denotes that the parameter is dependent on specific simulations. Please refer to the main text for specific values Table 4 Variables for network simulations Variable Description (one variable for each modeled neuron) Units [K +]o Extracellular potassium concentration (eqn. (13)) mM E K Potassium reversal potential (eqn. (15)) mV E L Leak reversal potential (eqn. (3)) mV [C a 2+]i Intracellular calcium concentration (eqn. (8)) mM [B] Glial free buffer concentration (eqn. set (11)) mM [K B] Glial bound buffer concentration ([K B]≡[B]max−[B]) mM k f Forward glial binding rate (eqn. sets (10) and (11)) ms −1⋅mM −1 s Synaptic variable of each neuron (eqn. set (18)) - g e Background excitatory conductance (eqn. set (19)) mS/cm 2 g i Background inhibitory conductance (eqn. set (19)) mS/cm 2 V Membrane potential mV h Sodium current inactivation variable - n Potassium current activation variable - m∞ Sodium current fast activation variable - t Time (independent variable) ms x,y,z Spatial coordinates (independent variables eqn. set (12)) cm Table 5 Synaptic parameters used for network simulations Parameter Description Units Value E i Inhibitory reversal potential mV -72 E e Excitatory reversal potential mV 0 τ i Inhibitory synaptic decay time constant ms 5 τ e Excitatory synaptic decay time constant ms 3 τ delay Axonal conduction delay ms 0.5 s max Maximal value of gating variable per spike - 1, except for Fig. 5 where it varies in time for interneurons smaxterm Terminal value of maximal gating variable per spike (Figure 5, interneurons only–equation (21)) - 70 τ rise see Eq. (21). Figure 5; interneurons only. ms 62.5 g e0 Mean background excitatory drive to model neurons mS/cm 2 S σ e Background excitatory fluctuation level mS/cm2 S g i0 Mean background inhibitory drive to model neurons mS/cm 2 S σ i Background inhibitory fluctuation level mS/cm2 S gsyni→i Unitary conductance from an inhibitory interneuron to another interneuron mS/cm2 S gsyne→e Unitary conductance from a pyramidal cell to another pyramidal cell mS/cm2 S gsyne→i Unitary conductance from a pyramidal cell to another interneuron mS/cm2 S gsyni→e Unitary conductance from an interneuron to another pyramidal cell mS/cm2 S P(i→i) Connection probability from one interneuron to another interneuron - 0.25 P(e→e) Connection probability from one pyramidal cell to another pyramidal cell - 0.15 P(e→i) Connection probability from one pyramidal cell to another interneuron - 0.15 P(i→e) Connection probability from one interneuron to another pyramidal cell - 0.25 Γm→n Connectivity matrix element (with values of either 1 or 0) between neuron m and n - assigned according to the four probability values above An “S” on the value column denotes that the parameter is dependent on a specific simulation. Please refer to the main text for specific values Firing properties of the single-neuron models under physiological and abnormal glia uptake conditions We used four single pyramidal cell model simulations to show the effects of the potassium pump and glial dynamics on neuronal excitability (Fig. 2). In each of these four simulations, Iext=0.36, [K+]o(eq)=3, [K+]o(th)=15, IKmax=7, gKCa=5. In this section, we define the “physiological” value 𝜃=−1.15, and the “pathological” value as 𝜃=−0.05 (equation set (11)). Table 3 specifies the units of these parameters. Our aim is to show that the neural system tends to settle into a single steady state for a wide range of initial conditions when the increase in the forward binding rate kf is smooth with increasing [K+]o (as in the physiological case, effected by a larger absolute value of 𝜃 – see equation set (11)). However, when there is a sharper increase in kf with increasing [K+]o (as in the pathological case, effected by a decreased absolute value of 𝜃), some of the initial conditions can lead the neural system into a state where virtually all of the glial buffers are free (i.e. [KB]→0). In this case, the system’s ability to return potassium to the extracellular space and maintain a target value of extracellular potassium level is impaired. The impairment occurs because Eq. (10) is no longer reversible when [KB]→0. In Fig. 2a and b, we show two single pyramidal neuron model simulations with the same set of “physiological” glial and potassium pump parameters but with different [K+]o and [B] initial conditions. In Fig. 2a, we set [K+]o(t=0)=1 mM and [KB] (t=0)=3 mM, and in Fig. 2b, [K+]o(t=0)=10 mM and [KB] (t=0)=10 mM. It is clear that, despite the difference in initial conditions between the two simulations, the model neuron and the glial buffering system in both simulations converge to the same steady state. The steady state is characterized by a dynamic equilibrium established between the glial buffer’s binding (kf) and unbinding (kb) of the extracellular potassium. This dynamic equilibrium is evidenced by the small but non-vanishing [KB] in the steady state. Moreover, all the steady state variables (including firing rates, [K+]o and [KB]) are identical between the two simulations, indicating that the steady states of the two simulations are the same. The glial buffering system enforces an upper ceiling for the extracellular potassium level in both simulations. In Fig. 2c and d, we show the same two simulations with “pathological” glial parameters. In contrast to Fig. 2a and b, we see that the same set of different initial conditions results in distinct steady states for the two simulations. In Fig. 2c, the model neuron and the glial buffer approach a state where there is a lower [K+]o and a vanishing value of [KB], while in Fig. 2d, the final [K+]o and [KB] values are higher and the neuron is excitable with a higher firing rate. Thus, bi- or multi-stable dynamics at the single neuron level are more easily obtained with a smaller absolute value of 𝜃, resulting into multiple fixed points (steady states) having either a lower [K+]o or higher [K+]o value. The main factor behind this bi-stable dynamics is the parameter 𝜃, which determines the level of bias for the forward binding rate kf against the backward rate kb (see equation sets (10) and (11)). With a small absolute value of 𝜃, at a critical [K+]o level lower than [K+]o(th), kb can be overwhelmingly larger than kf. The large kb value rapidly lowers [KB]. Although this backward unbinding action releases K+ back to the extracellular space, the IKpump (9) counteracts this effect and therefore [K+]o remains low. Moreover, the small kf value (due to the bias factor 𝜃) is not sufficient to replenish [KB] via the forward binding action. As a result, the system ends up in a state where virtually all the buffer is in the free state B with a low level of [K+]o (as in Fig. 2c). For the neuron and glial system to exit this state, as we have shown in Fig. 2d, one can either artificially increase the values of [K+]o or [KB] (for example, by setting up the initial conditions). Increasing [K+]o to a value closer to [K+]o(th) forces kb to be less biased against kf (equation set (11)), so that [KB] can be more efficiently replenished and thus allows a dynamic equilibrium between [B] and [KB]. Increasing [KB] artificially serves similar purposes by increasing the backward unbinding release of K+ to the extracellular space (thus increasing [K+]o). In essence, decreasing the absolute value of 𝜃 reduces the range of [K+]o over which the glial buffering system can efficiently regulate, and thus makes the single neuron more susceptible to multi-stable behaviour. We show in the next section that this apparent multistability also carries over to the network simulations with abnormal glial parameters. As mentioned above, we will associate the state in which the [K+]o is low with the normal physiological state, and the elevated [K+]o level state with the “seizure” state. We also note that, although the above single-neuron analysis was based on pyramidal neurons, the same low and elevated [K+]o states occur in the case of FS inhibitory neurons. The effect of elevated [K+]o on pyramidal and FS interneuron spiking will, however, differ in the network simulations presented below. Sustained gamma epileptiform activity results from an abnormal but balanced glial dynamics and inhibitory synaptic strength In this section we show how most features of gamma seizures observed in the human focal epilepsy data (Truccolo et al. 2011, 2014) can be replicated by our network model. These gamma seizures in the human data are characterized by sustained narrow band gamma LFP oscillations (∼30−60 Hz). However, few fine temporal synchrony transients exist in the spiking activities of the measured neurons. Neuronal spiking tends to be asynchronous and heterogeneous (Truccolo et al. 2011, 2014). Individual neurons also fire at a lower rate than the gamma frequency of observed LFP oscillations. Similar neural dynamics is observed in our model simulations (Fig. 3). Critical ingredients for the appearance of sustained gamma band LFP oscillations with asynchronous neuronal firing at a lower rate include high inhibitory conductance values (for both gsyni→i and gsyni→e) and an axonal conduction time delay (Brunel 2000; Brunel and Wang 2003). We also require a value of [K+]o(th) that is not too high (see equation set (11)). A moderate [K+]o(th) value ensures that the model FS inhibitory interneurons have a high excitability in the “epileptic” state, while preventing the interneurons from being overly excited to enter depolarization block. We used gsyne→e=0.0007, gsyne→i=0.0007, gsyni→i=0.025, gsyni→e=0.025, ge0=0.01026, σe=0.0025, gi0=0.084, σi=0.02 for this gamma seizure simulation (Table 5 introduces and defines these synaptic parameters). As for intrinsic and glial properties of model neurons, we chose 𝜃=−0.15, [K+]o(eq)=3.6, [K+]o(th)=7.5, IKmax=1.45, gKCa=5 for pyramidal cells and 𝜃=−0.15, [K+]o(eq)=3, [K+]o(th)=7.5, IKmax = 1.9, gK = 6.8 for interneurons. The value of gK for interneurons was chosen so that they only enter depolarization block at a relatively high level of [K+]o (≳8 mM). Figure 3a shows the first 90 seconds of the temporal progression of several biophysical variables in the model and the time-frequency spectrum of the average membrane potentials across all the pyramidal neurons in the network (Fig. 3a, lowest panel). We use this average value as a proxy for the LFP activity. The model cortical network during the first 40 seconds of the simulation is in the “low activity” state (in which [K+]o converges to ∼3mM). This “low-activity” state at the network level is similar to the single neuron level in Fig. 2c where the replenishment of [KB] is insufficient for the glial buffer to maintain a targeted level of [K+]o. At time t=40s, a DC stimulation (2.5 μA/cm 2 added to Iext) was applied to every neuron in the model network for a short period of 2.5 seconds. This stimulation kick-started the extracellular potassium accumulation process and forced the cortical network to enter into a “high-activity” or “epileptic” state in which the glial buffer attempts to lock into a higher [K+]o level (Fig. 3a, 4 th panel), as determined by the parameter [K+]o(th). Clear gamma band (∼40 Hz) oscillations are readily observed in the LFP power spectrum after the DC stimulation (Fig. 3a, lowest panel). Importantly, the gamma activity is sustained, even after DC stimulation is terminated, persisting until the end of the simulation. Furthermore, at the individual neuron level, both the fast-spiking interneurons and pyramidal cells spike at lower rates of ∼ 3 and 1 spike per second, respectively, than the frequency of the gamma LFP activity (see Fig. 3a, topmost panel). The second panels from the top of Fig. 3a and b show raster plots of 2 of the 81 minicolumns in the simulation. The neuron indices in the raster plots are grouped by neuron types (pyramidal or interneuron). It is also clear that while the LFP proxy (average membrane potential of pyramidal cells) shows gamma oscillations (Fig. 3a and b, 3 rd panels from top), the firing pattern of each individual neuron is not indicative of the global oscillation as observed in the LFP proxy. The density plots of the interspike intervals (ISIs) after the DC stimulation across each of the pyramidal and interneuronal populations (Fig. 3c, main plots) reveal a large variance of ISI values. Spiking in FS interneurons showed high irregularity, with the coefficient of variation (CV) of ISIs concentrating around 1 and higher values, while the CV for pyramidal neurons tended to be lower and more broadly distributed (Fig. 3c, inset plots). Irregularity of pyramidal cell spiking can be increased, nevertheless, by moderately increasing the unitary inhibition from the interneurons to the pyramidal cells. However, we observed a competition between the glial parameters (i.e. 𝜃 and [K+]o(th)) and the unitary inhibitory conductance values in terms of sustaining high enough [K+]o values to support gamma oscillations. Increasing the unitary inhibitory conductance values affects the sustainability of gamma oscillations. Too strong gsyni→e or gsyni→i values make the gamma oscillations less sustainable, with the system immediately settling back into the original “low-activity” state. A rough mean-field estimate of the synaptic currents entering the pyramidal cells after the DC stimulation suggests that the magnitude of the inhibitory current is a few times higher than that of the excitatory current. This average inhibitory synaptic current entering a pyramidal cell can, for example, be estimated by the formula: 20 Isyni→e=P(i→e)×gsyni→e×Ni×〈(V−Ei)〉×τi×〈νi〉, where νi is the spike rate of the inhibitory population and Ni is the number of inhibitory interneurons in the network. Transition into spike-and-wave complex (SWC) seizures: role of synaptic inhibition and depolarization block In this section we show how variations in synaptic inhibition and depolarization block in the examined neuronal network model can reproduce the dynamics of the observed SWC seizures. In this type of seizures, neural activity transitions into high amplitude rhythmic 2−3 Hz LFP discharges. Each SWC event in the LFP lasts for about 300 ms and is characterized by an initial fast “spike” followed by a slow wave potential. Neuronal spiking tends to occur during the LFP “spike” phase and is highly suppressed during the LFP wave phase (e.g. Truccolo et al. 2014). Preliminary examination of neuronal spiking characteristics has indicated that although both putative principal and FS inhibitory interneurons tend to increase their firing rates in the initial stages of spike-wave seizures (Ahmed et al. 2014), putative FS interneurons tend to shut down later, just before the emergence of full spike-wave discharges. Ahmed et al. (2014) observed that this termination of activity is preceded by a progressive decreasing of action potential amplitudes in FS interneurons – an indication that FS interneurons enter depolarization block, resulting eventually in the cessation of their activity. Figure 4 shows simulation results for the case where the overall inhibitory synaptic strength is kept at a low level during the entire simulation, leading to SWC seizures. Because the aim is to have FS interneurons to eventually enter depolarization block, we set a higher [K+]o(th) and a lower gK value for the interneurons than in the gamma seizure simulation (Fig. 3). Higher [K+]o(th) values increase FS interneuron excitability, while lower gK values lead FS interneurons to enter depolarization block at a lower [K+]o level. Parameters for model neurons and glial properties of model neurons corresponded to: 𝜃=−0.15, [K+]o(eq)=3.6, [K+]o(th)=8.5, IKmax=1.8, gKCa=12 for pyramidal cells and 𝜃=−0.15, [K+]o(eq)=3, [K+]o(th)=7.75, IKmax=1.9, gK=3.5 for interneurons. Unitary inhibitory conductance values gsyni→i and gsyni→e were set here to only one-tenth of that in the gamma seizure simulation (Fig. 3). Synaptic parameters corresponded to: gsyne→e=0.0007, gsyne→i=0.0007, gsyni→i=0.0025, gsyni→e=0.0025, ge0=0.01016, σe=0.0025, gi0=0.084, σi=0.02. As before, the simulation starts at a “low-activity” state (t=0−40 seconds) and a DC stimulation (1.6 μA/cm 2) is delivered between t=40 and 42.5 seconds, “kicking” the system into a high-activity “epileptic” state. Since the overall inhibitory synaptic strength is low, but [K+]o(th) is higher than that in the gamma seizure case, FS inhibitory interneurons fire at a higher frequency after the DC stimulation than in Fig. 3, creating a positive feedback which further increases [K+]o. When [K+]o reaches a critical value, inhibitory interneurons enter depolarization block, which eventually leads to cessation of inhibition (at around t=46 seconds in Fig. 4). We note that, in contrast to the gamma seizures, the attainment of the critical value for [K+]o is possible here because [K+]o(th) is set to a higher level, such that the glial potassium ceiling is higher. After cessation of synaptic inhibition, the network dynamics is dominated by the interaction amongst pyramidal cells and evolve into rhythmic SWC discharges. Throughout, abnormal glial buffering maintains the high [K+]o required for keeping the interneurons in the depolarization block regime, thus creating a long-lasting “seizure” state. In the model simulations, IKCa adaptation currents in the pyramidal cells and sufficiently strong excitatory-excitatory (gsyne→e) couplings (Van Vreeswijk and Hansel 2001; Dur-e-Ahmad et al. 2012; Nicola and Campbell 2013b; Ferguson et al. 2015) are critical for both the “spike” and “wave” firing suppression phases of the the SWC discharges. Transition into seizures consisting of initial low-voltage fast LFP oscillations followed by spike-and-wave discharges Next, we demonstrate the generation of the second type of SWC seizures. In this case, the emergence of rhythmic SWC discharges is preceded by low voltage higher frequency (∼10−20 Hz) LFP oscillations. Neuronal spiking activity during these fast oscillations tends to be asynchronous as in the previously examined gamma seizures. In other words, while there is a clear fast frequency oscillation at the LFP level, individual neurons spike at a lower rate without obvious synchrony with the global (i.e. LFP) oscillation patterns. The similarity of the firing dynamics of individual neurons between these transient oscillations and gamma seizures leads us to speculate that both phenomena might share a common synaptic and glial mechanism. As shown in Fig. 5, the seizure begins with low voltage fast (∼20 Hz) LFP oscillations which eventually evolve into low-frequency rhythmic SWC discharges. Similarly to the gamma seizures (Fig. 3), the initial fast oscillations in the model require a high overall inhibitory synaptic strength, while the evolution and maintenance of spike-wave discharges require the opposite. To reproduce this type of seizure dynamics, the following sequence of events was implemented in the model: a substantial increase in inhibitory synaptic strength at seizure transition (leading to low voltage fast oscillations) is followed by a rapid decrease, which is hypothesized here to occur via GABA depletion (Zhang et al. 2012). Once synaptic inhibition is below normal levels, the same path as in the first type of SWC seizures follows: inhibitory interneurons enter depolarization block and the network evolves into rhythmic SWC discharges supported mostly by the population of pyramidal neurons. Specifically, the synaptic parameters gsyni→i and gsyni→e were set to constant values during the entire simulation. To increase the effective inhibitory synaptic strength during seizure transition (i.e. during the DC stimulation phase, t=40−42.5s), we let the maximal value of the gating variable per spike, smax, for interneurons (equation set (18)), vary temporally instead of being a constant as in Figs. 3 and 4. The value of smax for interneurons was set to 1 at the beginning of the simulation (Table 5) and kept at this value during the “low activity” phase (t=0−40s). After that, smax for interneurons asymptotically approached a terminal value smaxterm during the DC stimulation phase according to the following formula: 21 smax(t+δt)=smax(t)×1+smaxtermsmax(t)−1×δtτrise, where τrise=62.5ms is the time constant of the increase in smax and smaxterm=70 is the asymptotic terminal value of smax (i.e. 70 times the value of original smax before the DC stimulation phase). The increased smax value for interneurons was then held constant during the fast oscillations phase (from 42.5 to 49.5 seconds). After that it was decreased to simulate the loss of inhibition. We set smax→0.1×smax for individual interneuron after each spike of the same interneuron during this “depletion” phase. (The start of this depletion phase is marked by a red triangle in Fig. 5.) Eventually the inhibition approaches zero, the “depleted” state. In summary, Fig. 5 shows the emergence of clear low voltage fast (∼ 20 Hz) LFP oscillations right after the onset of the DC stimulation. These fast oscillations disappear shortly after the start of the inhibition “depletion” phase (marked by a red triangle in Fig. 5). The strong build up of inhibitory synaptic strength (up to 70 times the initial strength) during the DC stimulation is necessary because it slows the rise of [K+]o during DC stimulation and keeps the interneurons from entering depolarization block. Moreover, this very high level of synaptic inhibition keeps the population frequency in the ∼10−20 Hz range. A lower, but still large inhibitory strength would have led to the emergence of gamma oscillations, as in Fig. 3. The initial decrease in [K+]o during the transient fast oscillations (Fig. 5a, fourth plot from top) is due to the strong inhibition which limits the firing of inhibitory interneurons. As inhibition begins to wear off at t=49.5 seconds, the trajectory of [K+]o reverses its course because both pyramidal cells and interneurons fire more as a result of decreased inhibitory synaptic strength. Finally, when [K+]o reaches a sufficiently high level (∼7 mM for this simulation), the interneurons enter depolarization block and inhibition ceases, as clearly shown by the complete termination of interneuron firing (Fig. 5b, raster plot at around t=63 seconds). After cessation of inhibition, SWC discharges emerge supported only by the activity of interacting pyramidal neurons. With the exception of the time varying nature of smax for interneurons and a DC stimulation here set to 2.5 instead of 1.6 μA/cm2, the numerical values of the synaptic (including gsyni→i and gsyni→e), intrinsic and glial parameters were the same as those for the first type of SWC seizures (Fig. 4). Discussion Our study is motivated by recent microelectrode array recordings of ensembles of single neurons and high-density LFP activity in human neocortex during propagated focal seizures (Truccolo et al. 2011; Schevon et al. 2012; Ahmed et al. 2014; Truccolo et al. 2014; Wagner et al. 2015). These recordings suggest different ways via which seizures starting in a focal site can recruit more distal neocortical areas. We used a biophysical cortical network model consisting of conductance-based neurons, coupled with glial buffer for [K+]o, to examine potential mechanisms underlying transitions into ictal states in recruited neocortical areas during seizure spread. Through the analyses and simulations of single neuron and network models, we have demonstrated that neocortical networks can be made susceptible to seizures as a result of abnormal glial potassium buffering and changes in synaptic inhibition. First, at the single neuron level, the imbalance in the conversion between potassium buffer in the bound and unbound forms leads to bi-stability such that one state has a lower [K+]o level (thus lower excitability) than the other. This bi-stability also carries over to the network level. Perturbation either by noise or by DC stimulation can “kick-start” the high [K+]o hyper-excitable state from the low [K+]o state. Second, we then showed that the three major types of seizure transitions as observed in the human microelectrode array recordings can be reproduced when these pathological glial potassium dynamics interact with various synaptic and intrinsic parameter settings for inhibitory interneurons. In our simulations, transitions into gamma (∼30−60 Hz) ictal activity were obtained with a high inhibitory conductance between cortical neurons. Such high inhibitory synaptic strength was required to prevent the interneurons from entering depolarization block during the seizure state. Moreover, since high inhibitory strength tends to decrease [K+]o, the long-lasting gamma seizure state was the result of a subtle balance between high inhibitory conductance values and the accumulation of [K+]o due to abnormal glial potassium buffering dynamics. For transitions into SWC ictal activity, the cortical network can either experience a generally low value of inhibitory conductance between cortical neurons, or an initial high inhibitory synaptic strength followed by inhibition breakdown. In the first scenario, the transition into SWC was direct and involved interneurons entering depolarization block because of the low inhibitory synaptic strength. As soon as the interneurons entered depolarization block, inhibition ceased. At that point, neuronal spiking activity supported only by interacting pyramidal cells led to recurring SWC discharges. In the second SWC scenario, there was a buildup of inhibitory synaptic strength prior to the transition into spike-wave discharges. The result of this inhibition buildup is the occurrence of transient low-amplitude fast oscillations around 10−20 Hz. This buildup of inhibitory synaptic strength was followed by a wearing off of inhibition (e.g. through GABA depletion), thus providing a low inhibition environment for interneurons to enter depolarization block and the population of pyramidal cells to eventually sustain recurring SWC discharges. Abnormal glial buffering and its relationship to seizure In this work we have established a computational model of seizure through the occurrence of bi-stability at the single neuron level. This bi-stability is the result of a hypothesized pathological glial potassium dynamics in which there is an imbalance between potassium buffers in their bound and unbound forms. Bi- (or multi)stability, either at the single neuron or at the network level, has been implicated in numerous physiological and pathological conditions in the brain (Van Ooyen et al. 1992; Sasaki et al. 2007; Freyer et al. 2009; Fröhlich et al. 2010; Anderson et al. 2012; Ho et al. 2012; He 2014; Ho et al. 2014; Hübel et al. 2014). In the particular case of epileptic seizures, Fröhlich et al. (2010) have demonstrated in computational models that it is possible for network bistability to exist even when the glial potassium dynamics are within a physiological range (through a higher absolute value of 𝜃, the parameter determining the bias favouring forward versus backward binding rate). In this scenario, Fröhlich et al. (2010) associate higher [K+]o states with epileptic seizures. In the models used here, we did not observe bi-stability either at the single neuron or network level when 𝜃 was set to physiological values. We only observed bi-stability with pathological (i.e. smaller in absolute value) values of 𝜃. We have singled out (via single neuron simulations) the differential in transition rates between [B] and [KB] as the important factor underlying such bi-stability. The fact that Fröhlich et al. (2010) used two-compartment neuron models, while we used single-compartment, might explain this difference between the two studies. Moreover, different synaptic connection parameters were used by Fröhlich et al. (2010). Different synaptic connectivity might affect the length and strength of perturbation required to elicit the bi-stable behaviour, if at all possible. A more thorough examination of how glial potassium buffering may affect neuronal dynamics should involve detailed bifurcation analyses of the combined neuron-glial system (Touboul 2008; Nicola and Campbell 2013a; Kim and Nykamp 2014; Nicola and Campbell 2014). Such analyses could determine the intrinsic and glial parameters that allow multistability. From the existence and stability of these states, one would expect to gain some insight into how resilient the network is to seizure initiation and propagation. Although we have focused on their role on seizure propagation, potassium and glial buffering may also play an important role in seizure termination (Kramer et al. 2012; González-Ramírez et al. 2015). Changes in inhibitory and excitatory conductances Our focus on changes in inhibition was motivated by several previous (in vivo/vitro) animal models (Trevelyan et al. 2006; Zhang et al. 2012; Grasse et al. 2013; žiburkus et al. 2013; Uva et al. 2015) and in vitro studies of human epileptic cortical tissue (e.g. D’Antuono et al. 2004), where inhibitory interneuron activity appears to change significantly during preictal and initial periods of the seizure, before major changes in principal cells are detected. Recent work by žiburkus et al. (2013) has also examined in detail how changes in both inhibitory and excitatory conductances may lead to imbalances in excitation/inhibition preceding and during seizure like events in hippocampal slices under the potassium channel blocker 4-aminopyridine and reduced extracellular magnesium. In that respect, we emphasize that the changes in inhibition as implemented in our simulations did not only affect inhibitory conductances. Changes in inhibition in our model were implemented in part by systematically varying the levels of unitary synaptic inhibition, which in turn affected several other network properties such as the number of neurons generating spikes at any given time, and thus firing rates in both populations of pyramidal neurons and FS interneurons. In this way, by varying unitary synaptic inhibition, the E-I balance in the simulated networks also varied dynamically, with both excitatory and inhibitory conductances being altered. In addition, after transitions into spike-wave discharges in the simulated SWC seizures (Figs. 4 and 5), these discharges were supported only by the synaptic interactions among pyramidal neurons. Furthermore, we note that there are a couple of important differences between the animal model in žiburkus et al. (2006, 2013) and our model simulations. First, apart from temporary depolarization block periods, the activity of inhibitory neurons remained throughout the seizures (i.e. until seizure termination) in žiburkus et al.’s model. In our computational model, it is only during gamma seizures that the spiking activity of FS inhibitory interneurons is preserved throughout the seizure and depolarization block is absent. By contrast, FS interneurons enter depolarization block and remain in that state during the SWC seizures. Second, while in žiburkus et al.’s study transient depolarization block events in pyramidal neurons are also present and play an important role, in our computational model pyramidal neurons tend to be robust to depolarization block both in SWC and gamma seizures. This robustness resulted primarily from the slow calcium-mediated potassium afterhyperpolarization (AHP) currents (IKCa, Eq. (7)) in the model pyramidal neurons, which prevented them from evolving into regimes of high enough firing rates required for depolarization block. These features of the biophysical model examined here seem consistent with the dynamics of propagated seizures (in neocortical patches distal from putative onset areas) observed in our human data. First, both inhibitory and excitatory spiking activity appears to be preserved throughout the gamma seizures (Truccolo et al. 2011). Second, in SWC seizures, spiking in putative FS interneurons appears to terminate before the transition into large amplitude spike wave discharges. This cessation of FS interneuron spiking activity is accompanied by signatures of depolarization block, while no similar signatures were found in putative principal cells (Ahmed et al. 2014). Network inhibition and its relationship to seizure propagation Using biophysical neuronal network models of neocortical patches, we have demonstrated how FS inhibitory interneurons may play an important role in three types of observed human seizures. In particular, we have quantified the role of FS interneurons by varying the inhibitory synaptic strength and controlling how easily interneurons enter depolarization block. For the two types of SWC seizures modeled here (Figs. 4 and 5), ultimately it is the failure of inhibition (when the FS inhibitory interneurons enter depolarization block) that leads the network to transition into SWC activity. We examined two main scenarios that led to the failure of inhibition, resulting in SWC seizures. In the first scenario, FS inhibitory interneurons enter depolarization block when there is an overall low inhibitory synaptic strength. When interneurons are excited during seizure initiation, the lower inhibitory synaptic strength allows interneurons to fire at a higher frequency, thus hastening the [K+]o accumulation process and leading eventually to depolarization block (Fig. 4). In the second scenario, an initially large inhibitory synaptic strength is followed by GABA depletion (Zhang et al. 2012). Initially, interneurons are mutually inhibited by the large inhibitory synaptic strength and their firing frequencies are too low to trigger [K+]o accumulation. If such level of inhibition was not perturbed, seizure propagation would fail. The decreased effective inhibitory strength resulting from subsequent GABA depletion allows for higher firing rates in interneurons (and principal cells) to develop, again leading eventually to depolarization block in inhibitory interneurons (Fig. 5). Some previous studies have emphasized the role of feedforward inhibition in seizure propagation in neocortex (Trevelyan et al. 2006; Schevon et al. 2012; for a review see Paz and Huguenard 2015). The main idea is that the propagation of initially localized ictal activity to more distal cortical sites involves predominantly feedforward inhibition. This feedforward inhibitory drive could be a mechanism for the initial strengthening of inhibition during seizure propagation assumed in our model in two of the three seizure propagation scenarios (gamma and SWC with preceding fast oscillations). Furthermore, Trevelyan and colleagues (Trevelyan et al. 2006; Schevon et al. 2012) have emphasized the role of an “inhibitory restraint” or “veto” mechanism during neocortical propagation of focal seizures. Inhibitory veto of seizure propagation has been shown in low Mg 2+ mouse models of epilepsy (Trevelyan et al. 2006) and has been argued to be present also in human focal seizures (Schevon et al. 2012). This inhibitory veto mechanism would lead to “penumbra areas” ahead of the ictal wavefront. The failure of such inhibitory veto would allow the successful propagation of the ictal wavefront into new recruited ictal areas. Otherwise, the inhibition would be strong enough to contain the seizure spread. In this way, one could also argue that the gamma seizures examined here would represent a failure in seizure propagation. This possibility remains an open and important experimental question. Given that these sustained gamma oscillations reflect abnormal dynamics during secondarily generalized focal seizures, we currently consider this activity as propagated seizures. The gamma seizures in our model (Fig. 3) represent a situation in which there is a balance between enhanced inhibition and [K+]o accumulation. In other words, the seizure propagates while inhibitory interneuronal activity is preserved. In a related experimental study, our group (Lu et al. 2015) has shown that constant, strong enough depolarization driven by optogenetic stimulation can induce sustained (∼50 Hz) gamma oscillations in healthy primate motor cortex. In view of the strong inhibition in simulations of gamma seizures (Fig. 3), we do not preclude the possibility that some of the “high-activity” states in the simulated gamma seizures may in fact be long-lasting transients as opposed to asymptotic stable states. In other words, the elevated [K+]o in the “high-activity” states may nevertheless settle back to a much lower value after a long enough time. However, this issue has little effect on our main claims in this paper, given the long time scale (at least in the order of minutes) over which [K+]o remains elevated and increasing. Model limitations and future work Our analyses and simulations show how different types of seizure transitions observed in intracortical recordings of propagated human focal seizures can be accounted for by variations in synaptic inhibition and extracellular potassium concentration. Several limitations of the model, however, warrant future improvement. First, although the interaction of [K+]o and inhibitory synaptic dynamics can account for the phenomena observed in our human epilepsy data (Ahmed et al. 2014), several other ionic or synaptic mechanisms are possible and have been studied in different animal models of epilepsy (e.g. Alfonsa et al. 2015). Second, our model exploration was restricted to the transition into ictal states. In particular, we did not examine how seizures terminate, which is also a complex topic and may involve multiple ionic and synaptic components (such as glutamate depletion–Lado and Moshé 2008; GABA upregulation–Wen et al. 2015; fluctuations in ionic concentrations–Krishnan and Bazhenov 2011; Kramer et al. 2012). A more detailed model, including more complex ionic dynamics, will be required to address this issue. Furthermore, gap junction effects were not included in the model. Although electrical synapses have been suggested as an underlying mechanism for interneuronal bursting (Skinner et al. 1999) and synchronous population gamma oscillations (Traub et al. 2001), neither of these two features appeared to be prominent in our human data where neuronal spiking during gamma seizures is largely asynchronous. Thus, it remains an open question whether gap junction effects are critical to account for transitions to and maintenance of gamma seizures. Nevertheless, gap junctions may still play an important role during SWC discharges as previously hypothesized by Traub and colleagues (e.g. Traub et al. 2005). We hope to address these open issues in the future. Third, in the hypothesized contributing factors to seizure transitions examined here, we have remained agnostic about what ultimately leads to transient dysfunctions in inhibitory activity and in potassium glial-buffering processes. We note that several other factors involving changes in oxygen, Na+/K+ ATP pumps and cell volume (e.g. Ingram et al. 2014; Wei et al. 2014a), for example, can also contribute to elevations in extracellular potassium. We also emphasize that the goal in this study is to replicate the dynamics observed in propagated seizures, in other words, ictal dynamics in neocortical patches distal to the putative seizure onset zones. As described in Truccolo et al. (2011, 2014) and Wagner et al. (2015), our recorded neocortical patches were close, but distal to the putative seizure onset zones. The dynamics involving seizure initiation in seizure onset/focus areas might involve different mechanisms. As stated above, we have focused on fast time-scale changes in synaptic inhibition which have been reported in several previous animal studies and in vitro studies of human epileptic cortical tissue. Although these studies have provided some initial evidence for transient changes in inhibitory interneuron activity preceding seizures, it remains a major open question how these dysfunctions in interneuron network activity (and potassium glial buffering) would arise in the longer time scale of epileptogenesis. The authors would like to thank the editors, reviewer, Drs. F. Skinner, Liang Zhang, Felipe Gerhard and Timothée Proix for constructive comments on the manuscript. We also thank Drs. Sydney Cash, Omar Ahmed and Mark Kramer for valuable discussions. This research was conducted using computational resources and services at the Center for Computation and Visualization, Brown University. Compliance with Ethical Standards Conflict of interest The authors declare that they have no conflict of interest. This research was supported by: the National Institute of Neurological Disorders and Stroke (NINDS), grant R01NS079533; the U.S. Department of Veterans Affairs, Merit Review Award I01RX000668; and the Pablo J. Salame ‘88 Goldman Sachs endowed Assistant Professorship of Computational Neuroscience. The contents do not represent the views of the U.S. Department of Veterans Affairs or the United States Government. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. ==== Refs References Ahmed OJ Kramer MA Truccolo W Naftulin JS Donoghue JA Eskandar EN Cosgrove GR Blum AS Potter NS Hochberg LR Cash SS Inhibitory single neuron control of seizures and epileptic traveling waves in humans BMC Neuroscience 2014 15 Suppl 1 F3 10.1186/1471-2202-15-S1-F3 Alfonsa H Merricks EM Codadu NK Cunningham MO Deisseroth K Racca C Trevelyan AJ The contribution of raised intraneuronal chloride to epileptic network activity The Journal of Neuroscience 2015 35 20 7715 7726 10.1523/JNEUROSCI.4105-14.2015 25995461 Anderson WS Kudela P Cho J Bergey GK Franaszczuk PJ Studies of stimulus parameters for seizure disruption using neural network simulations Biological Cybernetics 2007 97 173 194 10.1007/s00422-007-0166-0 17619199 Anderson W Azhar F Kudela P Bergey G Franaszczuk P Epileptic seizures from abnormal networks: why some seizures defy predictability Epilepsy Research 2012 99 202 213 10.1016/j.eplepsyres.2011.11.006 22169211 Bazhenov M Timofeev I Steriade M Sejnowski TJ Potassium model for slow (2-3 Hz) in vivo neocortical paroxysmal oscillations Journal of Neurophysiology 2004 92 1116 1132 10.1152/jn.00529.2003 15056684 Brunel N Dynamics of sparsely connected networks of excitatory and inhibitory spiking neurons Journal of Computational Neuroscience 2000 8 183 208 10.1023/A:1008925309027 10809012 Brunel N Wang XJ What determines the frequency of fast network oscillations with irregular neural discharges? I. Synaptic dynamics and excitation-inhibition balance Journal of Neurophysiology 2003 90 415 430 10.1152/jn.01095.2002 12611969 Buxhoeveden DP Casanova MF The minicolumn hypothesis in neuroscience Brain: A Journal of Neurology 2002 125 935 951 10.1093/brain/awf110 11960884 Cammarota M Losi G Chiavegato A Zonta M Carmignoto G Fast spiking interneuron control of seizure propagation in a cortical slice model of focal epilepsy The Journal of Physiology 2013 591 4 807 822 10.1113/jphysiol.2012.238154 23207591 Cressman JRJr Ullah G Ziburkus J Schiff S Barreto E The influence of sodium and potassium dynamics on excitability, seizures, and the stability of persistent states: I. Single neuron dyamics Journal of Computational Neuroscience 2009 26 159 170 10.1007/s10827-008-0132-4 19169801 D’Antuono M Louvel J Köhling R Mattia D Bernasconi A Olivier A Turak B Devaux A Pumain R Avoli M Gabaa receptor-dependent synchronization leads to ictogenesis in the human dysplastic cortex Brain: A Journal of Neurology 2004 127 1626 1640 10.1093/brain/awh181 15175227 Dur-e-Ahmad M Nicola W Campbell S Skinner F Network bursting using experimentally constrained single compartment CA3 hippocampal neuron models with adaptation Journal of Computational Neuroscience 2012 33 1 21 40 10.1007/s10827-011-0372-6 22131133 Ermentrout, B. (2002). Simulating, analyzing, and animating dynamical systems: A guide to XPPAUT for researchers and students (SIAM). Ferguson K Njap F Nicola W Skinner F Campbell S Examining the limits of cellular adaptation bursting mechanisms in biologically-based excitatory networks of the hippocampus Journal of Computational Neuroscience 2015 39 289 309 10.1007/s10827-015-0577-1 26464038 Freyer F Aquino K Robinson PA Ritter P Breakspear M Bistability and non-Gaussian fluctuations in spontaneous cortical activity The Journal of Neuroscience 2009 29 26 8512 8524 10.1523/JNEUROSCI.0754-09.2009 19571142 Fröhlich F Bazhenov M Iragui-Madoz V Sejnowski TJ Potassium dynamics in the epileptic cortex: New insights on an old topic The Neuroscientist 2008 14 422 10.1177/1073858408317955 18997121 Fröhlich F Sejnowski TJ Bazhenov M Network bistability mediates spontaneous transitions between normal and pathological brain states The Journal of Neuroscience 2010 30 32 10734 10743 10.1523/JNEUROSCI.1239-10.2010 20702704 González-Ramírez LR Ahmed OJ Cash SS Wayne CE Kramer MA A biologically constrained, mathematical model of cortical wave propagation preceding seizure termination PLoS Computational Biology 2015 11 2 e1004065 10.1371/journal.pcbi.1004065 25689136 Grasse D Karunakaran S Moxon K Neuronal synchrony and the transition to spontaneous seizures Experimental Neurology 2013 248 72 84 10.1016/j.expneurol.2013.05.004 23707218 He BJ Scale-free brain activity: past, present, and future Trends in Cognitive Sciences 2014 18 480 487 10.1016/j.tics.2014.04.003 24788139 Ho ECY Strüber M Bartos M Zhang L Skinner FK Inhibitory networks of fast-spiking interneurons generate slow population activities due to excitatory fluctuations and network multistability The Journal of Neuroscience 2012 32 29 9931 9946 10.1523/JNEUROSCI.5446-11.2012 22815508 Ho ECY Eubanks JH Zhang L Skinner FK Network models predict that reduced excitatory fluctuations can give rise to hippocampal network hyper-excitability in meCP2-null mice PloS One 2014 9 e91148 10.1371/journal.pone.0091148 24642514 Hübel N Scholl E Dahlem MA Bistable dynamics underlying excitability of ion homeostasis in neuron models PLoS Computational Biology 2014 10 e1003551 10.1371/journal.pcbi.1003551 24784149 Ingram J Zhang C Cressman J Hazra A Wei Y Koo Y žIburkus J Kopelman R Xu J Schiff S Oxygen and seizure dynamics: I. experiments Journal of Neurophysiology 2014 112 205 212 10.1152/jn.00540.2013 24598521 Jirsa VK Stacey WC Quilichini PP Ivanov AI Bernard C On the nature of seizure dynamics Brain: A Journal of Neurology 2014 133 2210 2230 10.1093/brain/awu133 24919973 Kager H Wadman WJ Somjen GG Simulated seizures and spreading depression in a neuron model incorporating interstitial space and ion concentrations Journal of Neurophysiology 2000 84 495 512 10899222 Kim C Nykamp D Dynamics of a network of excitatory and inhibitory neurons induced by depolarization block BMC Neuroscience 2014 15 Suppl 1 P76 10.1186/1471-2202-15-S1-P76 Kramer MA Truccolo W Eden UT Lepage KQ Hochberg LR Eskandar EN Madsen JR Lee JW Maheshwari A Halgren E Chu CJ Cash SS Human seizures self-terminate across spatial scales via a critical transition Proc Natl Acad Sci U S A 2012 109 21116 21121 10.1073/pnas.1210047110 23213262 Krishnan GP Bazhenov M Ionic dynamics mediate spontaneous termination of seizures and postictal depression state The Journal of Neuroscience 2011 31 24 8870 8882 10.1523/JNEUROSCI.6200-10.2011 21677171 Krishnan G Filatov G Bazhenov M Dynamics of high-frequency synchronization during seizures Journal of Neurophysiology 2013 109 2423 2437 10.1152/jn.00761.2012 23427308 Lado FA Moshé SL How do seizures stop? Epilepsia 2008 49 1651 1664 10.1111/j.1528-1167.2008.01669.x 18503563 Lu Y Truccolo W Wagner F Vargas-Irwin C Ozden I Zimmermann J May T Agha N Wang J Nurmikko A Optogenetically induced spatiotemporal gamma oscillations and neuronal spiking activity in primate motor cortex Journal of Neurophysiology 2015 113 3574 3587 10.1152/jn.00792.2014 25761956 Naze S Bernard C Jirsa V Computational modeling of seizure dynamics using coupled neuronal networks: Factors shaping epileptiform activity PLoS Computational Biology 2015 11 e1004209 10.1371/journal.pcbi.1004209 25970348 Nicola, W., & Campbell, S. (2013a). Bifurcations of large networks of two-dimensional integrate and fire neurons. Journal of Computational Neuroscience, 35, 87–108. Nicola, W., & Campbell, S. A. (2013b). Mean-field models for heterogeneous networks of two-dimensional integrate and fire neurons. Frontiers in Computational Neuroscience, 7(184), 1–25. Nicola, W., & Campbell, S. (2014). Non-smooth bifurcations of mean field systems of two-dimensional integrate and fire neurons. arXiv:1408.4767. Park EH Durand DM Role of potassium lateral diffusion in non-synaptic epilepsy: a computational study Journal of Theoretical Biology 2006 238 666 682 10.1016/j.jtbi.2005.06.015 16085109 Paz J Huguenard JR Microcircuits and their interactions in epilepsy: is the focus out of focus? Nature Neuroscience 2015 18 351 359 10.1038/nn.3950 25710837 Perucca P Dubeau F Gotman J Intracranial electroencephalographic seizure-onset patterns: effect of underlying pathology Brain: A Journal of Neurology 2014 137 183 196 10.1093/brain/awt299 24176980 Richardson, K., Fanselow, E., & Connors, B. (2008). Neocortical anatomy and physiology. In Engel, J., & Pedley, T. (Eds.) Epilepsy: A Comprehensive Textbook (pp. 323–336): Lippincott-Williams & Wilkins. Rudolph M Piwkowska Z Badoual M Bal T Destexhe A A method to estimate synaptic conductances from membrane potential fluctuations Journal of Neurophysiology 2004 91 6 2884 2896 10.1152/jn.01223.2003 15136605 Sasaki T Matsuki N Ikegaya Y Metastability of active CA3 networks The Journal of Neuroscience 2007 27 3 517 528 10.1523/JNEUROSCI.4514-06.2007 17234584 Schevon CA Weiss SA McKhann GJr Goodman RR Yuste R Emerson RG Trevelyan AJ Evidence Of an inhibitory restraint of seizure activity in humans Nature Communications 2012 3 1060 1 11 Skinner, F. K. (2006). Conductance-based models. Scholarpedia 1:1408 revision #125663. Skinner FK Zhang L Perez-Velazquez JL Carlen PL Bursting in inhibitory interneuronal networks: a role for gap-junctional coupling Journal of Neurophysiology 1999 81 1274 1283 10085354 Sritharan D Sarma SV Fragility in dynamic networks: application to neural networks in the epileptic cortex Neural Computation 2014 26 10 2294 2327 10.1162/NECO_a_00644 25058705 Thurman DJ Beghi E Begley CE Berg AT Buchhalter JR Ding D Hesdorffer DC Hauser WA Kazis L Kobau R Kroner B Labiner D Liow K Logroscino G Medina MT Newton CR Parko K Paschal A Preux PM Sander JW Selassie A Theodore W Tomson T Wiebe S Standards for epidemiologic studies and surveillance of epilepsy Epilepsia 2011 52 2 26 10.1111/j.1528-1167.2011.03121.x 21899536 Touboul J Bifurcation analysis of a general class of nonlinear integrate-and-fire neurons SIAM Journal of Applied Mathematics 2008 68 1045 1079 10.1137/070687268 Traub RD Kopell N Bibbig A Buhl EH LeBeau FEN Whittington MA Gap junctions between interneuron dendrites can enhance synchrony of gamma oscillations in distributed networks The Journal of Neuroscience 2001 21 23 9478 9486 11717382 Traub RD Contreras D Cunningham MO Murray H LeBeau FEN Roopun A Bibbig A Wilent WB Higley MJ Whittington MA Single-column thalamocortical network model exhibiting gamma oscillations, sleep spindles, and epileptogenic bursts Journal of Neurophysiology 2005 93 4 2194 2232 10.1152/jn.00983.2004 15525801 Trevelyan AJ Sussillo D Watson BO Yuste R Modular propagation of epileptiform activity: evidence for an inhibitory veto in neocortex The Journal of Neuroscience 2006 26 48 12447 12455 10.1523/JNEUROSCI.2787-06.2006 17135406 Truccolo W Donoghue JA Hochberg LR Eskandar EN Madsen JR Anderson WS Brown EN Halgren E Cash SS Single-neuron dynamics in human focal epilepsy Nature Neuroscience 2011 14 5 635 643 10.1038/nn.2782 21441925 Truccolo W Ahmed OJ Harrison MT Eskandar EN Cosgrove GR Madsen JR Blum AS Potter NS Hochberg LR Cash SS Neuronal ensemble synchrony during human focal seizures The Journal of Neuroscience 2014 34 30 9927 9944 10.1523/JNEUROSCI.4567-13.2014 25057195 Ullah G Cressman JRJr Barreto E Schiff SJ The influence of sodium and potassium dynamics on excitability, seizures, and the stability of persistent states: II. Network and glial dynamics Journal of Computational Neuroscience 2009 29 171 183 10.1007/s10827-008-0130-6 19655238 Uva L Breschi GL Gnatkovsky V Taverna S de Curtis M Synchronous inhibitory potentials precede seizure-like events in acute models of focal limbic seizures The Journal of Neuroscience 2015 35 3048 3055 10.1523/JNEUROSCI.3692-14.2015 25698742 Van Ooyen A Van Pelt J Corner MA Lopes da Silva FH The emergence of long-lasting transients of activity in simple neural networks Biological Cybernetics 1992 67 269 277 10.1007/BF00204400 1498190 Van Vreeswijk C Hansel D Patterns of synchrony in neural networks with spike adaptation Neural Computation 2001 13 959 992 10.1162/08997660151134280 11359640 Wagner F Eskandar E Cosgrove G Madsen J Blum A Potter N Hochberg L Cash S Truccolo W Microscale spatiotemporal dynamics during neocortical propagation of human focal seizures Neuroimage 2015 122 114 130 10.1016/j.neuroimage.2015.08.019 26279211 Wang Y Goodfellow M Taylor PN Baier G Dynamic mechanisms of neocortical focal seizure onset PLoS Computational Biology 2014 10 e1003787 10.1371/journal.pcbi.1003787 25122455 Wang XJ Buzsáki G Gamma oscillation by synaptic inhibition in a hippocampal interneuronal network model The Journal of Neuroscience 1996 16 20 6402 6413 8815919 Wei, Y., Ullah, G., & Schiff, S. (2014a). Oxygen and seizure dynamics: II. computational modeling. Journal of Neurophysiology, 112, 213–223. Wei, Y., Ullah, G., & Schiff, S. (2014b). Unification of neuronal spikes, seizures, and spreading depression. The Journal of Neuroscience, 34, 11733–11743. Wen B Qian H Feng J Ge RJ Xu X Cui ZQ Zhu RY Pan LS Lin ZP Wang JH A portion of inhibitory neurons in human temporal lobe epilepsy are functionally upregulated: an endogenous mechanism for seizure termination CNS Neuroscience & Therapeutics 2015 21 204 214 10.1111/cns.12336 25475128 Zhang ZJ Koifman J Shin DS Ye H Florez CM Zhang L Valiante TA Carlen PL Transition to seizure: Ictal discharge is preceded by exhausted presynaptic GABA release in the hippocampal CA3 region The Journal of Neuroscience 2012 32 7 2499 2512 10.1523/JNEUROSCI.4247-11.2012 22396423 žiburkus J Cressman J Barreto E Schiff S Interneuron and pyramidal cell interplay during in vitro seizure-like events Journal of Neurophysiology 2006 95 3948 3954 10.1152/jn.01378.2005 16554499 žiburkus J Cressman JR Schiff SJ Seizures as imbalanced up states: excitatory and inhibitory conductances during seizure-like events Journal of Neurophysiology 2013 109 1296 1306 10.1152/jn.00232.2012 23221405

PMC005xxxxxx/PMC5002284.txt

==== Front J Parasitol ResJ Parasitol ResJPRJournal of Parasitology Research2090-00232090-0031Hindawi Publishing Corporation 10.1155/2016/1084353Research ArticleCould kDNA-PCR in Peripheral Blood Replace the Examination of Bone Marrow for the Diagnosis of Visceral Leishmaniasis? de Godoy Natalia Souza 1 Andrino Marcos Luiz Alves 1 de Souza Regina Maia 1 Gakiya Erika 1 Amato Valdir Sabbaga 1 2 Lindoso José Ângelo Lauletta 3 4 http://orcid.org/0000-0002-6493-465XAlmeida Braz Lucia Maria 5 * 1Laboratório de Investigação Médica, Parasitologia LIM 46 do Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo (HCFMUSP), Avenida Dr. Enéas Carvalho de Aguiar 470/500, 05403-000 São Paulo, SP, Brazil2Departamento de Moléstias Infecciosas e Parasitárias, HCFMUSP, Avenida Dr. Enéas Carvalho de Aguiar 255, 05403-000 São Paulo, SP, Brazil3Laboratório de Soroepidemiologia e Imunobiologia, IMTSP-USP, Avenida Dr. Enéas Carvalho de Aguiar 470/500, 05403-000 São Paulo, SP, Brazil4Hospital Emilio Ribas, Avenida Dr. Arnaldo 165, 01246-900 São Paulo, SP, Brazil5Laboratório de Parasitologia, Instituto de Medicina Tropical de São Paulo, Universidade de São Paulo (IMTSP-USP), Avenida Dr. Enéas Carvalho de Aguiar 470/500, 05403-000 São Paulo, SP, Brazil*Lucia Maria Almeida Braz: lmabraz@usp.brAcademic Editor: José F. Silveira 2016 14 8 2016 2016 10843531 4 2016 29 6 2016 10 7 2016 Copyright © 2016 Natalia Souza de Godoy et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.The aim of this study was to evaluate whether the molecular (kDNA-PCR) and parasitological diagnosis in peripheral blood (PB) could replace the invasive and painful bone marrow collection (BM) in the diagnosis of visceral leishmaniasis (VL). PB from suspected VL patients was evaluated by parasitological and molecular techniques using as the gold standard (GS) a combination of clinical, epidemiological, and immunochromatographic test (PB-rK39) results and parasitological examination of BM. Based on the GS, 38 samples from 32 patients were grouped: Group 1, 20 samples of VL cases, and Group 2, 18 samples of non-VL cases. In order to evaluate the parasitological and molecular techniques in PB, the samples were examined. From Group 1, PB kDNA-PCR was positive in 20 samples and in 19 of 20 in BM kDNA-PCR examination. However, the parasitological examination of buffy coat was insensitive, being able to detect only 4 cases from Group 1. All samples from Group 2 were negative. We concluded that, for the diagnosis of visceral leishmaniasis, the parasitological examination of peripheral blood was not useful; however, molecular diagnosis by kDNA-PCR, performed in peripheral blood, could be useful to replace the parasitological examination of bone marrow. Fundação de Amparo à Pesquisa do Estado de São Paulo2010-50304-8 ==== Body 1. Introduction Visceral leishmaniasis (VL) has been reported in 88 countries, and 90% of the world's burden is localized in India, Brazil, and Sudan [1]. In the Americas, VL is known as American Visceral Leishmaniasis (AVL) and its etiologic agent is Leishmania (Leishmania) infantum (L. (L.) chagasi, syn.) [2]. VL is characterized by its chronicity and systemic dissemination capacity [3] and it can be fatal if treatment is not administered. To be correctly diagnosed, VL requires the use of highly sensitive and specific laboratory methods [4, 5]. In the context of VL, the techniques that are considered gold standards (GS) are the direct microscopic examination of bone marrow (BM) or spleen aspirate samples, with observation of amastigotes on smears (BM-S), and the isolation of promastigotes in culture (BM-C) [6, 7]. More recently, the rK39 immunochromatographic test has been included [8]. Although splenic aspirate smears show the highest sensitivity, followed by BM aspirate smears, both exams require invasive and more risky procedures [9–11]. Given the fact that a considerable number of VL patients are children and immunocompromised patients, it would be desirable to adopt less risky and painful procedures provided that similar sensitivity and specificity rates are obtained. Therefore, the aim of this study was to evaluate whether parasitological and molecular techniques performed in the BC (buffy coat) or WB (whole blood), from a small volume of peripheral blood (PB), could replace the parasitological examination of the bone marrow (BM), whose collection is invasive and painful, for the diagnosis of visceral leishmaniasis (VL). Peripheral blood (PB) is a biological material that can be easily collected and used in highly sensitive tests to investigate VL such as Leishmania DNA amplification by PCR performed in whole blood or buffy coat samples [4] and serological detection of anti-Leishmania antibodies [6]. Using samples of whole blood (WB) and buffy coat (BC) from patients with and without VL, parasitological and molecular techniques were evaluated. These samples were used to prepare smears (PB-S) and inoculated in NNN medium supplemented with BHI (PB-C). In addition, a kDNA-PCR was performed in WB and BC samples (PB kDNA-PCR) as well as in BM samples (BM kDNA-PCR). The performances of these techniques were compared to the GS laboratorial techniques (PB-rK39; BM-S and BM-C) results. 2. Methods 2.1. Patients This research was approved by the Institutional Research Ethics Committee (protocol number 0006/11). After signing the informed consent, participants suspected to have VL coming from different regions of Brazil were attended at the University Hospital (HC-FMUSP) of São Paulo. From 77 samples examined, we selected 38 samples from 32 patients, whose samples from BM and PB were matched. 2.2. Laboratorial Gold Standards (GS) and Definition of VL Cases The laboratorial GS were the parasitological techniques (direct microscopy examination and culture) of BM samples, aside from the immunochromatographic diagnostic test (rK39) in whole blood PB samples (PB-rK39). To be considered a true VL case the participant had to present clinical test, epidemiological test, and at least one positive test among the laboratorial GS. Thirty-eight samples from 32 patients met the inclusion criteria of the study and were grouped as follows:  Group 1: 20 samples from 18 true VL cases according to the clinical and epidemiological data associated with a positive result in at least one GS laboratorial techniques.   Group 2: 18 samples from 16 non-VL cases presented symptoms initially compatible with VL and had negative results in the GS laboratorial techniques. 2.3. Collection and Processing of Biological Samples BM aspirates and PB samples (3.5 mL) were collected in EDTA tubes and submitted to the following diagnostic procedures. 2.3.1. Parasitological Techniques (Gold Standard When Tested in Bone Marrow) Microscopy Examination of the Stained Smear from Bone Marrow (BM-S) and from Peripheral Blood Samples PB (PB-S). Five microliters of BM and PB samples (WB and BC at 1,506 ×g and 20,000 ×g, resp.) was used to prepare eight slides smears. From each procedure (BM, WB PB, and BCPB1 at 1,506 ×g and BCPB2 at 20,000 ×g) 2 slides were prepared, which were subsequently stained by Panótico method (Newprov Instant Prov, Paraná, Brazil), a kind of Leishman or Romanowsky dye. They were analyzed by microscopy (1000x magnification), searching for amastigotes, and 200 fields were examined in each smear [12]. Culture of BM (BM-C) and PB (PB-C) in NNN Medium Supplemented with BHI. Forty microliters of BM or the same volume of BC (1,506 ×g and 20,000 ×g) from PB samples was inoculated into 2 tubes, containing NNN medium (DIFCO, USA), to which 2 mL of BHI medium was added (DIFCO, USA). The tubes were incubated in a BOD incubator at 25°C and the samples were weekly analyzed for a total of 30 days by means of optical microscopy searching for promastigotes [13]. 2.3.2. Molecular Diagnostic Techniques of Peripheral Blood (PB) and Bone Marrow (BM) Polymerase Chain Reaction: PB kDNA-PCR and BM kDNA-PCR WB and BC obtained by centrifugation (1,506 ×g and 20,000 ×g) of PB samples as well as BM samples were submitted to DNA extraction with the aid of the Genomic DNA Extraction Kit (Real Biotech Corporation, Taiwan, China) starting with an initial volume of 300 μL. DNA samples were identified and stored at −20°C. The kDNA primers were designed within a conserved region of Leishmania sp. kDNA minicircles. The forward primer 20 (5′-GGGKAGGGGCGTTCTSCGA A-3′) and reverse primer 22 (5′-SSSWCTATWTTACACCAACCCC-3′) yielded a 120-base-pair amplification product [14]. Amplifications were performed in a total volume of 20 μL containing 100 ng of template DNA, 50 mM of KCl, 10 mM of Tris-HCl (pH 8.0), 0.2 mM of dNTPs (Fermentas, Thermo Fisher, Ontario, Canada), 1.0 mM of MgCl2, 0.4 μM of each primer, and 1 unit of Taq DNA Polymerase (Fermentas, Thermo Fisher, Ontario, Canada). In each experiment, two negative controls containing sterile water instead of template DNA were also tested. The positive control was Leishmania infantum DNA extracted from cultures. Each reaction was performed with an initial denaturation step of 94°C for 5 minutes, followed by 35 cycles of 94°C for 1 minute, 58°C for 1 minute, and 72°C for 30 seconds, ending with a final extension step of 72°C for 5 minutes. Reactions were performed in a minicycler thermocycler (MJ Research Corp/MJ Research, Quebec, Canada) and PCR products were visualized in 2% ethidium bromide-stained agarose gels by means of a UV transilluminator (Alpha Innotech Corp/Alpha Innotech Multimage, California, USA). To minimize the risk of contamination, reagents preparation and PCR master mix, DNA extraction, and electrophoresis were performed in three separate areas. To confirm that amplification inhibitors were not present, a fragment of the human beta-globin gene was tested in all the samples [15]. 2.3.3. Serological Technique Whole PB samples were tested by the PB-rK39 immunochromatographic kit that uses a recombinant peptide containing 39 amino acid repeats from the kinesin-like gene found in L. chagasi. This test is widely used for VL diagnosis in field studies [11]. 2.4. Statistical Analysis To determine the agreement of tests, the kappa test was used at a 95% confidence interval and p values ≤0.05 were considered significant. The data were analyzed with the statistical software STATA version 12.0 (Stata Corp LP, College Station, Texas, USA). 3. Results Regarding the GS techniques (Table 1), performed in Group 1 (20 true VL samples), we obtained the following results: 16 samples were positive by BM-S, including 8 also by BM-C; 15 were positive by PB-rK39 (four of which were negative by parasitological examination of BM), though there were five samples, from two HIV positive patients, that yielded negative results by PB-rK39 and were positive by parasitological examination of BM. Group 2 was composed of 18 samples from 16 non-VL cases that were negative by all the gold standard techniques (BM-S, BM-C, and PB-rK39). The clinical follow-up of these 16 non-VL cases revealed, in most of them, cutaneous leishmaniasis (1), lupus (2), prostatitis (1), hepatitis (1), urinary tract infection (1), Sjogren syndrome (1), lymphoma (2), gastric ulcers (2), sarcoidosis (1), pellagra (1), pharyngitis (1), hypothyroidism (1), and anemia (1). All DNA samples were amplified by the beta-globin gene demonstrating that there were no amplification inhibitors. According to Table 1, PB kDNA-PCR, performed in whole blood and buffy coat, was positive in 100% of Group 1 samples (20/20). PB-S was positive in 20% of the samples (4/20), and PB-C was negative in all samples. The four PB-S positive samples were obtained from BCPB1 (one buffy coat sample centrifuged at 1,506 ×g) and BCPB2 (two buffy coat samples centrifuged at 20,000 ×g) and one in BCPB1 and BCPB2 (buffy coat sample tested after centrifugation at 1,506 ×g and 20,000 ×g). It is noteworthy that none of these four samples was parasitologically positive when the analysis was performed directly in the corresponding whole blood samples. Analyzing the PB kDNA-PCR and BM kDNA-PCR they presented a specificity of 100% and a sensitivity of 100% and 95%, respectively (Table 2). A good concordance, with a kappa index of 0,79 and 0,74, was obtained (p < 0.001), when PB kDNA-PCR was compared with the gold standard techniques: the BM-S and the rK39, respectively (Table 3). Comparison between PB kDNA-PCR and BM kDNA-PCR results showed an almost perfect correlation (0.94, p < 0.001) (Table 3). 4. Discussion In this study PB samples were tested as a substitute of invasive and painful procedures to obtain BM samples for the diagnosis of VL. Abeijon and Campos-Neto [16] investigated a potential noninvasive urine-based antigen detection assay to diagnose active VL. One of the advantages of PB samples to investigate VL is that the same biological material can be concomitantly analyzed by parasitological, serological, and molecular techniques unlike urine samples. The main symptoms of VL patients in our study were hepatosplenomegaly, febrile, and pancytopenia as presented for all the patients in Group 1 (true VL cases, Table 1). They were all Brazilians and came from VL endemic regions [2, 3]. Regarding the parasitological investigation in peripheral blood, buffy coat PB-S detected 20% of positive samples, while none of the samples were positive when whole blood samples were examined, as expected, because parasites are concentrated in buffy coat facilitating the visualization of amastigotes within leucocytes [4]. Nevertheless, 20% detection is very low when compared to BM-S, one of the gold standards (80% of positivity). Similarly, PB-C did not find any positive sample while BM-C detected 40%. Sixteen out of 20 samples were negative by PB-S, and this low sensitivity can be explained by the fact that there are more parasites in the bone marrow and spleen samples [6, 7], justifying why they are the gold standard of VL laboratorial investigation. In the present study, cultures were not a sensitive method compared to other techniques, even in the case of BM, as this exam was able to confirm only 40% of truly positive samples. Aside from being very prone to contamination, cultures are time consuming requiring four weeks to release a final result, what has already been acknowledged by other authors [17] and was corroborated in this research. According to the ideal molecular target for detecting Leishmania, kDNA was chosen due to the large presence of minicircles in the cells of the parasite, about 104 copies per cell [17]. Moreover, the proven viability of parasites in PB, which is a biological material, obtained more easily than the BM leads us to search for kDNA in the PB. Nevertheless, differently of parasitological investigation in PB, molecular investigation (kDNA) in PB did not present differences between buffy coat (BCPB: 1506 ×g and BCPB: 20000 ×g) and whole blood. All samples 20/20 were positive. According to Srivastava et al. [4] PCR analysis of the whole blood or its buffy coat preparation may prove a useful screening test. A sensitivity of 100% was obtained by PB kDNA-PCR when compared to the gold standard techniques (BM-S, BM-C, and PB-rK39). By comparing molecular and parasitological techniques, Ozerdem et al. [18] obtained better results with kDNA-PCR (29/50 or 58%) in comparison with microscopic examination (10/50 or 20%) of Giemsa-stained smears from blood samples of suspected VL patients. In our study, PB kDNA-PCR showed a good concordance with the rapid immunochromatographic test (PB-rK39) (0.74, p < 0.001), which is a rapid and highly sensitive technique, so that it has been used as a reference test. Disch et al. [9] and Andresen et al. [19] obtained sensitivities of 91% (48/53) and 92.5% (37/40) when kDNA-PCR was used to test whole blood from PB samples of patients with VL, confirmed by clinical and microscopic examination of BM or lymph node samples. Fraga et al. [20] evaluated the effectiveness of a kDNA-PCR in PB and found a very good sensitivity (43/45; 95.6%), which was higher than that found in BM samples: kDNA (41/45; 91.1%); microscopic examination of smear (36/45; 80%); and culture (12/45; 26.7%). Antinori et al. [21] used PCR and obtained a sensitivity of 98.5% (64/65) and 95.7% (45/47) in PB and BM, respectively, once again confirming the better sensitivity of PB-PCR (whole blood) in comparison with BM-PCR. In contrast, Cruz et al. [10] found a higher positivity of Ln-PCR in BM (24/24) than in whole blood of PB (Ln-PCR) (19/24). By means of immunochromatography test, due to its sensitivity and rapidity, the PB-rK39 can screen suspected cases of VL, especially in immunocompetent patients. Although it is not our main objective, it is necessary to reinforce that the presence of VL antibodies, associated with clinical and epidemiological data, can assist in prompt medical decisions, but it cannot differentiate past from active infection. In order to diagnose an active VL infection in PB samples, from immunocompetent or immunodeficient patients, kDNA-PCR is the most appropriate. Unlike the BM, it also presents the advantage of being an easy to take biological sample. Nonetheless, parasitological examination of the peripheral blood (PB-S and PB-C) cannot substitute the parasitological examination of bone marrow (BM-S and BM-C). In conclusion, kDNA-PCR performed in small volumes of PB, in either whole blood or buffy coat, showed a good agreement with VL gold standard tests. Therefore peripheral blood could be useful to replace the invasive and painful procedures to obtain bone marrow samples for the diagnosis of visceral leishmaniasis. Acknowledgments The authors thank Professor Dr. Thelma S. Okay for her help in correction of the English language and Edite H. Y. Kanashiro, M.S., for providing help in the parasitological examination, the “Laboratório de Micologia (LIM-53)” and the “Laboratório de Bacteriologia (LIM 54)”. The research project was conducted fully sponsored by FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo) with Grant no. 2010-50304-8 under the supervision of Lucia Maria Almeida Braz. Ethical Approval Ethics Committee of HCFMUSP-BRASIL (given number 0006/11 CAPPesq) approved this study. Competing Interests The authors declare that they have no competing interests. Table 1 Results of clinical, epidemiological, and laboratorial diagnosis of visceral leishmaniasis in Group 1 (true VL cases) (1). Number of samples (N), age (A), and sample entrance (SE) Comorbidities (CM) Clinical and laboratorial manifestations Epidemiological data (place of origin) Laboratorial investigation for VL (gold standard) Laboratorial investigation for VL (tests) Parasitological techniques in bone marrow (BM) Serological technique (rK39) in peripheral blood (PB) Parasitological techniques in peripheral blood (PB) Molecular technique (kDNA) in bone marrow (BM) and peripheral blood (PB) BM-S BM-C PB-rK39 PB-S PB-C BM-kDNA PB-kDNA N: 1.1 A: 19 SE: 07/30/2009 Systemic lupus erythematosus Hepatosplenomegaly and pancytopenia Bahia Sister and dog with VL Neg Neg Pos Neg Neg Pos Pos N: 1.2 A: 20 SE: 05/25/2011 Systemic lupus erythematosus VL with death Hepatosplenomegaly and pancytopenia Bahia Pos Neg Pos Neg Neg Pos Pos N: 2.1 A: 31 SE: 08/14/2009 HIV Hepatosplenomegaly and fever Mato Grosso Pos Neg Neg Neg Neg Pos Pos N: 2.2 A: 33 SE: 02/22/2011 HIV Hepatosplenomegaly and fever Mato Grosso Pos Neg Neg Neg Neg Pos Pos N: 2.3 A: 33 SE: 03/11/2011 HIV Hepatosplenomegaly and fever Mato Grosso Pos Neg Neg Neg Neg Pos Pos N: 2.4 A: 33 SE: 03/30/2011 HIV Hepatosplenomegaly and fever Mato Grosso Pos Neg Neg Neg Neg Pos Pos N: 3 A: 26 SE: 09/18/2009 HIV VL severe with death Hepatosplenomegaly and fever São Paulo Pos Neg Neg Neg Neg Neg Pos N: 4 A: 69 SE: 12/07/2009   Hepatosplenomegaly Ceará Neg Neg Pos Neg Neg Pos Pos N: 5 A: 30 SE: 02/01/2010 No Hepatosplenomegaly, fever, and weight of loss Bahia Neg Neg Pos Neg Neg Pos Pos N: 6 A: 74 SE: 04/16/2010 No VL with death Pancytopenia and fever Minas Gerais Pos Pos Pos Neg Neg Pos Pos N: 7 A: 40 SE: 11/16/2010 No Splenomegaly, plateletopenia, anaemia, and fever Pernambuco Pos Pos Pos Neg Neg Pos Pos N: 8 A: 29 SE: 05/11/2011 Schistosomiasis Hepatosplenomegaly and fever Bahia Pos Pos Pos Pos Neg Pos Pos N: 9 A: 65 SE: 06/28/2011 Hemophagocytic syndrome Hepatosplenomegaly, fever, and pancytopenia Bahia Pos Pos Pos Pos Neg Pos Pos N: 10 A: 53 SE: 08/23/2011 Schistosomiasis hanseniasis Pancytopenia Alagoas Pos Pos Pos Pos Neg Pos Pos N: 11 A: 33 SE: 08/23/2011 No Hepatosplenomegaly and fever Bahia Pos Pos Pos Neg Neg Pos Pos N: 12 A: 4 SE: 08/16/2011 No Hepatosplenomegaly, fever pancytopenia, anaemia, weight loss, and lymphadenomegaly Mato Grosso Neg Neg Pos Neg Neg Pos Pos N: 13 A: 2 SE: 08/25/2011 No Hepatosplenomegaly, fever, and pancytopenia Piauí Pos Pos Pos Pos Neg Pos Pos N: 14 A: 48 SE: 08/29/2011 No Hepatosplenomegaly, fever, and pancytopenia Bahia Pos Neg Pos Neg Neg Pos Pos N: 15 A: 45 SE: 11/11/2011 No Hepatosplenomegaly, fever, and weight loss Bahia Pos Pos Pos Neg Neg Pos Pos N: 16 A: 20 SE: 03/27/2011 No Hepatosplenomegaly and fever Bahia Pos Neg Pos Neg Neg Pos Pos BM-S: smear with sample from bone marrow; BM-C: culture with sample from bone marrow; PB-S: smear with sample from peripheral blood; PB-C: culture with sample from peripheral blood; PB-rK39: rK39 with sample from peripheral blood; BM-kDNA: kDNA PCR in sample from bone marrow; PB-kDNA: kDNA PCR in sample from peripheral blood; pos: positive; neg: negative. Table 2 Sensitivity, specificity, predictive positive value, predictive negative value, probability of false positive, probability of false negative of PB kDNA-PCR and BM kDNA-PCR. Test Sensitivity (%) Specificity (%) PPV (%) PNV (%) PFP (%) PFN (%) Efficiency (%) (CI 95%) PB kDNA-PCR 100 100 100 100 0 0 100 (90.7–100) BM kDNA-PCR 95.0 100 100 95.2 0 4.8 97.3 (86.2–99.9) BM kDNA-PCR: kDNA PCR in sample from bone marrow; PB kDNA-PCR: kDNA PCR in sample from peripheral blood; PPV: predictive positive value; PNV: predictive negative value; PFP: probability of false positive; PFN: probability of false negative; CI 95%: 95% confidence intervals. Table 3 Comparative analysis between the results of PB kDNA-PCR and results of BM kDNA-PCR, BM-S, BM-C, and PB-rk39. Test PB kDNA-PCR Kappa (CI 95%) p value Efficiency (%) (CI 95%) Positive Negative Total BM kDNA-PCR       0.947 <0.001 97.4 (86.2–99.9) Positive 19 0 19 (0.846–1.000)     Negative 1 18 19 Almost perfect correlation     Total 20 18 38     BM-S       0.791 <0.001 89.5 (75.2–97.1) Positive 16 0 16 (0.602–0.980)     Negative 4 18 22 Good concordance     Total 20 18 38     BM-C       0.387 0.001 68.4 (51.3–82.5) Positive 8 0 8 (0.160–0.614)     Negative 12 18 30 Low concordance     Total 20 18 38     PB-rK39       0.740 <0.001 86.4 (71.9–95.6) Positive 15 0 15 (0.534–0.945)     Negative 5 18 23 Good concordance     Total 20 18 38     BM kDNA-PCR: kDNA PCR in sample from bone marrow; PB kDNA-PCR: kDNA PCR in sample from peripheral blood; PB-rk39: rK39 with sample from peripheral blood; CI 95%: 95% confidence intervals. ==== Refs 1 Bhattacharya S. K. Sur D. Sinha P. K. Karbwang J. Elimination of leishmaniasis (kala-azar) from the Indian subcontinent is technically feasible &operationally achievable Indian Journal of Medical Research 2006 123 3 195 196 2-s2.0-33745190020 16778303 2 Michel G. Pomares C. Ferrua B. Marty P. Importance of worldwide asymptomatic carriers of Leishmania infantum (L. chagasi ) in human Acta Tropica 2011 119 2-3 69 75 10.1016/j.actatropica.2011.05.012 2-s2.0-79960598850 21679680 3 Maia-Elkhoury A. N. S. Alves W. A. De Sousa-Gomes M. L. De Sena J. M. Luna E. A. Visceral leishmaniasis in Brazil: trends and challenges Cadernos de Saude Publica 2008 24 12 2941 2947 10.1590/s0102-311x2008001200024 2-s2.0-57749119428 19082286 4 Srivastava P. Dayama A. Mehtotra S. Sundar S. Diagnosisof visceral leishmaniasis Transactions of the Royal Society of Tropical Medicine and Hygiene 2011 105 1 1 6 21074233 5 Chappuis F. Sundar S. Hailu A. Visceral leishmaniasis: what are the needs for diagnosis, treatment and control? Nature Reviews Microbiology 2007 5 11 873 882 10.1038/nrmicro1748 2-s2.0-35348871028 17938629 6 de Assis T. S. M. Caligiorne R. B. Romero G. A. S. Rabello A. Detection of Leishmania kDNA in human serum samples for the diagnosis of visceral leishmaniasis Transactions of the Royal Society of Tropical Medicine and Hygiene 2009 103 12 1269 1272 10.1016/j.trstmh.2008.07.013 2-s2.0-70350621374 18774151 7 Rocha M. N. Margonari C. Presot I. M. Soares R. P. Evaluation of polymerase chain reaction protocols for cultured Leishmania spp. typing Diagnostic Microbiology and Infectious Disease 2010 68 4 401 409 10.1016/j.diagmicrobio.2010.08.007 2-s2.0-78649331772 20926219 8 Assis T. S. M. Braga A. S. C. Pedras M. J. Validation of the rapid immunochromatographic test IT-LEISH® for the diagnosis of human visceral leishmaniasis Epidemiologia e Serviços de Saúde 2008 17 2 107 116 9 Disch J. Maciel F. C. de Oliveira M. C. Orsini M. Rabello A. Detection of circulating Leishmania chagasi DNA for the non-invasive diagnosis of human infection Transactions of the Royal Society of Tropical Medicine and Hygiene 2003 97 4 391 395 10.1016/s0035-9203(03)90066-6 2-s2.0-7744240008 15259464 10 Cruz I. Chicharro C. Nieto J. Comparison of new diagnostic tools for management of pediatric mediterranean visceral leishmaniasis Journal of Clinical Microbiology 2006 44 7 2343 2347 10.1128/JCM.02297-05 2-s2.0-33746255296 16825347 11 Ritmeijer K. Melaku Y. Mueller M. Kipngetich S. O'Keeffe C. Davidson R. N. Evaluation of a new recombinant K39 rapid diagnostic test for Sudanese visceral leishmaniasis American Journal of Tropical Medicine and Hygiene 2006 74 1 76 80 2-s2.0-33646012333 16407349 12 São Paulo Manual de vigilância e controle da leishmaniose visceral americana do estado de São Paulo Secretaria de Estado da Saúde de São Paulo, Superintendência de Controle de Endemias/SUCEN, e Coordenadoria de Controle de Doenças/CCD 2006 Vera Lucia Fonseca de Camargo-Neves 13 Boggild A. K. Miranda-Verastegui C. Espinosa D. Optimization of microculture and evaluation of miniculture for the isolation of Leishmania parasites from cutaneous lesions in Peru The American Journal of Tropical Medicine and Hygiene 2008 79 6 847 852 2-s2.0-57649211318 19052291 14 Nicodemo A. C. Amato V. S. Tuon F. F. Souza R. M. Okay T. S. Braz L. M. Usefulness of kDNA PCR in the diagnosis of visceral leishmaniasis reactivation in co-infected patients Revista do Instituto de Medicina Tropical de São Paulo 2013 55 6 429 431 10.1590/s0036-46652013000600011 24213198 15 Nakajima-Iijima S. Hamada H. Reddy P. Kakunaga T. Molecular structure of the human cytoplasmic β -actin gene: Interspecies homology of sequences in the introns Proceedings of the National Academy of Sciences of the United States of America 1985 82 18 6133 6137 10.1073/pnas.82.18.6133 2-s2.0-0011238681 2994062 16 Abeijon C. Campos-Neto A. Potential non-invasive urine-based antigen (Protein) detection assay to diagnose active visceral leishmaniasis PLoS Neglected Tropical Diseases 2013 7 5 e2161 10.1371/journal.pntd.0002161 2-s2.0-84878472590 17 Schönian G. Nasereddin A. Dinse N. PCR diagnosis and characterization of Leishmania in local and imported clinical samples Diagnostic Microbiology and Infectious Disease 2003 47 1 349 358 12967749 18 Ozerdem D. Eroglu F. Genc A. Demirkazik M. Koltas I. S. Comparison of microscopic examination, rK39, and PCR for visceral leishmaniasis diagnosis in Turkey Parasitology Research 2009 106 1 197 200 10.1007/s00436-009-1650-3 2-s2.0-71449110996 19859739 19 Andresen K. Gasim S. Elhassan A. M. Diagnosis of visceral leishmaniasis by the polymerase chain reaction using blood, bone marrow and lymph node samples from patients from the Sudan Tropical Medicine and International Health 1997 2 5 440 444 10.1111/j.1365-3156.1997.tb00166.x 2-s2.0-0030975831 9217699 20 Fraga T. L. Brustoloni Y. M. Lima R. B. Polymerase chain reaction of peripheral blood as a tool for the diagnosis of visceral leishmaniasis in children Memorias do Instituto Oswaldo Cruz 2010 105 3 310 313 10.1590/S0074-02762010000300011 2-s2.0-77953491536 20512245 21 Antinori S. Calattini S. Longhi E. Clinical use of polymerase chain reaction performed on peripheral blood and bone marrow samples for the diagnosis and monitoring of visceral leishmaniasis in HIV-infected and HIV-uninfected patients: a single-center, 8-year experience in italy and review of the literature Clinical Infectious Diseases 2007 44 12 1602 1610 10.1086/518167 2-s2.0-34249872216 17516404

PMC005xxxxxx/PMC5002285.txt

==== Front Dis MarkersDis. MarkersDMDisease Markers0278-02401875-8630Hindawi Publishing Corporation 10.1155/2016/1868739Research ArticleOsteopontin in relation to Prognosis following Coronary Artery Bypass Graft Surgery Sbarouni Eftihia 1 http://orcid.org/0000-0002-5436-3144Georgiadou Panagiota 1 * Chatzikyriakou Sofia 1 Analitis Antonis 2 Chaidaroglou Antigoni 3 Degiannis Demitris 3 Voudris Vassilis 1 12nd Division of Interventional Cardiology, Onassis Cardiac Surgery Center, 17674 Athens, Greece2Department of Hygiene, Epidemiology and Medical Statistics, Medical School, University of Athens, 11527 Athens, Greece3Molecular Immunopathology and Histocompatibility Laboratory, Onassis Cardiac Surgery Center, 17674 Athens, Greece*Panagiota Georgiadou: georgiadou@ocsc.grAcademic Editor: Dennis W. T. Nilsen 2016 14 8 2016 2016 186873916 5 2016 18 7 2016 Copyright © 2016 Eftihia Sbarouni et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Cardiovascular events may occur even after complete revascularization in patients with coronary artery disease. We measured preoperative osteopontin (OPN) levels in 131 consecutive patients (66.5 ± 10 years old, 117 men and 14 women) with left ventricular ejection fraction of 50.7 ± 9.2% and low logistic EuroScore (3.5 ± 3.2%) undergoing elective Coronary Artery Bypass Grafting (CABG) surgery. Patients were prospectively followed up for a median of 12 months (range 11–24). The primary study endpoint was the composite of cardiovascular death, nonfatal myocardial infarction, need for repeat revascularization, and hospitalization for cardiovascular events. Pre-op OPN plasma levels were 77.9 (49.5, 150.9). Patients with prior acute myocardial infarction (AMI) had significantly higher OPN levels compared to those without [131.5 (52.2, 219) versus 73.3 (45.1, 125), p = 0.007]. OPN levels were positively related to EuroScore (r = 0.2, p = 0.031). Pre-op OPN levels did not differ between patients who had a major adverse event during follow-up compared to those with no event (p = 0.209) and had no effect on the hazard of future adverse cardiac events [HR (95% CI): 1.48 (0.43–4.99), p = 0.527]. The history of AMI was associated with increased risk of subsequent cardiovascular events at follow-up (p = 0.02). OPN is associated with preoperative risk assessment prior to low-risk CABG but did not independently predict outcome. ==== Body 1. Introduction Osteopontin (OPN) is a matrix cellular protein and cytokine, released when tissue injury and remodeling occur in various organs including the heart [1]. OPN has been isolated in human aorta, carotid, and coronary arteries and has been shown to be involved in the development and progression of atherosclerosis [2]. OPN may have opposite effects during several cardiovascular diseases [1]. OPN seems to affect prognosis in coronary artery disease (CAD), including acute coronary syndromes (ACS) and chronic stable angina [3, 4], and in heart failure [5]. Cardiovascular events may occur even after complete revascularization and there is constant interest in independent markers of subsequent risk including biomarkers [6]. We sought to investigate whether OPN pre-op has prognostic value for patients following elective Coronary Artery Bypass Grafting (CABG) surgery. 2. Materials and Methods 2.1. Study Population We studied 131 consecutive patients undergoing elective CABG in our institution. Patients with acute or chronic inflammatory disease, immunological disease, any cancer, and administration of any vitamin supplement during the last 6 months were excluded from the study. All patients gave informed consent and the study protocol was approved by the Ethics Committee of our institution. There were 117 men and 14 women with a mean age of 66.5 ± 10 years (Table 1). Demographic characteristics in relation to risk factors for CAD are shown in Table 2. Six (4.7%) of the diabetics were treated with insulin. Regarding previous acute myocardial infarction (AMI) and revascularization procedures, 42 (32.6%) had a history of AMI and 31 (23.8%) had undergone percutaneous coronary intervention (PCI) and 5 (3.9%) CABG. Almost a fourth, 29 (22.3%), suffered an ACS preoperatively. The majority of our patients presented with significant left main or 3-vessel disease, 98 (76%), and only 29 (22.5%) and 2 (1.6%) were operated on for 2- or 1-vessel disease, respectively. The study sample was relatively at low risk with preserved ejection fraction (50.7 ± 9.2%) and low logistic EuroSCORE (European system for cardiac operative risk evaluation) (3.5 ± 3.2%). 2.2. Biochemical Analysis Preoperative blood samples were obtained for OPN evaluation. All samples were centrifuged for 10 min in 3000 rounds per min within 15 min of sampling and stored at −80°C until assayed. Plasma OPN levels were determined by an enzyme-linked immunosorbent assay (ELISA) using the Quantikine human OPN kit (R&D Systems, UK) and are expressed in ng/mL. The average of duplicate readings for each standard, control, and sample was used. The standard curve was created by using a four-parameter logistic curve fit. Samples were usually diluted 25 times. The mean minimum detectable dose of OPN was 0.011 ng/mL. The intra- and interassay coefficients of variations were 4% and 7%, respectively. 2.3. Patient Follow-Up All patients but one were operated on on-pump. The number of grafts inserted and the cross-clumping and cardiopulmonary bypass time are shown in Table 1. Median hospital stay was 8 days (range 3–28). There was only one in-hospital death. Two patients developed acute renal failure and improved on dialysis and one suffered tamponade which was drained successfully. Seven patients had chronic atrial fibrillation; of those with sinus rhythm, 44 patients developed atrial fibrillation postoperatively, which was successfully cardioverted with intravenous amiodarone. All patients were prospectively followed up for a median of 12 months (m) (range 11–24). The primary study endpoint was the composite of cardiovascular death, nonfatal myocardial infarction, need for repeat revascularization, and hospitalization for cardiovascular events. Nine patients reached the primary endpoint during follow-up. There was one sudden cardiac death 9 m post-op. One patient suffered a nonfatal non-ST elevation myocardial infarction, 18 m post-op. We recorded 4 cases of repeat revascularization procedures: one PCI to circumflex artery 1 m post-op due to ACS and incomplete revascularization of the dominant circumflex, one PCI to the ungrafted right coronary artery which was disease-free at the time of the operation 11 m post-op and PCI to the left descending anterior artery distal to the internal mammary artery anastomosis, both at the same stage, and two PCIs to the intermediate and the first obtuse marginal, respectively, both new lesions in relation to the pre-op angiogram. Hospitalization occurred as follows: one patient was operated on for peripheral arterial disease with aortofemoral bypass 9 m post-op, another patient developed an embolic event in his right leg 2 m post-op, which was treated percutaneously, and one patient underwent coronary angiography for unstable angina and was subsequently managed medically. In addition, 3 patients were hospitalized for respiratory infections and 4 for minor surgery, 4 were operated on for cancer, and 1 was admitted for renal failure which improved with medical therapy and 1 patient for worsening heart failure, who was managed conservatively. Paroxysmal atrial fibrillation occurred in 3 patients; all of them had atrial fibrillation immediately postoperatively. 2.4. Statistical Analysis Association between categorical variables was tested using chi-square or Fisher's exact test. Spearman correlation was applied to investigate associations between OPN and continuous variables. Cox proportional hazard models were used to estimate the potential effect of OPN on the hazard of major cardiovascular event. All tests were two-sided at a = 5% level of statistical significance. 3. Results Pre-op OPN plasma levels were 77.9 (49.5, 150.9) [median (interquartile range 25th–75th percentiles)]. There was no difference in OPN levels in relation to age or gender and the prevalence of the examined cardiovascular risk factors (hypertension, diabetes, hyperlipidemia, smoking, and family history of coronary disease) (Table 2). Interestingly, however, patients with prior AMI had significantly higher OPN levels compared to those without [131.5 (52.2, 219) versus 73.3 (45.1, 125), p = 0.007] (Figure 1(a)). Likewise OPN was significantly higher in patients treated with insulin when compared to those not on insulin [170 (84.6, 196) versus 77.3 (49.5, 149), p = 0.05] (Figure 1(b)). OPN was lower in patients with 1- or 2-vessel disease, 65.3 (43.3, 149), compared to patients with 3-vessel disease or left main, 82.3 (53.3, 158), but this difference did not reach statistical significance (p = 0.23). In addition, OPN showed no relation to left ventricular ejection fraction (r = −0.09, p = 0.31). Conversely, OPN levels were positively related to EuroSCORE (r = 0.2, p = 0.031). Pre-op OPN levels did not differ between patients who had a major adverse event during follow-up compared to those who did not experience any event during this period [116 (54.9, 181.2) versus 77.4 (49.2, 149), p = 0.209]. Furthermore, pre-op OPN levels had no effect on the hazard of future major adverse cardiac events [hazard ratio (95% confidence interval (CI)): 1.48 (0.43–4.99), p = 0.527]. As expected, history of AMI was associated with increased risk of subsequent cardiovascular events at follow-up [hazard ratio (95% CI): 4.46 (1.27–15.66), p = 0.02], even after controlling for OPN levels. 4. Discussion We evaluated pre-op OPN levels in patients with stable CAD undergoing elective CABG on-pump and found that OPN is higher in patients with prior AMI and those on insulin; OPN does not vary in relation to CAD severity or left ventricular ejection fraction but increases with higher EuroSCORES. In addition, OPN does not predict postoperative cardiovascular events but, as expected, a history of AMI does. Plasma OPN levels may be associated with the presence and extent of CAD [7]; in multivariate analysis, however, when atherosclerotic risk factors are taken into account the severity of coronary atherosclerosis is not an independent factor for OPN levels [8]. In addition, OPN has been shown to predict risk for future cardiovascular events in patients undergoing PCI [9]; OPN, however, was higher in patients with coronary calcification and correlated with the number of calcified segments in this study [10] although it is known that cardiovascular events are often associated with the presence of soft atheromatous plaques [11]. We have previously shown that OPN levels do not differ between patients with and without CAD and we now found that OPN levels are similar in relation to the severity of CAD as expressed by the number of the diseased vessels [11]. OPN was inversely related to left ventricular ejection fraction [4] and OPN > 55 ng/mL was an independent predictor of adverse outcome. In this study, however, we did not confirm either the relation to left ventricular performance or the prognostic value. Circulating OPN seems to be positively related to left ventricular dimensions and inversely to ejection fraction mostly in postmyocardial infarction patients [12]; in our study only one-third of patients had a history of AMI and these patients had higher OPN levels and worse outcome. In addition, our short follow-up and complete surgical revascularization may have contributed to low event rate and lack of prognostic significance of OPN. High glucose increases OPN m-RNA in smooth muscle cells [13]; there is increased OPN in diabetic arteries [14]. We did not observe different OPN levels in relation to diabetes, when we compared diabetics, irrespective of treatment, with nondiabetics but OPN was higher in patients on insulin versus those without insulin. There is evidence that insulin-dependent diabetic patients with CAD present with increased levels of cytokines independently of glycemic control, the duration of diabetes, and the extent of CAD [15]. Our results suggest a positive relationship between plasma OPN levels and the surgical risk in stable patients undergoing elective CABG. EuroSCORE point system includes preoperative risk factors, such as age, peripheral arterial disease, left ventricular ejection fraction, and thoracic aorta surgery [16], parameters which in some studies are related to OPN levels [4]. In our cohort, however, OPN did not differ in relation to age and left ventricular performance. 5. Conclusion OPN is associated with preoperative risk assessment prior to low-risk patients undergoing CABG but did not independently predict outcome in the present study. Further research is required with larger number of patients to investigate if pre-op OPN could eventually replace or ameliorate the performance of EuroSCORE for the evaluation of cardiac operative mortality risk. Competing Interests Eftihia Sbarouni, Panagiota Georgiadou, Sofia Chatzikyriakou, Antonis Analitis, Antigoni Chaidaroglou, Demitris Degiannis, and Vassilis Voudris declare that there is no conflict of interests. Figure 1 (a) Significant differences in osteopontin levels between patients with and without history of acute myocardial infarction. (b) Significant differences in osteopontin levels between patients treated with insulin compared to those not on insulin (box plots showing median, interquartile range). Table 1 Baseline characteristics and data of coronary artery bypass grafting procedure in study population. Age (yrs), mean ± SDa 66.5 ± 10 Height (cm), mean ± SD 171 ± 8 Weight (kg), mean ± SD 84 ± 17 Body mass index (kg/m2), mean ± SD 29 ± 5 Osteopontin (ng/mL) median (interquartile range 25th–75th percentiles) 77.9 (49.5, 150.9) Ejection fraction (%), mean ± SD 51 ± 9 Euroscore (%), mean ± SD 3.5 ± 3.2 Number of grafts, n (%)    1 7 (5.4)  2 53 (40.8)  3 64 (49.2)  4 6 (4.6) ACCb (sec), mean ± SD 71.2 ± 34.2 CPBc (sec), mean ± SD 100.5 ± 35.3 aSD: standard deviation. bACC: aortic cross-clumping. cCPB: cardiopulmonary bypass. Table 2 Plasma osteopontin levels in study population. Characteristic Patients, n (%) Osteopontin, ng/mL Median (interquartile range) p value Gender        Female 14 (10.7) 110.8 (59.4, 200.0) 0.103  Male 117 (89.3) 76.9 (48.2, 149.0) Diabetes        Yes 50 (38.5) 83.8 (52.0, 160.0) 0.590  No   77.1 (49.4, 149.0) Hyperlipidemia        Yes 113 (86.9) 80.1 (50.5, 150.9) 0.928  No   72.5 (54.9, 149.0) Hypertension        Yes 121 (93.1) 80.1 (50.5, 149.0) 0.452  No   76.9 (54.9, 221.0) Smoking        Yes 70 (53.8) 75.0 (45.5, 134.0) 0.071  No   88.5 (54.9, 167.0) Family history of coronary disease        Yes 62 (48.1) 81.3 (52.0, 160.0) 0.607  No   76.9 (48.9, 134.0) ==== Refs 1 Waller A. H. Sanchez-Ross M. Kaluski E. Klapholz M. Osteopontin in cardiovascular disease: a potential therapeutic target Cardiology in Review 2010 18 3 125 131 10.1097/crd.0b013e3181cfb646 2-s2.0-77951483204 20395697 2 Fitzpatrick L. A. Severson A. Edwards W. D. Ingram R. T. Diffuse calcification in human coronary arteries: association of osteopontin with atherosclerosis The Journal of Clinical Investigation 1994 94 4 1597 1604 10.1172/jci117501 2-s2.0-0028136961 7929835 3 Bjerre M. Pedersen S. H. Møgelvang R. High osteopontin levels predict long-term outcome after STEMI and primary percutaneous coronary intervention European Journal of Preventive Cardiology 2013 20 6 922 929 10.1177/2047487313487083 2-s2.0-84887100788 23613225 4 Georgiadou P. Iliodromitis E. K. Kolokathis F. Osteopontin as a novel prognostic marker in stable ischaemic heart disease: a 3-year follow-up study European Journal of Clinical Investigation 2010 40 4 288 293 10.1111/j.1365-2362.2010.02257.x 2-s2.0-77949702401 20192976 5 Behnes M. Brueckmann M. Lang S. Diagnostic and prognostic value of osteopontin in patients with acute congestive heart failure European Journal of Heart Failure 2013 15 12 1390 1400 10.1093/eurjhf/hft112 2-s2.0-84890085360 23851388 6 Bradshaw P. J. Jamrozik K. Le M. Gilfillan I. Thompson P. L. Mortality and recurrent cardiac events after coronary artery bypass graft: long term outcomes in a population study Heart 2002 88 5 488 494 10.1136/heart.88.5.488 2-s2.0-0036839621 12381640 7 Ohmori R. Momiyama Y. Taniguchi H. Plasma osteopontin levels are associated with the presence and extent of coronary artery disease Atherosclerosis 2003 170 2 333 337 10.1016/s0021-9150(03)00298-3 2-s2.0-0242661371 14612215 8 Momiyama Y. Ohmori R. Fayad Z. A. Associations between plasma osteopontin levels and the severities of coronary and aortic atherosclerosis Atherosclerosis 2010 210 2 668 670 10.1016/j.atherosclerosis.2009.12.024 2-s2.0-77953286225 20074733 9 Kato R. Momiyama Y. Ohmori R. Prognostic significance of plasma osteopontin levels in patients undergoing percutaneous coronary intervention Circulation Journal 2009 73 1 152 157 10.1253/circj.CJ-08-0687 2-s2.0-58549091481 19023151 10 Abdel-Wahab M. Khattab A. Liska B. Relationship between cardiovascular risk as predicted by established risk scores and coronary artery plaque composition as detected by virtual histology intravascular ultrasound analysis: the PREDICT pilot study EuroIntervention 2008 3 4 482 489 10.4244/EIJV3I4A86 19736092 11 Georgiadou P. Iliodromitis E. K. Kolokathis F. Plasma levels of osteopontin before and 24 h after percutaneous coronary intervention Expert Opinion on Therapeutic Targets 2008 12 12 1477 1480 10.1517/14728220802510740 2-s2.0-57549085148 19007317 12 Ärnlöv J. Evans J. C. Benjamin E. J. Clinical and echocardiographic correlates of plasma osteopontin in the community: the Framingham Heart Study Heart 2006 92 10 1514 1515 10.1136/hrt.2005.081406 2-s2.0-33750042569 16973806 13 Takemoto M. Yokote K. Yamazaki M. Enhanced expression of osteopontin by high glucose in cultured rat aortic smooth muscle cells Biochemical and Biophysical Research Communications 1999 258 3 722 726 10.1006/bbrc.1999.0701 2-s2.0-0033583818 10329452 14 Takemoto M. Yokote K. Nishimura M. Enhanced expression of osteopontin in human diabetic artery and analysis of its functional role in accelerated atherogenesis Arteriosclerosis, Thrombosis, and Vascular Biology 2000 20 3 624 628 10.1161/01.atv.20.3.624 2-s2.0-0034061786 15 Antoniades C. Tousoulis D. Marinou K. Effects of insulin dependence on inflammatory process, thrombotic mechanisms and endothelial function, in patients with type 2 diabetes mellitus and coronary atherosclerosis Clinical Cardiology 2007 30 6 295 300 10.1002/clc.20101 2-s2.0-34250811431 17551966 16 Geissler H. J. Hölzl P. Marohl S. Risk stratification in heart surgery: comparison of six score systems European Journal of Cardio-Thoracic Surgery 2000 17 4 400 406 10.1016/s1010-7940(00)00385-7 2-s2.0-0343294320 10773562

PMC005xxxxxx/PMC5002286.txt

==== Front Case Rep RheumatolCase Rep RheumatolCRIRHCase Reports in Rheumatology2090-68892090-6897Hindawi Publishing Corporation 10.1155/2016/4039801Case ReportCoexistence of Ankylosing Spondylitis and Neurofibromatosis Type 1 Gundogdu Baris 1 Yolbas Servet 1 Yildirim Ahmet 1 Gonen Murat 2 http://orcid.org/0000-0003-4995-430XKoca Suleyman Serdar 1 * 1Department of Rheumatology, Faculty of Medicine, Firat University, 23200 Elazig, Turkey2Department of Neurology, Faculty of Medicine, Firat University, 23200 Elazig, Turkey*Suleyman Serdar Koca: kocassk@yahoo.comAcademic Editor: Jamal Mikdashi 2016 14 8 2016 2016 403980123 2 2016 12 7 2016 Copyright © 2016 Baris Gundogdu et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Ankylosing spondylitis (AS) is a systemic disease primarily characterized by the inflammation of sacroiliac joints and axial skeleton. Neurofibromatosis type 1 (NF1) is a multisystem genetic disease which is characterized by cutaneous findings, most importantly café-au-lait spots and axillary freckling, by skeletal dysplasia, and by the growth of both benign and malignant nervous system neoplasms, most notably benign neurofibromas. In this case report, we present a 43-year-old male with AS and NF1. ==== Body 1. Introduction Ankylosing spondylitis (AS) is a chronic inflammatory disease mainly involving the sacroiliac joints and the axial skeleton. Other clinical findings include enthesitis, peripheral arthritis, and extra-articular organ involvements. In this regard, neurologic complications may occur in patients with AS secondary to fractures of a fused spine. Patients have risk for atlantoaxial subluxation that may lead to cervical myelopathy. Cauda equina syndrome (CES) may also be seen in severe long-standing AS [1, 2]. There are three major clinically and genetically different forms of neurofibromatosis, neurofibromatosis types 1 and 2 (NF1 and NF2), and schwannomatosis. NF1, also known as von Recklinghausen disease, is the most common type. It is a multisystem genetic disease frequently associated with neurologic, cutaneous, and orthopedic manifestations. The typical clinical manifestations of NF1 are neurofibromas, café-au-lait macules, axillary and/or inguinal freckling, and iris hamartomas [3]. 2. Case Report A 43 year-old male was admitted to the rheumatology clinic with the complaints of long-standing neck and low back pain and left lower extremity paresthesia. Neck pain was neuropathic character while low back pain was fulfilling the criteria for inflammatory back pain. Inflammatory back pain is typically characterized by following features: age of onset <40 years, insidious onset, improvement with exercise (no improvement with rest), and pain at the second half of night [4]. His past medical history included NF1 (Figure 1). Sensorimotor polyneuropathy (SM-PNP) accompanied by mixed type axonal degeneration and demyelination in the lower extremities had been detected in electromyographic examination (EMG). Then, metabolic and paraneoplastic causes of PNP had been ruled out. SM-PNP was partially controlled by gabapentin treatment. Because of the prior diagnosis of NF1, brain magnetic resonance imaging (MRI) had been taken. Neurofibromas on the scalp (Figure 2(a)) and 30 × 14 mm in size arachnoid cyst in the right anterior temporal lobe (Figure 2(b)) had been detected on MRI. In addition, extruded disc herniation and myelopathy due to cord compression at C4-C5 level, syringohydromyelia cavity at T12-L1 level, and biforaminal disc protrusion at L3-L4 level had been identified by MRI. Partial relief in painful SM-PNP was provided through 6-day methylprednisolone treatment at dose of 80 mg/day with gabapentin 600 mg twice a day. On physical examination, lumbar spinal motion in sagittal and frontal planes was limited. Sacroiliac joint provocation tests were positive. Neurologically, the patient exhibited mild unsteadiness during the performance of tandem gait and Rhomberg's test. Upper and lower extremity motor strengths and deep tendon reflexes appeared to be decreased slightly. On admission, hemogram, urinalysis, erythrocyte sedimentation rate, and C-reactive protein tests were within normal limits. Initial investigations also revealed normal renal and liver function. However, human leukocyte antigen B27 test was positive. Radiograph of sacroiliac joints depicted bilateral grade III-IV sacroiliitis (Figure 3). Therefore, the patient was diagnosed as AS. Meloxicam at the dose of 15 mg/day was added to patient's current treatment and the patient's symptoms abated. The patient who had stable vital signs was discharged after 3 days. 3. Discussion Neurological complications in AS occur rarely and are quite variable extending from minimal joint instability of the spine to more severe and prominent clinical manifestations such as CES and cervical myelopathy [5, 6]. In the patient presented here, AS coexisted with NF1 which is characterized by the clinical triad of cutaneous findings, skeletal dysplastic deformities, and learning disabilities. NF1 has been reported in association with many infectious and chronic systemic diseases except AS [7–10]. If we take into account the epidemiological and etiopathogenetic features of both diseases, NF1 is an autosomal-dominant disease that occurs in 1 out of 3,000 births. It is induced by the inactivation of NF1 gene. NF1 gene is a tumor suppressor gene that encodes for the neurofibromin protein, a member of the Ras family. This inactivation may be a familial condition with an autosomal-dominant inheritance pattern; otherwise, it may be sporadic [10]. On the other hand, AS particularly occurs in young adults, with a peak age of onset between 20 and 30 years. Although the etiology of AS is not completely understood, it focuses on some environmental factors with a strong genetic predisposition that trigger the disease. A direct relationship between AS and the HLA-B27 gene has been determined [1]. Another possible mechanism of AS etiology is the presentation of an arthritogenic peptide from enteric bacteria via specific HLA molecules [11]. Another important point in this context is the presence of dural ectasia (DE). DE is expansion of the dural sac surrounding the spinal cord and it should also be considered as a common clinical manifestation of both diseases [6, 12] albeit not detected in our case by the present methods. DE most frequently occurs in the lumbosacral spine. It may lead to lower back pain, or neurologic deficits manifested with bowel, or bladder dysfunction. Thus, this finding is important in the differential diagnosis of both diseases. As a result, even though the etiopathogenesis and clinical findings of AS and NF1 differ greatly, these two disorders can coexist incidentally. Upon reviewing the literature, we have noticed that this is the first case report in terms of the association of both diseases. Competing Interests There are no competing interests regarding the publication of this paper. Figure 1 Abundant neurofibromas on dorsal side. Figure 2 Cranial MRI findings. (a) Axial T1-weighted image depicts multiple neurofibromas on the scalp (marked by arrows). (b) Axial T2-weighted image demonstrates an arachnoid cyst in the right anterior temporal lobe (marked by an asterisk). Figure 3 Pelvic X-ray showing bilateral grade III-IV sacroiliitis. ==== Refs 1 Khan M. A. Update on spondyloarthropathies Annals of Internal Medicine 2002 136 12 896 907 10.7326/0003-4819-136-12-200206180-00011 2-s2.0-0037129952 12069564 2 Hunter T. The spinal complications of ankylosing spondylitis Seminars in Arthritis and Rheumatism 1989 19 3 172 182 10.1016/0049-0172(89)90030-9 2-s2.0-0024814711 2513645 3 DeBella K. Szudek J. Friedman J. M. Use of the National Institutes of Health criteria for diagnosis of neurofibromatosis 1 in children Pediatrics 2000 105 3 608 614 10.1542/peds.105.3.608 2-s2.0-0034094731 10699117 4 Sieper J. van der Heijde D. Landewé R. New criteria for inflammatory back pain in patients with chronic back pain: a real patient exercise by experts from the Assessment of SpondyloArthritis international Society (ASAS) Annals of the Rheumatic Diseases 2009 68 6 784 788 10.1136/ard.2008.101501 2-s2.0-67349230644 19147614 5 Tyrrell P. N. M. Davies A. M. Evans N. Neurological disturbances in ankylosing spondylitis Annals of the Rheumatic Diseases 1994 53 11 714 717 10.1136/ard.53.11.714 2-s2.0-0027973340 7826132 6 Liu C.-C. Lin Y.-C. Lo C.-P. Chang T.-P. Cauda equina syndrome and dural ectasia: rare manifestations in chronic ankylosing spondylitis British Journal of Radiology 2011 84 1002 e123 e125 10.1259/bjr/45816561 2-s2.0-79956360996 21606066 7 Corominas H. Guardiola J. M. Matas L. Vázquez G. Neurofibromatosis and systemic lupus erythematosus. A matter of coincidence? Clinical Rheumatology 2003 22 6 496 497 10.1007/s10067-003-0764-8 2-s2.0-2342488222 14677041 8 Alrumaih H. Ilyas I. Kashif S. Neurofibromatosis induced hip arthritis. An unusual presentation American Journal of Case Reports 2014 15 79 81 10.12659/ajcr.889726 2-s2.0-84896735622 24587854 9 Till S. H. Amos R. S. Neurofibromatosis masquerading as monoarticular juvenile arthritis British Journal of Rheumatology 1997 36 2 286 288 10.1093/rheumatology/36.2.286 2-s2.0-0030860703 9133949 10 Takazawa Y. Sakurai S. Sakuma Y. Gastrointestinal stromal tumors of neurofibromatosis type I (von Recklinghausen's disease) The American Journal of Surgical Pathology 2005 29 6 755 763 10.1097/01.pas.0000163359.32734.f9 2-s2.0-19544389327 15897742 11 Sorrentino R. Böckmann R. A. Fiorillo M. T. HLA-B27 and antigen presentation: at the crossroads between immune defense and autoimmunity Molecular Immunology 2014 57 1 22 27 10.1016/j.molimm.2013.06.017 2-s2.0-84884412625 23916069 12 Woon C. Y.-L. Clinical images-dural ectasia: a manifestation of type 1 neurofibromatosis CMAJ 2010 182 13 p. 1448 10.1503/cmaj.092004 2-s2.0-77956898075

PMC005xxxxxx/PMC5002287.txt

==== Front Comput Intell NeurosciComput Intell NeurosciCINComputational Intelligence and Neuroscience1687-52651687-5273Hindawi Publishing Corporation 10.1155/2016/5089767EditorialTheory and Applications of Bioinspired Neural Intelligence for Robotics and Control http://orcid.org/0000-0002-6888-7993Yang Simon X. 1 * Luo Chaomin 2 Li Howard 3 http://orcid.org/0000-0002-7130-8331Ni Jianjun 4 Zhang Jianwei 5 1School of Engineering, University of Guelph, Guelph, ON, Canada N1G 2W12University of Detroit Mercy, Detroit, MI 48221, USA3University of New Brunswick, Fredericton, NB, Canada E3B 5A34College of IOT Engineering, Hohai University, Changzhou 213022, China5University of Hamburg, 22527 Hamburg, Germany*Simon X. Yang: syang@uoguelph.ca2016 14 8 2016 2016 508976724 5 2016 25 5 2016 Copyright © 2016 Simon X. Yang et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. ==== Body Computational intelligence approaches are nature-inspired methods, which offer a wealth of ideas for solutions to complex problems. In comparison to the traditional approaches, the computational intelligence approaches are more powerful so that they do not need the reformulation of the problem to search a nonlinear and a nondifferentiable space with real world conditions with the massive parallelism. Another advantage of the computational intelligence approaches is the flexibility of the fitness function formulation, which can be expressed as a proper function of the system outputs and are suitable for multiobjective problems. Recently, a new type of computational intelligence methods has been developed to overcome the limitations of traditional artificial intelligent methods. One of the most important features of these computational intelligence methods is that their working mechanisms are more lifelike to an individual or a group of organisms, which can be understood very well. These methods usually have higher efficiency than the traditional artificial intelligent methods. The computational intelligence methods of this type are defined as bioinspired intelligent algorithms to distinguish from the traditional artificial intelligent methods. Research on computational intelligence and neuroscience, particularly bioinspired intelligence, has made significant progress in both understanding the neuroscience and biological systems and applying to various robotic and control systems. The aim of this special issue is to provide a forum for different research efforts towards new activities in intelligent robotics and control systems with emphasis on humanoid robots and bioinspired robotics, where the methodologies for the robotic systems are mainly inspired from the strategies, mechanisms, and functionality of neural systems and biological systems, for example, biologically inspired neural networks, genetic algorithms, and fuzzy systems. This special issue includes six papers selected on a peer review basis. These papers present research results in tracking control of underactuated ship, coordinated path following of multiple marine vessels, driving behaviour modeling of electrical wheelchair users, gait planning and stability control of a quadruped robot, and optimization control of a production process. In addition, there is a survey paper in bioinspired intelligent algorithm and its applications for mobile robot control. The paper, which is by J. Yuan et al., is about the course control of underactuated surface vessel. In this paper, a backstepping controller with robust neural network is designed to deal with the uncertain and underactuated for the ship. The uniform stability and the convergence of tracking error to zero are quarantined by the Lyapunov stability theory. The paper by J. Ni et al. presents a survey of the state-of-the-art research in bioinspired intelligent algorithms (BIAs), which focuses on the realization of various BIAs based on different working mechanisms and the applications for mobile robot control, to help in understanding BIAs comprehensively and clearly. This survey paper includes four primary parts: a classification of BIAs from the biomimetic mechanism, a summary of several typical BIAs from different levels, an overview of current applications of BIAs in mobile robot control, and a description of some possible future directions for research. The paper presented by M. Fu and Y. Xu focuses on the coordinated path following of multiple marine vessels with speed saturation. The objective is to avoid the tracking error jump and solve the speed saturation problem in straight path. The virtual leader strategy is applied, and the neural dynamic model and passivity-based techniques are used together to yield a distributed control strategy. The paper, which is authored by J. Li et al., is about the gait planning and stability control of the quadruped robot. This paper presents a new CPG (central pattern generator) model controller and its gait switching strategy based on the Wilson-Cowan model, to generate smooth gait and shorten the adjusting time of the model oscillation system. An adaptive speed adjustment and gait switch are completed by the real-time computing of ZMP (zero moment point), to realize stability control. The paper by S. O. Onyango et al. discusses the driving behaviour modeling problem of electrical wheelchair users. This paper presents the method to obtain the steering data for parameter identification. The modeling method based on the improved Directed Potential Field (DPF) for trajectory planning is applied. The paper authored by D. He et al. is on the optimization control of the color-coating production process (CCPP). An effective optimization control strategy for the CCPP is proposed. To manage the model uncertainty, the robust optimization approach is introduced to improve the feasibility of the optimized solution and the iterative learning control is then utilized to further refine the model uncertainty. In order to deal with the upcoming challenges in robotics and control, it is necessary to develop more advanced bioinspired neural intelligence techniques and theories in the literature. This special issue not only captures a snapshot of the current research but also gives some future aspects. We hope it will motivate the readers to direct their effort into this rewarding area. Acknowledgments As the guest editors of the special issue, we would like to thank all the authors and reviewers for their contributions to our special issue. Simon X. Yang Chaomin Luo Howard Li Jianjun Ni Jianwei Zhang

PMC005xxxxxx/PMC5002288.txt

==== Front Biomed Res IntBiomed Res IntBMRIBioMed Research International2314-61332314-6141Hindawi Publishing Corporation 10.1155/2016/2694030Research ArticleAdvancing Implementation of Evidence-Based Public Health in China: An Assessment of the Current Situation and Suggestions for Developing Regions Shi Jianwei 1 2 http://orcid.org/0000-0002-8937-090XJiang Chenghua 2 Tan Duxun 3 Yu Dehua 1 Lu Yuan 1 Sun Pengfei 2 Pan Ying 1 Zhang Hanzhi 1 http://orcid.org/0000-0002-5022-4497Wang Zhaoxin 1 2 * Yang Beilei 4 1Yangpu Hospital, Tongji University School of Medicine, Shanghai 200090, China2Tongji University School of Medicine, Shanghai 200092, China3The Fifth Affiliated Hospital of Southern Medical University, Guangzhou 510900, China4Tongji University College of Economics and Management, Shanghai 200092, China*Zhaoxin Wang: supercell002@sina.comAcademic Editor: Sabine Rohrmann 2016 14 8 2016 2016 269403015 4 2016 6 7 2016 17 7 2016 Copyright © 2016 Jianwei Shi et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Objective. Existing research shows a serious scarcity of EBPH practice in China and other developing regions; as an exploratory study, this study aimed to assess the current EBPH implementation status in Shanghai of China qualitatively. Methods. Using semistructured key informant interviews, we examined the status of and impediments to the lagging EBPH in China. Data were analyzed based on the Consolidated Framework for Implementation Research (CFIR). Results. Chinese public health practitioners knew more about evidence-based medicine but less about EBPH. The situation was worse in community healthcare centers. Participants perceived that evidence sources were limited and the quality of evidence was low. Concerning the inner setting factors, the structural characteristics, networks and communications, implementation climate, and leadership engagement were confronted with many problems. Among the outer setting factors, external government policies and incentives and low patient compliance were the key problems. Additionally, public health practitioners in Shanghai lacked sufficient awareness of EBPH. Furthermore, the current project-based EBPH lacks a systematic implementation system. Conclusions. Existing practical perspectives on EBPH indicate a lag in the advocacy of this new ideology in China. It would be advisable for healthcare institutions to take the initiative to explore feasible and multiple methods of EBPH promotion. the Project of Central University Special Funds for Scientific Research1500219099Excellent Young Teacher Project (EYTP) of Fundamental Research Funds for the Central UniversitiesChina Postdoctoral Science Foundation2016M591719 ==== Body 1. Introduction 1.1. Evidence-Based Public Health as a Booster to Public Health Promotion Evidence-based public health (EBPH), which was formally proposed in the US approximately two decades ago, has gained widespread academic attention and obtained fast progress in high-income countries, especially since the 2000s [1]. Although there is not yet a generic definition, there appears to be a consensus among scholars and practitioners that EBPH is a combination of scientific evidence and values, resources, and context that should be used to inform public health decisions and improve the health of populations [1–4]. Over the past several years, many efforts have been devoted to developing evidence-based decision-making in public health fields in developed countries. The Cochrane Collaboration in many developed countries and the national guidelines of the “Community Guide” and “Cancer Control P.L.A.N.E.T.” in the US provide good approaches to planning and developing evidence-based programs and policies [5–7]. Under these circumstances, the use of evidence-based practices in public health has exerted numerous direct and indirect benefits, including access to more and higher-quality information on what works, a higher likelihood of successful programs and policies being implemented, greater workforce productivity, and more efficient use of public and private resources [2, 8, 9]. However, existing knowledge shows that EBPH has not achieved systematic dissemination. Studies have revealed that implementation or translation of evidence-based approaches in public health always encounters complex barriers, including personal and institutional factors and political environment and deficits in information systems, resources, and leadership [10, 11]. Additionally, lack of time and incentives, inadequate connections between research and practice, and the absence of cultural and managerial support are among the most commonly cited barriers [12–14]. 1.2. Evidence-Based Public Health in the Context of Public Health Reform in China Despite the growing body of research on EBPH implementation and promotion in high-income countries, there is a lack of studies that assess status of EBPH and its dissemination in China and other developing regions [15, 16]. To date, EBPH is still in its early stages in China. In 2003, the outbreak of SARS initiated China's public health reform, and precisely at that time the importance of EBPH was realized by scholars. By applying the principles and methods of evidence-based medicine, the Chinese Cochrane Center specifically developed evidence-based strategies for preventing and addressing unexpected public health events. After the “Wenchuan earthquake” in Sichuan Province in 2008, more efforts concerning evidence-based strategies were made, involving emergency management, modes of rescue, physical and mental interventions, rehabilitation, finance, and logistics [17]. After 2008, the scope of EBPH research in China gradually expanded to wide public health domains, such as epidemiology, health economics, and health management, all of which aim at making decisions on the basis of the best available scientific evidence [18]. The existing research in China has mostly focused on developing or improving evidence-based measures or approaches. However, in regard to how to translate evidence-based science into practice, studies of current EBPH utilization and its impeding factors are scarce. Consequently, the main objective of the study is to review and assess the current EBPH implementation status and explore its barriers in China. As an exploratory study, we chose Shanghai as the sample region because EBPH is new to China and we seek to investigate the situation in this most developed region to reveal the overall situation in China. The findings may also provide effective suggestions for low- and middle-income countries that are considering the introduction and practice of EBPH in the public sector to promote their EBPH implementation. 2. Methods 2.1. Analytical Framework To analyze the status and major determinants of EBPH practices in China, we formulated an interview instrument based on Damschroder et al.'s Consolidated Framework for Implementation Research (CFIR) [19], which has been widely used by prior studies addressing health innovation implementation [20–22]. The CFIR provides a systematic overarching typology to promote the development of implementation theory and verification of the target and scope of EBPH. It comprises the following five major domains: intervention characteristics, inner setting, outer setting, individuals, and implementation process (Table 1). The subcategories or constructs are also displayed in Table 1. In summary, the framework indicates that the obstacles to EBPH practices can essentially be reflected in different domains and constructs. By using the CFIR framework, we developed interview instruments for data collection in healthcare institutions. The interview instrument is available on request, and the brief sample questions are listed in Table 1. 2.2. Data Collection In our study, the participants consisted of 14 officials from the Chinese Centers for Disease Control and Prevention (CDC) and 82 practitioners from 10 community healthcare centers, 8 secondary hospitals, and 6 tertiary hospitals during the period of June 2015 to October 2015. We included secondary and tertiary hospitals because they also undertook much of the public health work in China. Ultimately, as an exploratory study, this study can reflect the most advanced level of EBPH implementation in China. For all the institutions we visited during our field study, we do not use their names in our analysis below due to privacy concerns. Instead, we refer to them as Institution A1, A2,…, A10 (A represents community healthcare centers), B1, B2,…, B8 (B represents secondary hospitals), and C1, C2,…, C6 (C represents tertiary hospitals). The CDC is assigned the label of D1. 2.3. Data Analysis In this study, data were analyzed for thematic content guided by the CFIR framework, and the indicative coding was obtained by iterative methods. We then identified key issues and problems in the implementation of EBPH in Shanghai, China. Additionally, attempting to develop a clear understanding of EBPH practices and their impediments in China, we collected and analyzed the literature from the Chinese National Knowledge Infrastructure Database and Wan Fang Database (the two largest databases in China), and we related policies issued by the government. All these data were collected from January 1, 2000, to March 31, 2016. 3. Results 3.1. Status of EBPH Practices in Shanghai Overall, we found that a growing emphasis has been placed on public health in recent years in China. For instance, prompted by the national health bureau's request that basic public health programs be conducted, the Shanghai health bureau created its key healthcare list and launched a public health campaign beginning in 2011 [23]. These programs referred to 12 categories, including disease control, vaccination, child healthcare, maternal healthcare, gerontology, and health education. These activities generally involved EBPH practices. Unfortunately, however, the condition of practice in Shanghai remains unsatisfactory. At the public health practitioners' level, on the whole, it was found that health practitioners knew and practiced more evidence-based medicine but clearly less evidence-based public health. “The evidence we use most often is that about the application of medication… For population intervention, personally, I'm not very familiar with that or how to do it” (A3-3). Furthermore, by comparing the status of EBPH practices in community healthcare institutions with that in higher-grade ones, we found that interviewees in higher-grade institutions had more access to EBPH. The discrepancy lay in the higher-grade institutions' access to more academic projects. Due to the lack of funding, EBPH practices were mainly initiated in the form of academic projects. The scope of the implementation depended on the project, such as the population and regions included and the budget dedicated. Moreover, it was noteworthy that although many public health practitioners had implemented evidence-based practices in their work because they had not received systematic and orthodox training on EBPH, they applied little EBPH. At the CDC officials' level, we were disappointed to find that among the 14 interviewees four senior officials were not even familiar with the concept of EBPH. Others knew a bit but had not recognized its importance and significance. In fact, their work was highly related to EBPH because they must take measures based on the upper government's guidance, such as the policy of “Tertiary Prevention” targeting chronic disease prevention and control. Overall, they seemed to be passive policy receivers but not proactive in proposing or improving new approaches/policies based on accumulated data or experience. 3.2. Impediments to EBPH Implementation 3.2.1. Intervention Characteristics Among the 82 interviewees from healthcare institutions, most answered with external channels, such as published literature, guidelines, policies, seminars, and academic conferences. Moreover, practitioners in community healthcare centers deemed it more difficult to obtain good-quality intervention sources. Indeed, most of the database of published literature was not free of charge, and it was a great burden for fundamental institutions to afford. 3.2.2. Inner Setting Factors The inner setting comprises the spectrum of organization-level strategies that have the potential to influence EBPH implementation. In this study, the interviewees expressed that there were many institutional inhibitors. According to the CFIR framework, the inner setting factors comprise internal structural characteristics, networks and communications, implementation climate, and leadership engagement [20]. In fact, these internal constructs are correlated with each other. (1) Structural Characteristics. Interviewees in this study stated that to better promote EBPH practices, “a team with strong ability in all aspects is required” (C1-1). However, extracted results showed that although inner hierarchical structure was reasonable, implementation was restricted by the capacity of both the leaders and the subordinates, reflecting their lack of capacities to cooperate and operate the given programs by observing the mechanism instead of practicing it spontaneously. (2) Networks and Communications. Another important issue is how we can effectively transmit our ideas or experiences to others. As one practitioner noted, “everyone's part in the organization is written in detail and should be very clear… Also, the bridge between the doctors and the management level is important, and it's quite good in our organization” (A3-2). (3) Implementation Climate. In our analysis, we found that among the diverse climate categories, the lack of organizational incentives was mentioned most frequently. Extracted information indicated that until now there had been no direct incentive from the government in the form of funds to encourage public health practitioners to apply EBPH-related programs, although some were encouraged by the government. Instead, the most common way to practice EBPH was in the form of academic projects. When asked “how much support do you feel the department provides for the processes necessary for utilizing evidence-based programs/policies?” interviewees answered “we have our own research” (A1-1) and “the support is project-based, like the project approved by the Shanghai Municipal Commission of Health and Family Planning, or by our hospital” (C1-4). In this case, the duration of the funding depended on the term of the project. (4) Leadership Engagement. In China, executive leaders in healthcare institutions are appointed administratively by the health bureau. Additionally, an interesting and common phenomenon is that even institutions are endowed with certain autonomy; if not forced by the upper government, executives are unwilling to take measures because their effects may not be known during their tenure [24]. Therefore, new technologies or approaches cannot be utilized and promoted timely. As stated by one physician in C2: “The same is true with those officials who are willing to make their own achievements. They manage to hold to a point and make it his/her achievement, and leave over other things. For example, one official worked on this, but the new leader changed ideas and changed everything. This is common in China.” Currently, because an academic project is the conventional way to implement EBPH, many public health practitioners desire their administrators to provide more such opportunities. As stated by a physician in a community hospital, “I do not know what the chronic disease management programs in other community-based health services centers look like, ours used to be a mess. But after the new dean got here, our programs become sound” (A3-2). 3.2.3. Outer Setting Factors Generally, the outer setting comprises the economic, political, and social context within which an organization resides. Normally, the implementation of a new approach will be influenced by outer setting factors mediated through changes in the inner setting [25]. (1) External Policies and Incentives. Among the broad constructs that encompass external factors to spread interventions, the policies, regulations, guidelines, and public or benchmark reporting count for a great deal [26], in this study, when asked “can you think of any external condition that can help you to implement the evidence-based public health interventions?” impressively, among the 82 practitioners, 53 noted that microlevel policies were the most important and urgent measures. Related problems with the policy can be manifested as lack of funding and execution power. Practitioners noted that “encouraging young and promising doctors to do EBPH work in community-based health centers requires money; more funding should be provided by the government” (A2-2). Concerning the power to execute current policy, some interviewees said that although the policy of “classification and treatment of constipation among different health institutions” was advocated because the referral medical mechanism did not function well, patients with constipation usually went directly to the higher-grade hospital and it was difficult for the community health centers to track them and follow up. Notably, the third-party-neighbourhood committees in China's administrative system are helpful in promoting the public health. Although they are not health institutions, their role of communication and organization between health institutions and the public facilitates public health work, which is a typical phenomenon in China. The neighbourhood committee is the basic autonomous organization of self-management, self-supervision, and self-service, and each community has its own neighbourhood committee. Typically, when new policies are issued by the government, this type of committee is also asked by the government to offer assistance to other targeted institutions. In this case, it can help the public realize and accept health practitioners' EBPH practices and facilitate their work. (2) Patient Needs and Resources. Patients are the direct recipients of health programs; conversely, their feedback or requirements can be beneficial for health practitioners. If practitioners can extract experience and apply it to new practice, doing so is obviously a good EBPH practice. In reality, interviewees in this study reflected that this practice was weak, especially in the community health centers. The key problem lay in low patient compliance. One physician stated, “we always provide the patients with appropriate guidance or treatment according to the guidelines, such as coronary heart disease and hyperlipemia. However, many of the patients' compliance is not good. Because health management requires cooperation between the physicians and patients, a lack of this will further result in low EBPH practice” (B1-1). 3.2.4. Individual Factors Greenhalgh et al. (2004) described the significant role of individuals as carriers of cultural, organizational, professional, and individual mindsets, norms, interests, and affiliations. Individuals make choices and can wield power and influence over others with predictable or unpredictable consequences for implementation [27]. In this study, such individual factors were reflected as lack of time and energy to think and participate, ineffective ability to conduct the program, and lack of awareness of EBPH. By comparison, it was shown that feedback from community healthcare centers was more related to limited skill and lack of consideration, while general hospitals reflected that they failed to practice EBPH due to high workloads and limited time. Another interesting phenomenon among the Chinese public health practitioners was that although some had conducted EBPH programs, they failed to identify EBPH practices. This low consciousness was directly caused by a lack of professional training. 3.2.5. Implementation Process According to the CFIR framework, the implementation process contains five essential activities that range across organizational change models: planning, engaging, executing, reflecting, and evaluating [20]. Through analysis, we found that there was no clear EBPH implementation process for those institutions that already practiced EBPH in Shanghai. One interviewee stated, “we are doing research in our communities or we are working towards this” (A3-1), signifying a lack of systematic implementation process. However, establishing a regular EBPH process is not as easy as it sounds because it requires department management, quality management, evaluation, and improvement. In summary, the CFIR's overarching structure supports the exploration of essential factors that may be encountered during EBPH implementation in Shanghai. A brief summary of our interviews is presented in Table 2. 4. Discussion 4.1. Key Challenges of EBPH Implementation in China Regarding the five CFIR domains, evidence, the institutional system, government support, and individuals' awareness persist as the key challenges to implementing EBPH. 4.1.1. Insufficient and Low-Quality Public Health Evidence in China The primary problem with public health evidence in China lies in its insufficiency and low quality. Comparatively, developed countries have accumulated persuasive national guidelines and policies in these aspects. In the US, the national level of evidence, such as the Guide to Community Preventive Services, has been put into practice [28]. In Australia, “Health-Evidence.org” is a widely accepted evidence source. Canada has established an extensive online set of evidence reviews and resources for scholars or practitioners to disseminate EBPH (http://www.healthevidence.org/) [29, 30]. Additionally, the Cochrane Collaboration provides an abundant repository in many developed countries [31]. The results showed that Chinese public health professionals have a pivotal position in efforts to change the current circumstances. The change requires the collective efforts of both the external government and internal healthcare institutions. For the government, construction of a decision-making database that suits Chinese people should be advanced, and both the CDC and other public health administrations must utilize population data. For institutions, the priority should be providing more opportunities for their employees to obtain evidence, such as internal training and external academic meetings. 4.1.2. Irrational Institutional System without Effective Communication and Incentive Mechanisms Qualitative results indicated that there were few effective communication systems for EBPH practice within institutions. Additionally, there was a lack of sufficient incentives for practitioners to promote EBPH while they are already busy with work. All these factors made it disadvantageous for public health practitioners to acquire and spread evidence-based philosophy and methods. Another vital obstacle was the irrational financial incentive mechanism. The extracted information showed that the initiation of EBPH was based on academic projects, making implementation restricted by the duration and amount of the funding. To establish favourable mechanisms, much should be considered involving intensified training and other communication methods. Furthermore, multiple incentives such as performance reviews, promotions, and less tangible incentives like increased stature or respect should be explored. 4.1.3. Lack of Resources for EBPH Practices due to Limited Government Support Although the government health agencies aimed to achieve universal public health program coverage in China, the coverage rate remained far from the established standards [32]. As revealed, one important reason was that these projects incurred expenses far beyond their budget. As a result, institutions and practitioners claimed they could not sustain the projects. Another reason was the government's lack of EBPH consciousness, with the results showing that some public health officials did not even understand EBPH. Except for more direct invested funds, other government support should be exploited. For instance, government health agencies can provide more public health-related projects to induce practitioners or institutions to focus on public health. To increase health officials' awareness, much support should be provided to improve their innovation and scientific decision-making abilities. 4.1.4. Public Health Practitioners' Weak Awareness of EBPH Practices in China EBPH is a relatively new ideology in China. Although it has been recognized and applied in public health practitioners' daily work to some degree in Shanghai, we found that many practitioners failed to realize its essence and potential advantages due to their conventional awareness and work patterns. Under the current limitations in internal and external incentives, more active approaches should be explored to inspire individuals. Personal presentations and institution-based training (e.g., information retrieval, EBPH methodology popularization) are good methods that can assist individuals to achieve a broad reach. Additionally, because practitioners in higher-grade health institutions in Shanghai are more aware of and practice more EBPH, policies should be made to encourage the workforces in community healthcare centers, who are also the main force of future public health according to China's strategic health policy, to better stimulate EBPH practice. 4.2. International Comparison of EBPH Practices The necessity of using EBPH to improve public health practice is well recognized by practitioners and researchers [2, 4, 6]. However, we found developing countries seriously delayed in this region. Numerous studies have shown that in developed countries, such as the US and many European countries, scholars are focusing on disseminating science to exploit effective interventions or strategies to elevate the quality and efficiency of evidence-based public health programs [2, 16, 33]. For instance, Allen et al.'s (2013) research designed multiphase dissemination approaches that included multiday in-person training workshops, electronic information exchange modalities, and remote technical assistance. Yost et al. noted that in Canada the National Collaborating Center for Methods and Tools (NCCMT) has established an online, searchable, and freely accessible collection of methods and tools to support knowledge translation in public health [34]. However, in China and most developing regions, existing research on EBPH is still in its early stages and exploratory studies of current status and its influencing factors are scarce, as is implementation science [35, 36]. Scattered studies mainly address the acquisition of health-related evidence. By exploring the current status of and impediments to EBPH in China, this study can provide inspiration for EBPH implementation and public health promotion in other developing countries and regions that suffer from the same limitations. 5. Limitations Several limitations of the current review should be noted. First, the data were self-reported, and we had no ability to objectively compare the reported activities with actual practices. Second, a convenience sample of key informants limits the generalizability of our study but was an appropriate approach in this exploratory study. Additionally, we could not avoid social desirability because a potential source of bias existed. Third, the international literature may not be able to convey the panorama of EBPH studies in other developing regions; however, restricted by language, we cannot obtain the local literature. Fourth, this study only depicted the present situation in Shanghai; however, Shanghai's experiences with EBPH are likely not generalizable to other areas in China because Shanghai is unique in its high education and income and access to international resources. However, because EBPH is new to China, we presume that healthcare providers in Shanghai have more opportunities to obtain evidence-based public health knowledge and practice it for us to acquire the up-to-date situation of China. We realized that much still remains to be explored in other regions of China. 6. Conclusions In summary, by applying the CFIR theory, which accurately describes the overarching contexts of the obstructive constructs qualitatively, we determined that China lags in the cognition and utilization of EBPH and that the operation of EBPH is primarily project-based. Considering the key challenges of EBPH promotion in China, it seems more feasible for healthcare institutions to take the initiative to explore multiple methods to facilitate and improve the development of EBPH. This study will become the baseline information on EBPH evaluation in China, and we hope it will be beneficial for other developing regions in launching and deepening their research and practice. Acknowledgments This study was supported by the Project of Central University Special Funds for Scientific Research (Grant no. 1500219099), the Excellent Young Teacher Project (EYTP) of Fundamental Research Funds for the Central Universities, and the China Postdoctoral Science Foundation Funded Project (Grant no. 2016M591719). The authors sincerely acknowledge the technical assistance and help from the Chinese Center for Disease Control and Prevention and the healthcare institutions in Shanghai Province. They appreciate all the interviewees' contributions to this study. In addition, they would like to express their appreciation to Ross C. Brownson from Washington University in St. Louis, who provided them with enlightenment to deepen the study of evidence-based public health in China. Abbreviations EBPH:Evidence-based public health CDC:Centers for Disease Control and Prevention CFIR:Consolidated Framework for Implementation Research. Ethical Approval All research activities were conducted with integrity and in line with generally accepted ethical principles and approved by the Ethics Committees of Tongji University (Ref. LL-2016-ZRKX-017). None of the personal information from the interviewees involved in the survey was available to people outside of the research team. Consent This study received the consent form from all the interviewees. Competing Interests The authors have declared that no competing interests exist. Table 1 The Consolidated Framework for Implementation Research. Domain Constructs Questions and examples Intervention characteristics Intervention source, evidence strength and quality, relative advantage, adaptability, trialability, complexity, design quality and packaging, cost How would you describe the current evidence-based interventions? Inner setting Structural characteristics, networks and communications, culture, implementation climate, leadership engagement Can you identify any organizational barriers that impede your ability to implement EBPH? Outer setting External policies and incentives, patient needs and resources, cosmopolitanism, peer pressure Can you identify any external factors that impede your ability to implement EBPH? Individual characteristics Knowledge and beliefs about the intervention, self-efficacy, individual stage of change, individual identification with organization, other personal attributes Can you identify any personal barriers that impede your ability to implement EBPH? Implementation process Planning, engaging, executing, reflecting and evaluating How would you describe the situation of your department as it relates to implementing evidence-based processes? Note: the framework was cited from [19]. Table 2 CFIR perspectives on EBPH implementation issues. CFIR dimension Key issues and problems Intervention characteristics Intervention sources were limited, and most were from external channels; the quality of the evidence was not ideal. Inner setting The implementation was restricted by the capacity of both the leaders and the subordinates; there was a lack of organizational incentives and rewards for EBPH implementation. Outer setting Although external policies proposed that practitioners implement evidence-based practices, the execution of the policies was not good, and the government funding was insufficient; there was low patient compliance with the practitioners' EBPH practices. Individual characteristics There was a lack of time and energy to practice EBPH and low awareness and identification of EBPH. Implementation process There was no clear process for EBPH implementation. ==== Refs 1 Kohatsu N. D. Robinson J. G. Torner J. C. Evidence-based public health: an evolving concept American Journal of Preventive Medicine 2004 27 5 417 421 10.1016/j.amepre.2004.07.019 2-s2.0-8844233487 15556743 2 Brownson R. C. Baker E. A. Leet T. L. Gillespie K. N. True W. R. Evidence-Based Public Health 2011 2nd New York, NY, USA Oxford University Press 3 Muir Gray J. A. Evidence-Based Healthcare: How to Make Health Policy and Management Decisions 1997 New York, NY, USA Churchill Livingstone 4 Rychetnik L. Hawe P. Waters E. Barratt A. Frommer M. A glossary for evidence based public health Journal of Epidemiology and Community Health 2004 58 7 538 545 10.1136/jech.2003.011585 2-s2.0-3042535701 15194712 5 Zaza S. Briss P. A. Harris K. W. The Guide to Community Preventive Services: What Works to Promote Health? 2005 New York, NY, USA Oxford University Press 6 Briss P. A. Brownson R. C. Fielding J. E. Zaza S. Developing and using the guide to community preventive services: lessons learned about evidence-based public health Annual Review of Public Health 2004 25 281 302 10.1146/annurev.publhealth.25.050503.153933 2-s2.0-2142705148 7 Kerner J. F. Guirguis-Blake J. Hennessy K. D. Translating research into improved outcomes in comprehensive cancer control Cancer Causes & Control 2005 16 Supplement 1 27 40 10.1007/s10552-005-0488-y 2-s2.0-26044450541 16208572 8 Hausman A. J. Implications of evidence-based practice for community health American Journal of Community Psychology 2002 30 3 453 467 10.1023/A:1015345321658 2-s2.0-0036614957 12054038 9 Tang K. C. Ehsani J. P. McQueen D. V. Evidence based health promotion: recollections, reflections, and reconsiderations Journal of Epidemiology and Community Health 2003 57 11 841 843 10.1136/jech.57.11.841 2-s2.0-0242487810 14600105 10 Jacobs J. A. Dodson E. A. Baker E. A. Deshpande A. D. Brownson R. C. Barriers to evidence-based decision making in public health: a national survey of chronic disease practitioners Public Health Reports 2010 125 5 736 742 2-s2.0-77957962479 20873290 11 Stamatakis K. A. McQueen A. Filler C. Measurement properties of a novel survey to assess stages of organizational readiness for evidence-based interventions in community chronic disease prevention settings Implementation Science 2012 7 1, article 65 10.1186/1748-5908-7-65 2-s2.0-84865112261 12 Armstrong R. Waters E. Roberts H. Oliver S. Popay J. The role and theoretical evolution of knowledge translation and exchange in public health Journal of Public Health 2006 28 4 384 389 10.1093/jpubhealth/fdl072 2-s2.0-33845256686 17082462 13 Baker E. A. Brownson R. C. Dreisinger M. McIntosh L. D. Karamehic-Muratovic A. Examining the role of training in evidence-based public health: a qualitative study Health Promotion Practice 2009 10 3 342 348 10.1177/1524839909336649 2-s2.0-74049161312 19574586 14 Maylahn C. Bohn C. Hammer M. Waltz E. C. Strengthening epidemiologic competencies among local health professionals in New York: teaching evidence-based public health Public Health Reports 2008 123 supplement 1 35 43 2-s2.0-44849144345 18497017 15 Cuijpers P. de Graaf I. Bohlmeijer E. Adapting and disseminating effective public health interventions in another country: towards a systematic approach European Journal of Public Health 2005 15 2 166 169 10.1093/eurpub/cki124 2-s2.0-20644448820 15755779 16 Brownson R. Dreisinger M. Colditz G. Proctor E. K. Brownson R. Colditz G. Proctor E. The path forward in dissemination and implementation research Dissemination and Implementation Research in Health: Translating Science to Practice 2012 New York, NY, USA Oxford University Press 498 508 17 Li Y. P. Yang X. Y. Chen Y. L. Evidence-based decision-making of public health in China—challenge and exploration Chinese Journal of Evidence-Based Medicine 2008 8 11 945 950 18 Chai G. L. Wang Z. X. The status quo and issues study on EBPH used in chronic disease control in China Proceedings of the Zhejiang General Medical Academic Annual Meeting 2012 187 189 19 Damschroder L. J. Aron D. C. Keith R. E. Kirsh S. R. Alexander J. A. Lowery J. C. Fostering implementation of health services research findings into practice: a consolidated framework for advancing implementation science Implementation Science 2009 4 1, article 50 1035 1039 10.1186/1748-5908-4-50 2-s2.0-69749093725 20 Damschroder L. J. Lowery J. C. Evaluation of a large-scale weight management program using the consolidated framework for implementation research (CFIR) Implementation Science 2013 8 1, article 51 1 17 10.1186/1748-5908-8-51 2-s2.0-84877264845 23279972 21 Aarons G. A. Glisson C. Green P. D. Hoagwood K. Kelleher K. J. Landsverk J. A. The organizational social context of mental health services and clinician attitudes toward evidence-based practice: a United States national study Implementation Science 2012 7 1, article 56 10.1186/1748-5908-7-56 2-s2.0-84862568168 22 Manuel J. K. Hagedorn H. J. Finney J. W. Implementing evidence-based psychosocial treatment in specialty substance use disorder care Psychology of Addictive Behaviors 2011 25 2 225 237 10.1037/a0022398 2-s2.0-79959291786 21668085 23 Proposals on organization and implementation of basic and major public health services in Shanghai 2012, http://www.hs.sh.cn/website/b/81111.shtml 24 Dai T. You C. M. He P. Analysis on functional operation and major reform advances of public hospitals in China Chinese Journal of Health Policy 2009 12 8 14 21 25 Institute of Medicine Crossing the Quality Chasm: A New Health System for the 21st Century 2001 Washington, DC, USA National Academy Press 26 Mendel P. Meredith L. S. Schoenbaum M. Sherbourne C. D. Wells K. B. Interventions in organizational and community context: a framework for building evidence on dissemination and implementation in health services research Administration and Policy in Mental Health and Mental Health Services Research 2008 35 1-2 21 37 10.1007/s10488-007-0144-9 2-s2.0-37649013511 17990095 27 Greenhalgh T. Robert G. Macfarlane F. Bate P. Kyriakidou O. Diffusion of innovations in service organizations: systematic review and recommendations Milbank Quarterly 2004 82 4 581 629 10.1111/j.0887-378x.2004.00325.x 2-s2.0-12444346743 15595944 28 Guide to Community Preventive Services http://www.thecommunityguide.org/index.html 29 Lee E. Dobbins M. Decorby K. McRae L. Tirilis D. Husson H. An optimal search filter for retrieving systematic reviews and meta-analyses BMC Medical Research Methodology 2012 12, article 51 10.1186/1471-2288-12-51 2-s2.0-84859790190 30 Yost J. Dobbins M. Traynor R. DeCorby K. Workentine S. Greco L. Tools to support evidence-informed public health decision making BMC Public Health 2014 14 1, article 728 10.1186/1471-2458-14-728 2-s2.0-84904273715 31 EPPI Center 2016, http://eppi.ioe.ac.uk/cms/ 32 Zheng Y. Gong Y. M. Research and practice of screening for colorectal cancer in population of Shanghai China Cancer 2013 22 2 86 89 33 Allen P. Sequeira S. Jacob R. R. Promoting state health department evidence-based cancer and chronic disease prevention: a multi-phase dissemination study with a cluster randomized trial component Implementation Science 2013 8 1, article 141 10.1186/1748-5908-8-141 2-s2.0-84889958855 34 Yost J. Mackintosh J. Read K. Dobbins M. Promoting awareness of key resources for evidence-informed decision-making in public health: an evaluation of a webinar series about knowledge translation methods and tools Frontiers in Public Health 2016 4, article 72 10.3389/fpubh.2016.00072 35 Lee E. Needs assessment for the computer-interpretable hypertension guideline at public health centers in Korea 129 Proceedings of the 12th World Congress on Health (Medical) Informatics 2007 p. 1455 36 Aljunid S. M. Srithamrongsawat S. Chen W. Health-care data collecting, sharing, and using in Thailand, China Mainland, South Korea, Taiwan, Japan, and Malaysia Value in Health 2012 15 supplement 1 S132 S138 10.1016/j.jval.2011.11.004 2-s2.0-84862924721 22265060

PMC005xxxxxx/PMC5002289.txt

==== Front Can J Gastroenterol HepatolCan J Gastroenterol HepatolCJGHCanadian Journal of Gastroenterology & Hepatology2291-27892291-2797Hindawi Publishing Corporation 10.1155/2016/2878149Review ArticleColorectal Cancer Screening in Average Risk Populations: Evidence Summary http://orcid.org/0000-0002-0264-2176Tinmouth Jill 1 2 3 4 5 * http://orcid.org/0000-0001-9304-7956Vella Emily T. 6 * Baxter Nancy N. 1 4 7 8 Dubé Catherine 1 9 http://orcid.org/0000-0003-2784-9576Gould Michael 10 11 Hey Amanda 12 http://orcid.org/0000-0003-2074-1896Ismaila Nofisat 13 http://orcid.org/0000-0003-1291-4638McCurdy Bronwen R. 1 http://orcid.org/0000-0002-3595-1493Paszat Lawrence 2 3 4 1Prevention and Cancer Control, Cancer Care Ontario, Toronto, ON, Canada2Department of Medicine, Sunnybrook Health Sciences Centre, Toronto, ON, Canada3Institute for Clinical Evaluative Sciences, Toronto, ON, Canada4Institute of Health Policy, Management and Evaluation, Dalla Lana School of Public Health, University of Toronto, Toronto, ON, Canada5Department of Medicine, University of Toronto, Toronto, ON, Canada6Program in Evidence-Based Care, Cancer Care Ontario, Hamilton, ON, Canada7Department of Surgery, St. Michael's Hospital, Toronto, ON, Canada8Li Ka Shing Knowledge Institute, St. Michael's Hospital, Toronto, ON, Canada9Department of Medicine, Division of Gastroenterology, University of Ottawa, The Ottawa Hospital, Ottawa, ON, Canada10William Osler Health Centre, Etobicoke, ON, Canada11Vaughan Endoscopy Clinic, Vaughan, ON, Canada12Northeast Cancer Centre Health Sciences North/Horizon Santé-Nord, Sudbury Outpatient Centre, Sudbury, ON, Canada13American Society of Clinical Oncology, Alexandria, VA, USA*Jill Tinmouth: jill.tinmouth@sunnybrook.ca and *Emily T. Vella: ccopgi@mcmaster.caAcademic Editor: Mark Borgaonkar 2016 14 8 2016 2016 28781491 4 2016 29 6 2016 Copyright © 2016 Jill Tinmouth et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Introduction. The objectives of this systematic review were to evaluate the evidence for different CRC screening tests and to determine the most appropriate ages of initiation and cessation for CRC screening and the most appropriate screening intervals for selected CRC screening tests in people at average risk for CRC. Methods. Electronic databases were searched for studies that addressed the research objectives. Meta-analyses were conducted with clinically homogenous trials. A working group reviewed the evidence to develop conclusions. Results. Thirty RCTs and 29 observational studies were included. Flexible sigmoidoscopy (FS) prevented CRC and led to the largest reduction in CRC mortality with a smaller but significant reduction in CRC mortality with the use of guaiac fecal occult blood tests (gFOBTs). There was insufficient or low quality evidence to support the use of other screening tests, including colonoscopy, as well as changing the ages of initiation and cessation for CRC screening with gFOBTs in Ontario. Either annual or biennial screening using gFOBT reduces CRC-related mortality. Conclusion. The evidentiary base supports the use of FS or FOBT (either annual or biennial) to screen patients at average risk for CRC. This work will guide the development of the provincial CRC screening program. ==== Body 1. Introduction Canada has one of the highest rates of colorectal cancer (CRC) in the world, with an estimated 25,100 cases in 2015 [1, 2]. CRC is also one of the leading causes of cancer-related death for men and women combined in Canada, with an estimated 9300 deaths in Canada in 2015. However, if CRC is found in its early stages, there is a 90% chance that it can be cured [1]. Cancers detected through screening tend to be earlier stage compared with cancers detected outside of screening [3–6]. In 2008, Ontario launched its CRC screening program, which offers screening to Ontarians aged 50 to 74. People at average risk are offered the guaiac fecal occult blood test (gFOBT) once every two years, while people at increased risk, defined as having one or more first-degree relatives with CRC, are offered colonoscopy. In 2013, approximately 58.5% of Ontarians were up-to-date with CRC screening tests [7]. In light of emerging evidence, the provincial CRC screening program is seeking guidance for CRC screening of average risk individuals in Ontario. Cancer Care Ontario's Prevention and Cancer Control portfolio and the Program in Evidence-Based Care (PEBC) developed this evidentiary base to help inform the CRC screening program in Ontario, ColonCancerCheck. The evidence supporting primary screening tests for CRC, ages of initiation and cessation for CRC screening, and screening intervals for selected CRC screening tests in people at average risk for CRC was systematically reviewed to develop this evidentiary base. Below, the methods and key findings of the systematic review are summarized. The full evidentiary base is available online [8]. 2. Systematic Review Objectives The purpose of this systematic review is to evaluate the existing evidence concerning screening of adults at average risk for CRC in the context of an organized, population-based screening program. The main objectives are to identify the following:The benefits and harms of screening in this population. The optimal primary CRC screening test(s) for this population. The appropriate ages of initiation and cessation for screening in this population. The intervals at which people at average risk should be recalled for CRC screening. 3. Target Population The target population includes primary care providers, endoscopists, policy-makers, and program planners in Ontario. 4. Research Questions 4.1. Primary Research Question Primary Research Question is as follows:How do different screening tests, individually or in combination, perform in average risk people in preventing CRC-related mortality or all-cause mortality or in decreasing the incidence of CRC? Secondary outcomes include the detection of cancer or its precursors, screening participation rate, adverse effects of tests, and test characteristics, such as sensitivity, specificity, positive predictive value, negative predictive value, and proportion of false-positives or of false-negatives. 4.2. Secondary Research Questions Secondary Research Questions are as follows:What are the appropriate ages of initiation and cessation for screening in people at average risk for CRC? Is there a relationship between age and the effectiveness of CRC screening? What are the appropriate intervals between CRC screening tests (by test)? Is there a relationship between screening intervals and the effectiveness and risks of screening? 5. Methods The authors of this evidentiary base (working group) consisted of one primary care physician, one colorectal surgeon, one expert in public health screening, one policy analyst from the Ontario CRC screening program, two methodologists, and three gastroenterologists. The PEBC, a provincial program of Cancer Care Ontario, is supported by the Ontario Ministry of Health and Long-Term Care. All work produced by the PEBC and any associated programs is editorially independent from the ministry. A two-stage method was used. It is summarized here and described in more detail as follows:Search and evaluation of existing systematic reviews: if existing systematic reviews were identified that addressed the research questions and were of reasonable quality, then they were included as a part of the evidentiary base. Original systematic review of the primary literature: this review focused on areas not covered by existing and accepted reviews. 5.1. Literature Search Strategy A systematic search was conducted in OVID MEDLINE (2006 to September 3, 2014), EMBASE (2006 to September 3, 2014), the Cochrane Library (Issue: 2–4, October 2013), and the American Society of Clinical Oncology (ASCO) conference proceedings (2009 to 2013). Details of the literature search strategy can be found online [8]. 5.1.1. Study Selection Criteria and Protocol Systematic reviews were included ifthey addressed at least one of the research questions, they evaluated randomized or nonrandomized control trials of asymptomatic average risk subjects undergoing CRC screening, the literature search strategy for the existing systematic review was reproducible (i.e., reported) and appropriate, the existing systematic review reported the sources searched, as well as the dates that were searched. Identified systematic reviews were assessed using the Assessing of Methodological Quality of Systematic Reviews (AMSTAR) tool [82]. In cases where multiple systematic reviews were identified for a particular outcome, only evidence from the most recent systematic review with the highest quality was used in the evidence base. The literature was searched for new primary studies published after the end search date of included systematic reviews. Individual study quality from the studies included in the systematic reviews as well as any new primary studies was assessed in order to complete the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) tables for risk of bias [9]. If no existing systematic review was identified for a given test or question, or if identified reviews were incomplete, a systematic review of the primary literature was performed. Articles in reference lists from included studies were also searched. The scope of the primary literature review was tailored to address the gaps in the incorporated existing systematic reviews (e.g., subject areas covered and time frames covered). The criteria for the primary literature are described as follows. Inclusion Criteria. Inclusion criteria are as follows:(1) Randomized controlled trials (RCTs) (primary research question and secondary research questions 1 and 2) that could be identified directly from the search or from reference sections of systematic reviews. (2) Cohort/case-control studies, minimum study size n = 30 (secondary research questions 1 and 2). (3) Evidence from nonrandomized prospective comparative studies with historical or contemporaneous controls, with the consensus of the working group, when there were gaps in available evidence from RCTs. (4) Studies preferred with asymptomatic average risk subjects and population-based studies that did not oversample adults with symptoms of CRC or a family history of CRC which were also considered acceptable. (5) For conference abstracts: RCTs (all questions). (6) The following screening tests considered for inclusion: (i) fecal-based tests including gFOBT, fecal immunochemical test (FIT), stool DNA panel (stool DNA), and fecal M2-PK, (ii) blood tests (Cologic®, ColonSentry®, mSEPT9, metabolomics, and hydroxylated polyunsaturated long chain fatty acids), (iii) endoscopic tests including flexible sigmoidoscopy (FS), colonoscopy, and capsule colonoscopy, (iv) radiological tests including double-contrast barium enema (DCBE) and computed tomography colonography (CT colonography). Exclusion Criteria. Exclusion criteria are as follows:Letters, comments, or editorials. Studies that included a population enriched with subjects with symptoms of CRC or a family history of CRC. Nonsystematic reviews. Non-English-language publications. One of the two reviewers (NI and EV) independently reviewed the titles and abstracts resulting from the search. For items that warranted full-text review, NI or EV reviewed each item independently. However, in uncertain cases, a second reviewer (JT) was asked to review them. 5.1.2. Data Extraction and Assessment of Study Quality and Potential for Bias Data from the included studies were independently extracted by NI and EV. If there was more than one publication for the same study, only the most updated or recent versions of the data were reported in the result. All extracted data and information were audited by an independent auditor. Important quality features, such as randomization details, sample size and power, intention-to-screen (ITS) analysis, length of follow-up, and funding, for each RCT, were extracted. The quality of observational studies was assessed using a modified Newcastle-Ottawa Scale [83]. The quality of diagnostic studies was assessed using a modified Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool [84]. The GRADE method for assessing the quality of aggregate evidence was used for each comparison using the GRADEpro Guideline Development Tool [9, 85]. The working group used the GRADE system for ranking outcomes and scored each outcome from the evidence review on a scale from 1 to 9. Outcomes with a score from 1 to 3 were considered of limited importance, from 4 to 6 important, and from 7 to 9 critical in the development of recommendations for the CRC screening program. Only outcomes that were considered critical or important were included in the GRADE evidence tables. 5.1.3. Synthesizing the Evidence When clinically homogenous results from two or more trials were available, a meta-analysis was conducted using review manager software (RevMan 5.3) provided by the Cochrane Collaboration [86]. For all outcomes, the dichotomous model with random effects was used. The number of person-years, rather than the total number of subjects, was used, if available. The number of person-years takes into account the fact that different people in the study may have been followed up for different lengths of time. In order to have comparable control rates across all gFOBT and FS trials, the control rates for the no screening groups in the gFOBT and FS trials were combined and calculated from the total number of cases across all gFOBT and FS trials over the total number of person-years across all gFOBT and FS trials. Statistical heterogeneity was calculated using χ 2 test for heterogeneity and I 2 percentage. A probability level for χ 2 statistic less than or equal to 10% (p ≤ 0.10) and/or an I 2 greater than 50% was considered indicative of statistical heterogeneity. 5.1.4. Process for Developing Conclusions The working group members met in person on four occasions to develop evidence-based conclusions through consensus. For each comparison (e.g., gFOBT versus no screening) the working group assessed the quality of the body of evidence for each outcome using the GRADE process [9]. Five factors were assessed for each outcome in each comparison: the risk of bias, inconsistency, indirectness, imprecision, and publication bias. Observational studies were initially graded as low quality and RCTs as high quality; the quality of the evidence was downgraded when serious threats were identified to one or more factors. At the in person meetings, the working group discussed each comparison and agreed on the overall certainty of the evidence across outcomes (Table 1), whether the desirable anticipated effects were large, whether the undesirable anticipated effects were small, and whether the desirable effects were large relative to the undesirable effects. For each comparison, conclusions were developed that reflected these working group discussions. 6. Results 6.1. Literature Search Results A total of 7538 studies were identified and 378 were selected for full-text review. Of those, 48 met the predefined eligibility criteria for this systematic review. An additional 27 articles were found from the reference lists. After our literature search, we became aware of and included an updated publication for one of the FS screening RCTs that had already been identified [21]. A total of 76 articles were included of which eight were systematic reviews [10, 29, 30, 59, 68, 87–89], 39 [6, 11–28, 60–66, 69–81] were from 30 RCTs, 19 were prospective studies [31–43, 52, 53, 55–57, 67], five were retrospective studies [44–48], and five were case-control studies [49–51, 54, 58]. Evidence from five of the eight systematic reviews was included either because the reviews were the most recent systematic review with the highest quality evidence for a particular outcome or because they included an outcome of interest not covered by other high-quality reviews [10, 29, 30, 59, 68]. After the search process and quality assessment, a total of 73 articles were included in this systematic review. The search flow diagram, the characteristics and quality of the included systematic reviews and studies, and the meta-analyses can be found online or in Supplementary Figures  1 to  19 (see Supplementary Material available online at http://dx.doi.org/10.1155/2016/2878149) [8]. Table 2 provides a summary of the number and type of studies used for each comparison. 7. Interpretation and Conclusions The following are the conclusions developed by the working group based on the review of the evidence and meta-analyses. When discussing the effects of various screening tests, the reported outcomes vary by test. There was strong agreement among the members of the working group that CRC-related mortality and complications from screening tests were critical outcomes for recommendation development. All-cause mortality, CRC incidence, participation rate, and diagnostic outcomes were considered important outcomes of interest. 7.1. Fecal Tests for Occult Blood There was strong evidence to support the use of fecal tests for occult blood to screen people at average risk for CRC. 7.1.1. Guaiac Fecal Occult Blood Test (gFOBT) versus No Screening The overall certainty of the evidence was high, suggesting a definite reduction in CRC-related mortality (Table 3 and Supplementary Figures  1 to  3). The magnitude of the effect was small (relative risk [RR], 0.87; 95% confidence interval [CI], 0.82 to 0.92). The magnitude of benefit was comparable to the disease-specific reduction in mortality from mammography for breast cancer screening (RR, 0.79; 95% CI, 0.68 to 0.90) [90] but was less than that from the human papillomavirus (HPV) test for cervical cancer screening (hazard ratio [HR], 0.52; 95% CI, 0.33 to 0.83) [91]. The anticipated harms associated with gFOBT (including follow-up colonoscopy for people with positive tests) are small and outweighed by the benefits. 7.1.2. Fecal Immunochemical Test (FIT) versus gFOBT The overall certainty of the evidence was moderate (Table 4 and Supplementary Figures  7 to  10). The magnitude of the desirable anticipated effects was at least equivalent to gFOBT, and it is likely that the desirable effects of FIT are greater than for gFOBT. The anticipated undesirable effects associated with FIT (including follow-up colonoscopy for people with positive tests) are small and outweighed by the benefits. While there were well-designed randomized controlled trials (RCTs) comparing FIT with gFOBT, the outcomes of these trials (participation and detection rates) were considered to be less important than CRC-related mortality. However, it was anticipated that the reduction in CRC-related mortality and the complications resulting from screening with FIT would be at least equivalent to those observed from screening with gFOBT. FIT's greater sensitivity for detection of CRC and advanced adenomas compared with gFOBT suggests that the reduction in CRC incidence with FIT could be greater than with gFOBT; however, the magnitude and significance of any additional benefit of FIT over gFOBT are unknown. It is important to highlight that the FIT positivity threshold selected would be an important determinant of the magnitude of the benefits and harms of FIT relative to gFOBT. 7.2. Lower Bowel Endoscopy There was strong evidence to support the use of flexible sigmoidoscopy (FS) to screen people at average risk for CRC. There was no direct evidence to support the use of colonoscopy to screen people at average risk for CRC, but evidence from FS informed the assessment of the benefits and harms of colonoscopy in screening people at average risk for CRC. 7.2.1. FS versus No Screening The overall certainty of the evidence was high, suggesting that FS has a definite effect on CRC-related mortality and incidence when compared with no screening (Table 5 and Supplementary Figures  4 to  6). The magnitude of the effect on CRC mortality was modest (RR, 0.72; 95% CI, 0.65 to 0.80); it exceeds the anticipated disease-specific reduction in mortality from gFOBT for CRC screening (RR, 0.87; 95% CI, 0.82 to 0.92) and is similar to the effects of mammography on breast cancer mortality (RR, 0.79; 95% CI, 0.68 to 0.90) [90] and of the HPV test on cervical cancer mortality (HR, 0.52; 95% CI, 0.33 to 0.83) [91]. The effect on survival with FS was also comparable to the benefit achieved with the current standard of care for patients with completely resected stage III CRC (5-fluorouracil/leucovorin plus oxaliplatin [FOLFOX or FLOX] versus 5-fluorouracil/leucovorin alone, HR for overall survival at six years, 0.80; 95% CI, 0.65 to 0.97) [92]. The anticipated harms associated with FS (including follow-up colonoscopy for people with positive tests) were small and outweighed by the benefits. 7.2.2. Colonoscopy versus No Screening The overall certainty of direct evidence supporting the use of colonoscopy to screen people at average risk for CRC was very low when compared with no screening (Table 6). The desirable and undesirable anticipated effects were uncertain. It is anticipated that the benefit of screening with colonoscopy would be at least equivalent to that observed for screening with FS; however, the magnitude of additional benefit over FS, if any, is unknown. The magnitude of additional undesirable effects of colonoscopy relative to FS is also unknown. 7.3. Fecal Tests for Occult Blood versus Lower Bowel Endoscopy There was insufficient evidence to determine how fecal tests for occult blood perform compared with lower bowel endoscopy to screen people at average risk for CRC (Supplementary Tables  1 to  5 and Supplementary Figures  11 to  19).The studies that compared one-time fecal tests for occult blood to lower bowel endoscopy were heterogeneous, with few comparisons where data could be pooled. However, in general, the evidence suggested that participation was higher and detection rate was lower with fecal-based tests compared with endoscopic tests. The overall certainty of the evidence was low. CRC-related mortality was not evaluated and the design of the studies favoured endoscopic tests because the comparison was to one-time fecal-based testing (rather than repeated testing over time, which is how these tests are used in usual practice). There was significant heterogeneity in participation. The undesirable anticipated effects of endoscopy (including follow-up endoscopy for people with positive fecal tests) are probably small. It is uncertain whether the desirable effects are large relative to the undesirable effects. 7.4. Radiological Tests 7.4.1. Computed Tomography Colonography versus Colonoscopy There was insufficient evidence to determine how computed tomography colonography performs compared with colonoscopy to screen people at average risk for CRC (results not shown; see [8]).The overall certainty of the evidence was low. The desirable and undesirable anticipated effects were uncertain. 7.4.2. Capsule Colonoscopy versus Colonoscopy There was insufficient evidence to determine how capsule colonoscopy performs compared with colonoscopy to screen people at average risk for CRC (results not shown; see [8]).The overall certainty of the evidence was very low. The desirable and undesirable anticipated effects were uncertain. 7.4.3. Double-Contrast Barium Enema (DCBE) There was no evidence to support the use of DCBE to screen people at average risk for CRC.Since 2006, there has been no new published evidence on this topic. Most recent CRC guidelines except for a 2008 guideline by the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology [93] have not endorsed the use of DCBE for screening [94–98]. 7.5. DNA Tests 7.5.1. Stool DNA versus Fecal Occult Blood Tests (gFOBT or FIT) There was insufficient evidence to determine how stool DNA performs compared with gFOBT or FIT to screen people at average risk for CRC (results not shown; see [8]).The overall certainty of the evidence was very low. The desirable and undesirable anticipated effects were uncertain. 7.5.2. Other DNA Tests There was insufficient evidence to support the use of mSEPT9 to screen people at average risk for CRC (results not shown; see [8]).The overall certainty of the evidence was very low. The desirable and undesirable anticipated effects were uncertain. 7.6. Metabolomic Tests 7.6.1. Fecal M2-PK There was insufficient evidence to support the use of fecal M2-PK to screen people at average risk for CRC (results not shown; see [8]).The overall certainty of the evidence was very low. The desirable and undesirable anticipated effects were uncertain. 7.6.2. Other Metabolomic Tests There was no evidence to support the use of other metabolomic tests (e.g., low levels of hydroxylated polyunsaturated long chain fatty acids [Cologic]) to screen people at average risk for CRC. 7.7. Age of Initiation/Cessation 7.7.1. Age of Initiation/Cessation with gFOBT Currently, the Ontario CRC screening program recommends that average risk individuals initiate screening with gFOBT beginning at 50 years of age and ending at age 74. There was insufficient evidence to support changing the ages of initiation and cessation for CRC screening with gFOBT in Ontario (results not shown; see [8]).The overall certainty of the evidence was very low. There was insufficient evidence to demonstrate differences in reduction of CRC mortality using gFOBT across age groups. The desirable and undesirable anticipated effects across age groups were uncertain. 7.7.2. Age of Initiation/Cessation with FS There was insufficient evidence to recommend ages of initiation or cessation when screening with FS in people at average risk for CRC (results not shown; see [8]).The overall certainty of the evidence was very low. There was insufficient evidence to demonstrate differences in reduction of CRC mortality or incidence using FS across age groups. The desirable and undesirable anticipated effects across age groups were uncertain. Of the four large FS RCTs, three examined “once in a lifetime” FS between the ages of 55 and 64, while the fourth RCT examined baseline FS between the ages of 55 and 74 with a second FS after three or five years. 7.7.3. Age of Initiation/Cessation with Colonoscopy There was insufficient evidence to recommend an age of initiation or cessation to screen with colonoscopy in people at average risk for CRC (results not shown; see [8]).The overall certainty of the evidence was very low. There was insufficient evidence to demonstrate differences in CRC detection using colonoscopy across age groups. The desirable and undesirable anticipated effects across age groups were uncertain. Currently, the Ontario CRC screening program does not recommend colonoscopy to screen persons at average risk for CRC. The program does recommend colonoscopy in people at increased risk (one or more first-degree relatives with CRC) starting at 50 years of age or 10 years younger than the age at which the relative was diagnosed, whichever occurred first. 7.7.4. Age of Initiation/Cessation with FIT There were no studies that met our inclusion criteria for age of initiation/cessation for FIT. 7.8. Screening Intervals 7.8.1. gFOBT Intervals There was evidence to suggest that either annual or biennial screening using gFOBT in people at average risk for CRC reduces CRC-related mortality (results not shown; see [8]).The overall certainty of the evidence was moderate. The desirable anticipated effects on CRC mortality were small and similar for annual or biennial screening. The undesirable anticipated effects were not reported for each interval group. Anticipated harms associated with gFOBT (including follow-up colonoscopy for people with positive tests) were small for biennial screening and were likely to be greater for annual screening. In addition, annual screening is anticipated to increase burden to the participant. 7.8.2. FIT Intervals There was insufficient evidence to recommend an interval to screen people at average risk for CRC using FIT (results not shown; see [8]).The overall certainty of the evidence was very low. The desirable and undesirable anticipated effects were uncertain. 7.8.3. FS and Colonoscopy Intervals There were no studies that met our inclusion criteria for screening intervals for FS or colonoscopy. 8. Summary and Next Steps This evidentiary base summarizes the known clinical effectiveness and safety of CRC screening tests. Concurrently, the Canadian Task Force on Preventive Health Care (CTFPH) [99] and the US Preventive Services Task Force (USPSTF) have published guidelines on CRC screening [100]. The audience for the 3 reviews differed slightly: the current review seeks to provide guidance to the CRC screening program in light of emerging evidence in CRC screening while the CTFPH and the USPSTF specify “primary care physicians” and “primary care clinicians and patients” as their target audiences, respectively. All 3 reviews support the use of fecal testing (gFOBT or FIT) and flexible sigmoidoscopy but treat colonoscopy slightly differently. The CTFPH recommends against colonoscopy while the USPSTF endorses it as one of a range of options. In the current review, we grade the evidence only (recommendations for Ontario's CRC screening program are released separately; see below). Our interpretation of the evidence acknowledges that the strong evidence in favour of flexible sigmoidoscopy does inform the assessment of colonoscopy. However, this evidence is both indirect and incomplete (magnitude of additional benefit from colonoscopy and of additional harms is unknown). In order to have complete understanding of the benefits and harms of colonoscopy, we await the results of ongoing randomized controlled trials in colonoscopy for average risk screening, anticipated in the 2020s. The evidence from the current review is central to the ongoing development of Ontario's CRC screening program. However, this evidentiary base is necessary but not sufficient to guide program development as other context-specific criteria such as cost-effectiveness, existing program design, and public acceptability and feasibility (from an organizational and economic perspective) must be considered. In addition, the program must also consider the balance between choice and informed decision making and issues not well addressed by the evidence such as how to best implement CRC screening when there is more than one CRC screening test supported by high-quality evidence. An expert panel comprising members from national and international screening programs, primary care physicians, general surgeons, gastroenterologists, pathologists and laboratory medicine professionals, nurse endoscopists, and members of the public was convened to provide guidance on how to incorporate this evidence in light of the other issues listed above. Their level of agreement with the conclusions is reflected in Table 7. The CCC program will use findings from the evidence summary as well as expert panel recommendations to guide its ongoing development. The specific recommendations resulting from this process have been released recently and can be accessed online at https://www.cancercare.on.ca/common/pages/UserFile.aspx?fileId=358486. Supplementary Material The supplementary material includes the GRADE evidence profiles and meta-analyses comparing fecal tests for occult blood with lower bowel endoscopy, the meta-analyses comparing screening with gFOBT or screening with FS versus no screening, and the meta-analyses comparing screening with FIT versus screening with gFOBT. Acknowledgments The authors would like to thank the following individuals for their assistance in developing this report: (i) Melissa Brouwers, Meghan Hatcher, Sheila McNair, and Hans Messersmith for providing feedback on draft versions, (ii) Waseem Hijazi for conducting a data audit, and (iii) Sara Miller and Jenny Lass for copyediting. Competing Interests Jill Tinmouth was a lead scientist at Cancer Care Ontario for ColonCancerCheck and paid as a consultant for this work. Catherine Dubé published an editorial in Can J. Gastro 2012 26 417-18 [101]. Michael Gould has an endoscopy clinic in Toronto, is a consultant for Abbott Laboratories, is on the Board of Directors for the Ontario Association of Gastroenterology and the Canadian Digestive Health Foundation, and is the Clinical Lead for ColonCancerCheck. For the remaining authors none were declared. The Program in Evidence-based Care (PEBC) is a provincial initiative of Cancer Care Ontario supported by the Ontario Ministry of Health and Long-Term Care. All work produced by the PEBC is editorially independent from the Ontario Ministry of Health and Long-Term Care. Table 1 Description of the quality of evidence grades according to Grading of Recommendations, Assessment, Development and Evaluations (GRADE) [9]. Grade Definition High We are very confident that the true effect lies close to that of the estimate of the effect Moderate We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect Very low We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect Table 2 Summary of included studies by research question. Research question and references∗ Systematic reviews Outcomes RCTs Prospective studies Retrospective studies Case-control studies Primary question: CRC screening tests ∗∗ gFOBT versus no screening [10–20] 1 CRC mortality 4       Complications from tests 3       All-cause mortality 4       CRC incidence 5       FS versus no screening [6, 10, 21–28] 1 CRC mortality 4       Complications from tests 5       All-cause mortality 4       CRC incidence 4       Colonoscopy versus no screening [29–51] 2 CRC mortality   2 1   Complications from tests   11 4   CRC incidence   2   3 mSEPT9 alone [52]   Diagnostic test accuracy outcomes   1     Fecal M2-PK alone [53, 54]   Diagnostic test accuracy outcomes   1   1 Stool DNA versus gFOBT or FIT [55–58]   Diagnostic test accuracy outcomes   3   1 FIT versus gFOBT [59–65] 1 Complications from tests 1       CRC/advanced adenoma detection rate (ITS) 5       CRC/advanced adenoma detection rate (PP) 5       Participation rate 6       Diagnostic test accuracy outcomes, false-positives/total screened 5       CT colonography versus colonoscopy [66]   Complications from tests 1       CRC/advanced adenoma detection rate (ITS) 1       CRC/advanced adenoma detection rate (PP) 1       Participation rate 1       Capsule colonoscopy versus colonoscopy [67]   Complications from tests   1     Adenoma detection rate (PP)   1     Fecal-based tests versus endoscopy [10, 59, 63, 68–80]              FIT versus colonoscopy 3 Complications from tests 2       CRC/advanced adenoma detection rate (ITS) 3       Participation rate 3        FIT versus FS Complications from tests 1       CRC/advanced adenoma detection rate (ITS) 3       Participation rate 3        gFOBT versus colonoscopy Complications from tests 1       Participation rate 2        gFOBT versus FS Complications from tests 1       CRC/advanced adenoma detection rate (ITS) 2       Participation rate 4      gFOBT versus gFOBT + FS Complications from tests 1       CRC/advanced adenoma detection rate (ITS) 2       Participation rate 3       Secondary question number 1: age of initiation and cessation by test gFOBT versus no screening [19, 20]   CRC mortality by age group 2       FS versus no screening [6, 21, 23, 26]   CRC mortality by age group 2       CRC incidence by age group 4       Colonoscopy versus no screening [50]   CRC risk by age group       1 Secondary question number 2: screening interval by test gFOBT [15, 16, 20]   CRC mortality by interval (annual versus biennial) 1       All-cause mortality by interval (annual versus biennial) 1       CRC incidence by interval (annual versus biennial) 1       FIT [81]   CRC/advanced adenoma detection rate by interval (1 versus 2 versus 3 years) 1       Participation rate by interval (1 versus 2 versus 3 years) 1       CRC: colorectal cancer; CT: computed tomographic; DNA: deoxyribonucleic acid; FIT: fecal immunochemical test; FS: flexible sigmoidoscopy; gFOBT: guaiac fecal occult blood test; ITS: intention to screen; PP: per protocol; RCT: randomized controlled trial. ∗Some RCTs published multiple articles. ∗∗Tests that were included in our search strategy but yielded no results were not included. Table 3 GRADE evidence profile—gFOBT versus no screening. Quality assessment Number of patients Effect Quality1 Importance Number of studies Design Risk of bias Inconsistency Indirectness Imprecision Other considerations gFOBT (cases/person-years) No screening (cases/person-years) Relative (95% CI) Absolute (cases/person-years) CRC mortality (followup: range 17–30 years) 4 Randomised trials Not serious Not serious Not serious Not serious Not serious 2027/2674854 (0.08%) 2326/2669246 (0.09%) RR 0.87  (0.82 to 0.92) 113 fewer per 1000000 (from 70 fewer to 157 fewer) ⨁⨁⨁⨁ High Critical Control (gFOBT + FS) = 0.06%∗ 78 fewer per 1000000 (from 48 fewer to 108 fewer) Complications from tests (from Holme et al. 2013 [10])2 3 Randomized trials Not serious Not serious Not serious Not serious Not serious     N/A3   ⨁⨁⨁⨁ High Critical All-cause mortality (follow-up: range 17–30 years) 4 Randomized trials Not serious Not serious Not serious Not serious Not serious 74,481/2,674,854 (2.8%) 74,174/2,669,246 (2.8%) RR 1  (0.99 to 1.01) 0 fewer per 1,000,000 (from 278 fewer to 278 more) ⨁⨁⨁⨁ High Important Control (gFOBT + FS) = 1.85%∗ 0 fewer per 1000000 (from 185 fewer to 185 more) CRC incidence (follow-up: range 17–30 years) 5 Randomized trials Not serious Not serious Serious4 Not serious Not serious 4324/2,434,487 (0.2%) 4489/2,431,961 (0.2%) RR 0.96 (0.9 to 1.02) 74 fewer per 1,000,000 (from 37 more to 185 fewer) ⨁⨁⨁◯ Moderate Important Control (gFOBT + FS) = 0.16%∗ 64 fewer per 1,000,000 (from 32 more to 160 fewer) CI: confidence interval; CRC: colorectal cancer; FS: flexible sigmoidoscopy; gFOBT: guaiac fecal occult blood test; GRADE: Grading of Recommendations, Assessment, Development and Evaluations; ITS: intention to screen; N/A: not applicable; RR: relative risk. ∗In order to have comparable control rates across all gFOBT and FS trials, the control rates for the no screening groups in the gFOBT and FS trials were combined and calculated from the total number of cases across all gFOBT and FS trials over the total number of person-years across all gFOBT and FS trials. 1GRADE working group grades of evidence: (i) High quality: we are very confident that the true effect lies close to that of the estimate of effect. (ii) Moderate quality: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. (iii) Low quality: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. (iv) Very low quality: we have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect. 2Major complication defined as bleeding, perforation, or death within 30 days of screening, follow-up colonoscopy, or surgery. 3Holme et al. 2013 [10] reported a major complication rate of 0.03%. 4Goteborg trial used sigmoidoscopy and double-contrast barium enema as reference standard; other trials used colonoscopy. Table 4 GRADE evidence profile—gFOBT versus FIT. Quality assessment Number of patients Effect Quality Importance Number of studies Design Risk of bias Inconsistency Indirectness Imprecision Other considerations FIT gFOBT Relative (95% CI) Absolute Complications from tests 1 Randomized trial Not serious Not serious Not serious Serious1 Not serious     Not pooled   ⨁⨁⨁◯  Moderate Critical CRC/advanced adenoma detection rate (ITS) 5 Randomized trials Not serious Not serious Serious2 Not serious Not serious 278/24,288 (1.1%) 129/27,346 (0.5%) RR 2.15 (1.58–2.94) 5 more per 1000 (from 3 more to 9 more) ⨁⨁⨁◯  Moderate Important CRC/advanced adenoma detection rate (PP) 5 Randomized trials Not serious Not serious Serious2 Not serious Not serious 278/12,146 (2.3%) 129/10,976 (1.2%) RR 1.83 (1.30–2.57) 10 more per 1000 (from 4 more to 18 more) ⨁⨁⨁◯  Moderate Important False-positive screening test results 5 Randomized trials Not serious Serious3 Not serious Not serious Not serious 385/24,288 (1.6%) 188/27,346 (0.7%) RR 2.12 (1.02–4.39) 8 more per 1000 (from 0 fewer to 23 more) ⨁⨁⨁◯  Moderate Important Participation rate 6 Randomized trials Not serious Serious4 Not serious Not serious Not serious 12,271/24,490 (50.1%) 11075/27,548 (40.2%) RR 1.16 (1.05–1.28) 64 more per 1000 (from 20 more to 113 more) ⨁⨁⨁◯  Moderate Important CI: confidence interval; CRC: colorectal cancer; FIT: fecal immunochemical test; gFOBT: guaiac fecal occult blood test; GRADE: Grading of Recommendations, Assessment, Development and Evaluations; ITS: intention to screen; PP: per protocol; RR: relative risk. 1Only one study. 2Surrogate outcome for CRC mortality. 3Heterogeneity: Tau2 = 0.61; Chi2 = 56.96, df = 3 (p < 0.00001); I 2 = 93%. 4Heterogeneity: Tau2 = 0.01; Chi2 = 107.18, df = 5 (p < 0.00001); I 2 = 95%. Table 5 GRADE evidence profile—FS versus no screening. Quality assessment Number of patients Effect Quality Importance Number of studies Design Risk of bias Inconsistency Indirectness Imprecision Other considerations FS (cases/person-years) No screening (cases/person-years) Relative (95% CI) Absolute (cases/person-years) CRC mortality (follow-up: 6–12 years) 4 Randomized trials Not serious Not serious Not serious Not serious Not serious 576/1,902,184 (0.03%) 1321/3,114,546 (0.04%) RR 0.72  (0.65–0.80) 119 fewer per 1,000,000 (from 85 fewer to 148 fewer) ⨁⨁⨁⨁  High Critical Control (gFOBT + FS) = 0.06%∗ 168 fewer per 1,000,000 (from 120 fewer to 210 fewer) Complications from tests (from Holme  et al. 2013 [10])1 5 Randomized trials Not serious Not serious Not serious Not serious Not serious   N/A2     ⨁⨁⨁⨁  High Critical All-cause mortality (follow-up: 6–12 years) 4 Randomized trials Not serious Not serious Not serious Not serious Not serious 19,525/1,902,184 (1.0%) 32,903/3,114,546 (1.1%) RR 0.97  (0.96–0.99) 317 fewer per 1,000,000 (from 106 fewer to 423 fewer) ⨁⨁⨁⨁   High Important Control (gFOBT + FS) = 1.85%∗ 555 fewer per 1,000,000 (from 185 fewer to 740 fewer) CRC incidence (follow-up: 6–12 years) 4 Randomized trials Not serious Not serious Not serious Not serious Not serious 2218/1,860,990 (0.1%) 4579/3,067,081 (0.1%) RR 0.78  (0.74–0.83) 328 fewer per 1,000,000 (from 254 fewer to 388 fewer) ⨁⨁⨁⨁  High Important Control (gFOBT + FS) = 0.16%∗ 352 fewer per 1,000,000 (from 272 fewer to 416 fewer) CI: confidence interval; CRC: colorectal cancer; FS: flexible sigmoidoscopy; gFOBT: guaiac fecal occult blood test; GRADE: Grading of Recommendations, Assessment, Development and Evaluations; RR: relative risk; N/A: not applicable. ∗In order to have comparable control rates across all gFOBT and FS trials, the control rates for the no screening groups in the gFOBT and FS trials were combined and calculated from the total number of cases across all gFOBT and FS trials over the total number of person-years across all gFOBT and FS trials. 1Major complication rate included bleeding, perforation, or death within 30 days of screening, follow-up colonoscopy, or surgery. 2Holme et al. 2013 [10] reported a major complication rate of 0.08%. Table 6 GRADE evidence profile—colonoscopy and CRC-related mortality and incidence. Quality assessment Effect Quality Importance Number of studies Design Risk of bias Inconsistency Indirectness Imprecision Other considerations Relative (95% CI) CRC mortality (from Brenner et al. 2014) [29] 3 Observational studies Serious1 Not serious Not serious Not serious Not serious RR 0.32 (0.23–0.43) ⨁◯◯◯  Very low Critical Complications from tests (perforations, bleeding, and deaths) 15 Observational studies Not serious Not serious Not serious Not serious Not serious N/A2 ⨁⨁◯◯  Low Critical CRC incidence (from Brenner et al. 2014) [29] 5 Observational studies Serious1 Serious3 Not serious Not serious Not serious RR 0.31 (0.12 to 0.77) ⨁◯◯◯  Very low Important CI: confidence interval; CRC: colorectal cancer; GRADE: Grading of Recommendations, Assessment, Development and Evaluations; N/A: not applicable. 1Mixed study designs included case-control and retrospective. 2The risks of perforation or bleeding were less than 1% ranging from 0% to 0.22% for perforations and 0% to 0.19% for bleeding. 3Heterogeneity: Tau2 = 1.0; (p < 0.0001); I 2 = 94%. Table 7 Responses of the expert panel to the working group's conclusions.   Reviewer ratings (N = 27) Conclusions Strongly disagree (%) Disagree (%) Neither agree nor disagree (%) Agree (%) Strongly agree (%) (1) Strong evidence to support use of fecal tests for occult blood to screen people at average risk for CRC 0 0 1 (4) 10 (37) 16 (59) (2) Strong evidence to support the use of FS to screen people at average risk for CRC 0 0 0 7 (26) 20 (74) (3) No direct evidence to support the use of colonoscopy to screen people at average risk for CRC, but evidence from FS informs the assessment of benefits and harms of colonoscopy to screen people at average risk for CRC 0 2 (8) 2 (8) 14 (54) 8 (31) (4) Insufficient evidence to determine how fecal tests for occult blood perform compared with lower bowel endoscopy to screen people at average risk for CRC 0 2 (8) 4 (15) 13 (50) 7 (27) (5) Insufficient evidence to determine how CT colonography performs compared with colonoscopy to screen people at average risk for CRC 0 0 1 (4) 8 (33) 15 (63) (6) Insufficient evidence to determine how capsule endoscopy performs compared with colonoscopy to screen people at average risk for CRC 0 0 0 1 (4) 23 (96) (7) No evidence to support the use of double-contrast barium enema to screen people at average risk for CRC 0 0 2 (8) 0 23 (92) (8) Insufficient evidence to determine how fecal DNA performs compared with guaiac fecal occult blood test (gFOBT) or fecal immunochemical test (FIT) to screen people at average risk for CRC 0 0 0 8 (33) 16 (67) (9) Insufficient evidence to support the use of mSEPT9 to screen people at average risk for CRC 0 0 0 2 (8) 23 (92) (10) Insufficient evidence to support the use of fecal M2-PK to screen people at average risk for CRC 0 0 0 3 (12) 23 (89) (11) Insufficient evidence to support the use of other metabolomic tests to screen people at average risk for CRC 0 0 1 (4) 2 (9) 20 (87) (12) Insufficient evidence to support changing ages of initiation and cessation for CRC screening with gFOBT in Ontario 0 1 (4) 1 (4) 9 (36) 14 (56) (13) Insufficient evidence to recommend an age of initiation or cessation to screen with FS in people at average risk for CRC 0 0 3 (12) 9 (36) 13 (52) (14) Insufficient evidence to recommend an age of initiation or cessation to screen with colonoscopy in people at average risk for CRC 0 0 4 (16) 10 (40) 11 (44) (15) Evidence suggests annual or biennial screening using gFOBT in people at average risk for CRC reduces CRC mortality 0 0 1 (4) 11 (42) 14 (54) (16) Insufficient evidence to recommend an interval to screen people at average risk for CRC using FIT 0 1 (4) 4 (15) 13 (50) 8 (31) ==== Refs 1 Colorectal Cancer Association of Canada, 2016, http://www.colorectal-cancer.ca/en/just-the-facts/colorectal/ 2 World Cancer Research Fund International website, 2016, http://www.wcrf.org/int/cancer-facts-figures/data-specific-cancers/colorectal-cancer-statistics 3 Brenner H. Chang-Claude J. Seiler C. M. Hoffmeister M. Interval cancers after negative colonoscopy: population-based case-control study Gut 2012 61 11 1576 1582 10.1136/gutjnl-2011-301531 2-s2.0-84864373288 22200840 4 Kronborg O. Fenger C. Olsen J. Jørgensen O. D. Søndergaard O. Randomised study of screening for colorectal cancer with faecal-occult-blood test The Lancet 1996 348 9040 1467 1471 10.1016/s0140-6736(96)03430-7 2-s2.0-0030607609 5 Morris E. J. A. Whitehouse L. E. Farrell T. A retrospective observational study examining the characteristics and outcomes of tumours diagnosed within and without of the English NHS Bowel Cancer Screening Programme British Journal of Cancer 2012 107 5 757 764 10.1038/bjc.2012.331 2-s2.0-84865406351 22850549 6 Segnan N. Armaroli P. Bonelli L. Once-only sigmoidoscopy in colorectal cancer screening: Follow-up findings of the italian randomized controlled trial - SCORE Journal of the National Cancer Institute 2011 103 17 1310 1322 10.1093/jnci/djr284 2-s2.0-80052676840 21852264 7 Canadian Quality Cancer of Ontario, 2016, http://www.csqi.on.ca/by_patient_journey/screening/colorectal_screening_participation 8 Tinmouth J. Vella E. Baxter NN. Colorectal cancer screening in average risk populations: evidence summary Program in Evidence-Based Care Evidence Summary 2015 15-14 Toronto, Canada CCO https://www.cancercare.on.ca/toolbox/qualityguidelines/clin-program/screening-ebs/ 9 Guyatt G. Oxman A. D. Akl E. A. GRADE guidelines: 1. Introduction—GRADE evidence profiles and summary of findings tables Journal of Clinical Epidemiology 2011 64 4 383 394 10.1016/j.jclinepi.2010.04.026 2-s2.0-79951952372 21195583 10 Holme Ø. Bretthauer M. Fretheim A. Odgaard-Jensen J. Hoff G. Flexible sigmoidoscopy versus faecal occult blood testing for colorectal cancer screening in asymptomatic individuals Cochrane Database of Systematic Reviews 2013 9 CD009259 2-s2.0-84890939875 11 Kewenter J. Brevinge H. Endoscopic and surgical complications of work-up in screening for colorectal cancer Diseases of the Colon and Rectum 1996 39 6 676 680 10.1007/BF02056949 2-s2.0-0029943153 8646956 12 Kronborg O. Jørgensen O. D. Fenger C. Rasmussen M. Randomized study of biennial screening with a faecal occult blood test: results after nine screening rounds Scandinavian Journal of Gastroenterology 2004 39 9 846 851 10.1080/00365520410003182 2-s2.0-4544310710 15513382 13 Lindholm E. Brevinge H. Haglind E. Survival benefit in a randomized clinical trial of faecal occult blood screening for colorectal cancer British Journal of Surgery 2008 95 8 1029 1036 10.1002/bjs.6136 2-s2.0-48249127689 18563785 14 Mandel J. S. Bond J. H. Church T. R. Reducing mortality from colorectal cancer by screening for fecal occult blood. Minnesota Colon Cancer Control Study The New England Journal of Medicine 1993 328 19 1365 1371 10.1056/nejm199305133281901 2-s2.0-0027223762 8474513 15 Mandel J. S. Church T. R. Bond J. H. The effect of fecal occult-blood screening on the incidence of colorectal cancer The New England Journal of Medicine 2000 343 22 1603 1607 10.1056/nejm200011303432203 2-s2.0-0034735826 11096167 16 Mandel J. S. Church T. R. Ederer F. Bond J. H. Colorectal cancer mortality: effectiveness of biennial screening for fecal occult blood Journal of the National Cancer Institute 1999 91 5 434 437 10.1093/jnci/91.5.434 2-s2.0-0033518901 10070942 17 Paimela H. Malila N. Palva T. Hakulinen T. Vertio H. Järvinen H. Early detection of colorectal cancer with faecal occult blood test screening British Journal of Surgery 2010 97 10 1567 1571 10.1002/bjs.7150 2-s2.0-77956989959 20603855 18 Robinson M. H. E. Hardcastle J. D. Moss S. M. The risks of screening: data from the Nottingham randomised controlled trial of faecal occult blood screening for colorectal cancer Gut 1999 45 4 588 592 10.1136/gut.45.4.588 2-s2.0-0032879959 10486370 19 Scholefield J. H. Moss S. M. Mangham C. M. Whynes D. K. Hardcastle J. D. Nottingham trial of faecal occult blood testing for colorectal cancer: a 20-year follow-up Gut 2012 61 7 1036 1040 10.1136/gutjnl-2011-300774 2-s2.0-84861573867 22052062 20 Shaukat A. Mongin S. J. Geisser M. S. Long-term mortality after screening for colorectal cancer The New England Journal of Medicine 2013 369 12 1106 1114 10.1056/nejmoa1300720 2-s2.0-84884183059 24047060 21 Holme Ø. Løberg M. Kalager M. Effect of flexible sigmoidoscopy screening on colorectal cancer incidence and mortality: a randomized clinical trial The Journal of the American Medical Association 2014 312 6 606 615 10.1001/jama.2014.8266 2-s2.0-84905901559 25117129 22 Atkin W. S. Cook C. F. Cuzick J. Single flexible sigmoidoscopy screening to prevent colorectal cancer: baseline findings of a UK multicentre randomised trial The Lancet 2002 359 9314 1291 300 10.1016/s0140-6736(02)08268-5 23 Atkin W. S. Edwards R. Kralj-Hans I. Once-only flexible sigmoidoscopy screening in prevention of colorectal cancer: a multicentre randomised controlled trial The Lancet 2010 375 9726 1624 1633 10.1016/s0140-6736(10)60551-x 2-s2.0-77951879165 24 Hoff G. Sauar J. Vatn M. H. Polypectomy of adenomas in the prevention of colorectal cancer: 10 years' follow-up of the Telemark Polyp Study I. A prospective, controlled population study Scandinavian Journal of Gastroenterology 1996 31 10 1006 1010 10.3109/00365529609003121 2-s2.0-0029818613 8898422 25 Hoff G. Grotmol T. Skovlund E. Bretthauer M. Risk of colorectal cancer seven years after flexible sigmoidoscopy screening: randomised controlled trial British Medical Journal 2009 338 b1846 10.1136/bmj.b1846 2-s2.0-67650076135 26 Schoen R. E. Pinsky P. F. Weissfeld J. L. Colorectal-cancer incidence and mortality with screening flexible sigmoidoscopy The New England Journal of Medicine 2012 366 25 2345 2357 10.1056/nejmoa1114635 2-s2.0-84862583882 22612596 27 Segnan N. Senore C. Andreoni B. Baseline findings of the Italian multicenter randomized controlled trial of ‘once-only sigmoidoscopy’—SCORE Journal of the National Cancer Institute 2002 94 23 1763 1772 10.1093/jnci/94.23.1763 2-s2.0-0037021672 12464648 28 Thiis-Evensen E. Hoff G. S. Sauar J. Langmark F. Majak B. M. Vatn M. H. Population-based surveillance by colonoscopy: effect on the incidence of colorectal cancer: Telemark Polyp Study I Scandinavian Journal of Gastroenterology 1999 34 4 414 420 10.1080/003655299750026443 2-s2.0-0032947335 10365903 29 Brenner H. Stock C. Hoffmeister M. Effect of screening sigmoidoscopy and screening colonoscopy on colorectal cancer incidence and mortality: systematic review and meta-analysis of randomised controlled trials and observational studies British Medical Journal 2014 348, article g2467 10.1136/bmj.g2467 2-s2.0-84898657761 30 Niv Y. Hazazi R. Levi Z. Fraser G. Screening colonoscopy for colorectal cancer in asymptomatic people: a meta-analysis Digestive Diseases and Sciences 2008 53 12 3049 3054 10.1007/s10620-008-0286-y 2-s2.0-55949084566 18463980 31 Chiu H.-M. Wang H.-P. Lee Y.-C. A prospective study of the frequency and the topographical distribution of colon neoplasia in asymptomatic average-risk Chinese adults as determined by colonoscopic screening Gastrointestinal Endoscopy 2005 61 4 547 553 10.1016/S0016-5107(05)00121-5 2-s2.0-16244382879 15812407 32 DiSario J. A. Foutch P. G. Mai H. D. Pardy K. Manne R. K. Prevalence and malignant potential of colorectal polyps in asymptomatic, average-risk men The American Journal of Gastroenterology 1991 86 8 941 945 2-s2.0-0025887546 1858757 33 Imperiale T. F. Wagner D. R. Lin C. Y. Larkin G. N. Rogge J. D. Ransohoff D. F. Risk of advanced proximal neoplasms in asymptomatic adults according to the distal colorectal findings The New England Journal of Medicine 2000 343 3 169 174 10.1056/nejm200007203430302 2-s2.0-0034691750 10900275 34 Johnson D. A. Gurney M. S. Volpe R. J. A prospective study of the prevalence of colonic neoplasms in asymptomatic patients with an age-related risk The American Journal of Gastroenterology 1990 85 8 969 974 2-s2.0-0025364414 2375325 35 Kahi C. J. Imperiale T. F. Juliar B. E. Rex D. K. Effect of screening colonoscopy on colorectal cancer incidence and mortality Clinical Gastroenterology and Hepatology 2009 7 7 770 775 10.1016/j.cgh.2008.12.030 2-s2.0-67649373345 19268269 36 Kim D. H. Lee S. Y. Choi K. S. The usefulness of colonoscopy as a screening test for detecting colorectal polyps Hepatogastroenterology 2007 54 80 2240 2242 18265641 37 Kim D. H. Pickhardt P. J. Taylor A. J. CT colonography versus colonoscopy for the detection of advanced neoplasia The New England Journal of Medicine 2007 357 14 1403 1412 10.1056/nejmoa070543 2-s2.0-34948891641 17914041 38 Lieberman D. A. Weiss D. G. Bond J. H. Ahnen D. J. Garewal H. Chejfec G. Use of colonoscopy to screen asymptomatic adults for colorectal cancer. Veterans Affairs Cooperative Study Group 380 The New England Journal of Medicine 2000 343 3 162 168 10.1056/nejm200007203430301 2-s2.0-0034691827 10900274 39 Manser C. N. Bachmann L. M. Brunner J. Hunold F. Bauerfeind P. Marbet U. A. Colonoscopy screening markedly reduces the occurrence of colon carcinomas and carcinoma-related death: a closed cohort study Gastrointestinal Endoscopy 2012 76 1 110 117 10.1016/j.gie.2012.02.040 2-s2.0-84862679542 22498179 40 Regula J. Rupinski M. Kraszewska E. Colonoscopy in colorectal-cancer screening for detection of advanced neoplasia The New England Journal of Medicine 2006 355 18 1863 1872 10.1056/nejmoa054967 2-s2.0-33750528561 17079760 41 Rogge J. D. Elmore M. F. Mahoney S. J. Low-cost, office-based, screening colonoscopy The American Journal of Gastroenterology 1994 89 10 1775 1780 2-s2.0-0028000262 7942665 42 Schoenfeld P. Cash B. Flood A. Colonoscopic screening of average-risk women for colorectal neoplasia The New England Journal of Medicine 2005 352 20 2061 2068 10.1056/nejmoa042990 2-s2.0-20844438843 15901859 43 Soon M.-S. Kozarek R. A. Ayub K. Soon A. Lin T.-Y. Lin O. S. Screening colonoscopy in Chinese and Western patients: a comparative study The American Journal of Gastroenterology 2005 100 12 2749 2755 10.1111/j.1572-0241.2005.00355.x 2-s2.0-33644818822 16393230 44 Choe J. W. Chang H.-S. Yang S.-K. Screening colonoscopy in asymptomatic average-risk Koreans: analysis in relation to age and sex Journal of Gastroenterology and Hepatology 2007 22 7 1003 1008 10.1111/j.1440-1746.2006.04774.x 2-s2.0-34347387165 17608845 45 Chung S. J. Kim Y. S. Yang S. Y. Prevalence and risk of colorectal adenoma in asymptomatic Koreans aged 40–49 years undergoing screening colonoscopy Journal of Gastroenterology and Hepatology 2010 25 3 519 525 10.1111/j.1440-1746.2009.06147.x 2-s2.0-77649211474 20370730 46 Nishihara R. Wu K. Lochhead P. Long-term colorectal-cancer incidence and mortality after lower endoscopy The New England Journal of Medicine 2013 369 12 1095 1105 10.1056/nejmoa1301969 2-s2.0-84884189727 24047059 47 Prajapati D. N. Saeian K. Binion D. G. Volume and yield of screening colonoscopy at a tertiary medical center after change in medicare reimbursement The American Journal of Gastroenterology 2003 98 1 194 199 10.1111/j.1572-0241.2003.07172.x 2-s2.0-0037259269 12526957 48 Strul H. Kariv R. Leshno M. The prevalence rate and anatomic location of colorectal adenoma and cancer detected by colonoscopy in average-risk individuals aged 40-80 years American Journal of Gastroenterology 2006 101 2 255 262 10.1111/j.1572-0241.2006.00430.x 2-s2.0-33644889911 16454827 49 Brenner H. Chang-Claude J. Jansen L. Knebel P. Stock C. Hoffmeister M. Reduced risk of colorectal cancer up to 10 years after screening, surveillance, or diagnostic colonoscopy Gastroenterology 2014 146 3 709 717 10.1053/j.gastro.2013.09.001 2-s2.0-84894354632 24012982 50 Cotterchio M. Manno M. Klar N. McLaughlin J. Gallinger S. Colorectal screening is associated with reduced colorectal cancer risk: a case-control study within the population-based Ontario Familial Colorectal Cancer Registry Cancer Causes and Control 2005 16 7 865 875 10.1007/s10552-005-2370-3 2-s2.0-23844504991 16132797 51 Doubeni C. A. Weinmann S. Adams K. Screening colonoscopy and risk for incident late-stage colorectal cancer diagnosis in average-risk adults: a nested case-control study Annals of Internal Medicine 2013 5, part 1 312 320 10.7326/0003-4819-158-5-201303050-00003 23460054 52 Church T. R. Wandell M. Lofton-Day C. Prospective evaluation of methylated SEPT9 in plasma for detection of asymptomatic colorectal cancer Gut 2014 63 2 317 325 10.1136/gutjnl-2012-304149 2-s2.0-84891743167 23408352 53 Haug U. Hundt S. Brenner H. Sensitivity and specificity of faecal tumour M2 pyruvate kinase for detection of colorectal adenomas in a large screening study British Journal of Cancer 2008 99 1 133 135 10.1038/sj.bjc.6604427 2-s2.0-46349099567 18542075 54 Tonus C. Neupert G. Sellinger M. Colorectal cancer screening by non-invasive metabolic biomarker fecal tumor M2-PK World Journal of Gastroenterology 2006 12 43 7007 7011 10.3748/wjg.v12.i43.7007 2-s2.0-33751513942 17109496 55 Ahlquist D. A. Sargent D. J. Loprinzi C. L. Stool DNA and occult blood testing for screen detection of colorectal neoplasia Annals of Internal Medicine 2008 149 7 441 450 10.7326/0003-4819-149-7-200810070-00004 2-s2.0-53749091845 18838724 56 Imperiale T. F. Ransohoff D. F. Itzkowitz S. H. Multitarget stool DNA testing for colorectal-cancer screening The New England Journal of Medicine 2014 370 14 1287 1297 10.1056/nejmoa1311194 2-s2.0-84897509377 24645800 57 Imperiale T. F. Ransohoff D. F. Itzkowitz S. H. Turnbull B. A. Ross M. E. Fecal DNA versus fecal occult blood for colorectal-cancer screening in an average-risk population The New England Journal of Medicine 2004 351 26 2704 2714 10.1056/nejmoa033403 2-s2.0-10844281210 15616205 58 Koga Y. Yamazaki N. Takizawa S. Gene expression analysis using a highly sensitive DNA microarray for colorectal cancer screening Anticancer Research 2014 34 1 169 176 2-s2.0-84897022707 24403458 59 Hassan C. Giorgi Rossi P. Camilloni L. Meta-analysis: adherence to colorectal cancer screening and the detection rate for advanced neoplasia, according to the type of screening test Alimentary Pharmacology and Therapeutics 2012 36 10 929 940 10.1111/apt.12071 2-s2.0-84867574204 23035890 60 Chubak J. Bogart A. Fuller S. Laing S. S. Green B. B. Uptake and positive predictive value of fecal occult blood tests: a randomized controlled trial Preventive Medicine 2013 57 5 671 678 10.1016/j.ypmed.2013.08.032 2-s2.0-84885835204 24029556 61 Federici A. Rossi P. G. Borgia P. Bartolozzi F. Farchi S. Gausticchi G. The immunochemical faecal occult blood test leads to higher compliance than the guaiac for colorectal cancer screening programmes: a cluster randomized controlled trial Journal of Medical Screening 2005 12 2 83 88 10.1258/0969141053908357 2-s2.0-20644468405 15949119 62 Hoffman R. M. Steel S. Yee E. F. T. Massie L. Schrader R. M. Murata G. H. Colorectal cancer screening adherence is higher with fecal immunochemical tests than guaiac-based fecal occult blood tests: a randomized, controlled trial Preventive Medicine 2010 50 5-6 297 299 10.1016/j.ypmed.2010.03.010 2-s2.0-77952583613 20307568 63 Hol L. Van Leerdam M. E. Van Ballegooijen M. Screening for colorectal cancer: randomised trial comparing guaiac-based and immunochemical faecal occult blood testing and flexible sigmoidoscopy Gut 2010 59 1 62 68 10.1136/gut.2009.177089 2-s2.0-73449097516 19671542 64 Levi Z. Birkenfeld S. Vilkin A. A higher detection rate for colorectal cancer and advanced adenomatous polyp for screening with immunochemical fecal occult blood test than guaiac fecal occult blood test, despite lower compliance rate. A prospective, controlled, feasibility study International Journal of Cancer 2011 128 10 2415 2424 10.1002/ijc.25574 2-s2.0-79953169888 20658527 65 van Rossum L. G. van Rijn A. F. Laheij R. J. Random comparison of guaiac and immunochemical fecal occult blood tests for colorectal cancer in a screening population Gastroenterology 2008 135 1 82 90 10.1053/j.gastro.2008.03.040 2-s2.0-46049103362 18482589 66 Stoop E. M. de Haan M. C. de Wijkerslooth T. R. Participation and yield of colonoscopy versus non-cathartic CT colonography in population-based screening for colorectal cancer: a randomised controlled trial The Lancet Oncology 2012 13 1 55 64 10.1016/s1470-2045(11)70283-2 2-s2.0-84855329616 22088831 67 Groth S. Krause H. Behrendt R. Capsule colonoscopy increases uptake of colorectal cancer screening BMC Gastroenterology 2012 12, article 80 10.1186/1471-230x-12-80 2-s2.0-84862700953 68 Littlejohn C. Hilton S. MacFarlane G. J. Phull P. Systematic review and meta-analysis of the evidence for flexible sigmoidoscopy as a screening method for the prevention of colorectal cancer British Journal of Surgery 2012 99 11 1488 1500 10.1002/bjs.8882 2-s2.0-84867103707 23001715 69 Berry D. P. Clarke P. Hardcastle J. D. Vellacott K. D. Randomized trial of the addition of flexible sigmoidoscopy to faecal occult blood testing for colorectal neoplasia population screening British Journal of Surgery 1997 84 9 1274 1276 10.1002/bjs.1800840922 2-s2.0-0030768664 9313712 70 Brevinge H. Lindholm E. Buntzen S. Kewenter J. Screening for colorectal neoplasia with faecal occult blood testing compared with flexible sigmoidoscopy directly in a 55-56 years' old population International Journal of Colorectal Disease 1997 12 5 291 295 10.1007/s003840050108 2-s2.0-0030701834 9401844 71 Federici A. Marinacci C. Mangia M. Borgia P. Giorgi Rossi P. Guasticchi G. Is the type of test used for mass colorectal cancer screening a determinant of compliance?. A cluster-randomized controlled trial comparing fecal occult blood testing with flexible sigmoidoscopy Cancer Detection and Prevention 2006 30 4 347 353 10.1016/j.cdp.2006.03.009 2-s2.0-33748923945 16965874 72 Gondal G. Grotmol T. Hofstad B. Bretthauer M. Eide T. J. Hoff G. The Norwegian Colorectal Cancer Prevention (NORCCAP) screening study: baseline findings and implementations for clinical work-up in age groups 50–64 years Scandinavian Journal of Gastroenterology 2003 38 6 635 642 10.1080/00365520310003002 2-s2.0-0038649963 12825872 73 Lisi D. Hassan C. Crespi M. Group A. S. Participation in colorectal cancer screening with FOBT and colonoscopy: an Italian, multicentre, randomized population study. [Erratum appears in Dig Liver Dis. 2012 Feb;44(2):182 Note: Hassan, C Cesare [corrected to Hassan, Cesare]] Digestive and Liver Disease 2012 42 5 371 376 19747888 74 Multicentre Australian Colorectal-neoplasia Screening (MACS) Group A comparison of colorectal neoplasia screening tests: a multicentre community-based study of the impact of consumer choice Medical Journal of Australia 2006 184 11 546 550 16768659 75 Quintero E. Castells A. Bujanda L. Colonoscopy versus fecal immunochemical testing in colorectal-cancer screening The New England Journal of Medicine 2012 366 8 697 706 10.1056/nejmoa1108895 2-s2.0-84857422764 22356323 76 Rasmussen M. Kronborg O. Fenger C. Jørgensen O. D. Possible advantages and drawbacks of adding flexible sigmoidoscopy to hemoccult-II in screening for colorectal cancer. A randomized study Scandinavian Journal of Gastroenterology 1999 34 1 73 78 10.1080/00365529950172862 2-s2.0-0032973615 10048736 77 Segnan N. Senore C. Andreoni B. Randomized trial of different screening strategies for colorectal cancer: patient response and detection rates Journal of the National Cancer Institute 2005 97 5 347 357 10.1093/jnci/dji050 2-s2.0-20044394716 15741571 78 Segnan N. Senore C. Andreoni B. Comparing attendance and detection rate of colonoscopy with sigmoidoscopy and FIT for colorectal cancer screening Gastroenterology 2007 132 7 2304 2312 10.1053/j.gastro.2007.03.030 2-s2.0-34250006056 17570205 79 Verne J. E. C. W. Aubrey R. Love S. B. Talbot I. C. Northover J. M. A. Population based randomised study of uptake and yield of screening by flexible sigmoidoscopy compared with screening by faecal occult blood testing British Medical Journal 1998 317 7152 182 185 10.1136/bmj.317.7152.182 2-s2.0-0032543792 9665902 80 Zauber A. G. Winawer S. J. Mills G. M. Adherence to screening in a randomized controlled trial of a one-time screening colonoscopy versus program of annual fecal occult blood test (gFOBt): implications of lower gFOBt adherence to screening on colorectal cancer mortality reduction Gastroenterology 2012 142 5 S82 S83 10.1016/s0016-5085(12)60312-6 81 van Roon A. H. C. Goede S. L. van Ballegooijen M. Random comparison of repeated faecal immunochemical testing at different intervals for population-based colorectal cancer screening Gut 2013 62 3 409 415 10.1136/gutjnl-2011-301583 2-s2.0-84873410891 22387523 82 Shea B. J. Grimshaw J. M. Wells G. A. Development of AMSTAR: a measurement tool to assess the methodological quality of systematic reviews BMC Medical Research Methodology 2007 7, article 10 7 10.1186/1471-2288-7-10 2-s2.0-33847606952 17291355 83 Wells G. A. Shea B. O'Connell D. The Newcastle-Ottawa Scale (NOS) for Assessing the Quality of Nonrandomised Studies in Meta-Analyses 2012 Ottawa, Canada Ottawa Hospital Research Institute http://www.ohri.ca/programs/clinical_epidemiology/oxford.htm 84 Whiting P. Rutjes A. W. S. Reitsma J. B. Bossuyt P. M. M. Kleijnen J. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews BMC Medical Research Methodology 2003 3, article 25 10.1186/1471-2288-3-25 2-s2.0-84927565732 85 Brożek J. Nowak A. Kunstman P. Schünemann H. J. GRADEpro Guideline Development Tool (G2DT) 2014 (Computer Program) Version 2.xx http://www.guidelinedevelopment.org 86 Review Manager (Rev Man). 4.2 for Windows 2003 Copenhagen, Denmark The Nordic Cochrane Centre, The Cochrane Collaboration 87 Elmunzer B. J. Hayward R. A. Schoenfeld P. S. Saini S. D. Deshpande A. Waljee A. K. Effect of flexible sigmoidoscopy-based screening on incidence and mortality of colorectal cancer: a systematic review and meta-analysis of randomized controlled trials PLoS Medicine 2012 9 12 e1001352 10.1371/journal.pmed.1001352 2-s2.0-84871666335 88 Hewitson P. Glasziou P. Irwig L. Towler B. Watson E. Screening for colorectal cancer using the faecal occult blood test, Hemoccult Cochrane Database of Systematic Reviews 2011 1 CD001216 89 Massat N. J. Moss S. M. Halloran S. P. Duffy S. W. Screening and primary prevention of colorectal cancer: a review of sex-specific and site-specific differences Journal of Medical Screening 2013 20 3 125 148 10.1177/0969141313501292 2-s2.0-84901372841 24197771 90 Canadian Task Force on Preventive Health Care Recommendations on screening for breast cancer in average-risk women aged 40–74 years Canadian Medical Association Journal 2011 183 17 1991 2001 10.1503/cmaj.110334 22106103 91 Sankaranarayanan R. Nene B. M. Shastri S. S. HPV screening for cervical cancer in rural India The New England Journal of Medicine 2009 360 14 1385 1394 10.1056/nejmoa0808516 2-s2.0-63849251690 19339719 92 André T. Boni C. Navarro M. Improved overall survival with oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment in stage II or III colon cancer in the MOSAIC trial Journal of Clinical Oncology 2009 27 19 3109 3116 10.1200/jco.2008.20.6771 2-s2.0-67650315187 19451431 93 Levin B. Lieberman D. A. McFarland B. Screening and surveillance for the early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American cancer society, the US multi-society task force on colorectal cancer, and the American college of radiology Gastroenterology 2008 134 5 1570 1595 10.1053/j.gastro.2008.02.002 2-s2.0-42949127635 18384785 94 Scottish Intercollegiate Guidelines Network (SIGN) Diagnosis and Management of Colorectal Cancer. A National Clinical Guideline 2011 Edinburgh, UK Scottish Intercollegiate Guidelines Network (SIGN) SIGN Publication no. 126 95 European Commission European Guidelines for Quality Assurance in Colorectal Cancer Screening and Diagnosis 2010 1st Rue Mercier, Luxembourg Publications Office of the European Union Edited by N. Segnan, J. Patnick, L. von Karsa 96 U.S. Preventive Services Task Force Screening for colorectal cancer: U.S. Preventive Services Task Force recommendation statement Annals of Internal Medicine 2008 149 9 627 637 10.7326/0003-4819-149-9-200811040-00243 18838716 97 Leddin D. The Canadian Association of Gastroenterology position on colon cancer screening Canadian Journal of Gastroenterology 2003 17 2 p. 133 10.1155/2003/837621 98 Rex D. K. Johnson D. A. Anderson J. C. American College of Gastroenterology guidelines for colorectal cancer screening The American Journal of Gastroenterology 2009 104 3 739 750 19240699 99 Canadian Task Force on Preventive Health C Recommendations on screening for colorectal cancer in primary care Canadian Medical Association Journal 2016 188 5 340 348 10.1503/cmaj.151125 26903355 100 US Preventive Services Task Force Screening for colorectal cancer: us preventive services task force recommendation statement The Journal of the American Medical Association 2016 315 23 2564 2575 10.1001/jama.2016.5989 27304597 101 Dubé C. Tackling colorectal cancer as a public health issue: what can the gastroenterologist do? Canadian Journal of Gastroenterology 2012 26 7 417 418 10.1155/2012/507989 2-s2.0-84863757308 22803014

PMC005xxxxxx/PMC5002290.txt

==== Front Case Rep Obstet GynecolCase Rep Obstet GynecolCRIOGCase Reports in Obstetrics and Gynecology2090-66842090-6692Hindawi Publishing Corporation 10.1155/2016/4169565Case ReportBleeding versus Clotting: A Complex Case of a Large Fibroid Uterus Causing Menorrhagia and a DVT http://orcid.org/0000-0001-8195-1200Ramanan Sangeeta 1 * http://orcid.org/0000-0002-4268-3896Chapman-Wardy Jude 2 http://orcid.org/0000-0002-9457-0359Watson Roy 3 1Lyell McEwin Hospital, Adelaide, SA 5112, Australia2Modbury Hospital, Adelaide, SA 5092, Australia3The Queen Elizabeth Hospital, Adelaide, SA 5011, Australia*Sangeeta Ramanan: sangeeta.ramanan@gmail.comAcademic Editor: Eing Mei Tsai 2016 14 8 2016 2016 416956516 4 2016 13 7 2016 Copyright © 2016 Sangeeta Ramanan et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.A 43-year-old woman presented with severe anaemia secondary to menorrhagia. Pelvic ultrasound showed a large intramural posterior fundal fibroid. Hysteroscopy showed the fibroid distorting the endometrial cavity, precluding Mirena® device insertion. As she was initially hesitant to have a hysterectomy, medical management with the oral contraceptive pill (OCP) and tranexamic acid was instituted, with good effect. Months later, after a long road trip, she presented with left leg swelling, and a Doppler ultrasound confirmed an extensive deep vein thrombosis (DVT). She was commenced on warfarin for anticoagulation but presented again with menorrhagia precipitated by overanticoagulation. After initial stabilization with multiple blood transfusions and reversal of anticoagulation, the warfarin was ceased in favour of enoxaparin and she underwent inferior vena cava (IVC) filter insertion prior to a total abdominal hysterectomy. Mass effect from large uterine fibroids can cause venous thromboembolism (VTE). A duplex ultrasound of the lower limbs if a woman presents with a large fibroid could identify asymptomatic DVTs in such women. A prehysterectomy IVC filter would then reduce their risk of postoperative pulmonary embolism. Medical management of menorrhagia with procoagulants should be avoided for management of menorrhagia in such women given their higher risk of developing VTE. ==== Body 1. Introduction Virchow's triad of venous stasis, endothelial damage, and hypercoagulability has been used to describe the pathogenesis of venous thromboembolism for over a century [1]. Venous stasis is caused by immobility due to recent surgery or illness or a sedentary lifestyle. Stasis is also due to any obstruction to blood flow. Endothelial damage results from trauma, such as lower limb fractures, or from excess venous compression during surgery or travel. A hypercoagulable state is caused by multiple factors, such as malignancy, pregnancy, polycythaemia, thrombocytosis, acquired and inherited thrombophilias, use of exogenous oestrogen (e.g., the oral contraceptive pill or hormone replacement therapy) [2], and other medications such as tranexamic acid [3]. Leiomyomas or uterine fibroids are benign, monoclonal tumours of the smooth muscle cells in the uterus [4]. Women with fibroids can remain asymptomatic [5] or may have symptoms of abnormal uterine bleeding (menorrhagia, prolonged bleeding, etc.), dyspareunia, and pelvic pain [6]. Women with leiomyomas are also at an increased risk of developing venous thromboembolism due to multiple different mechanisms. Polycythaemia and reactive thrombocytosis have been seen in women with menorrhagia due to fibroids [2], and these are risk factors for venous thromboembolism [7, 8]. Mass effect from benign space occupying lesions, including large uterine fibroids, can result in venous stasis of the lower limbs, leading to venous thromboembolism (VTE) [2, 9–12]. Pathological, nonuniform enlargement of the uterus is more likely to cause impingement of pelvic veins than uniform enlargement of the uterus (as seen in pregnancy) [2]. We report a case of a multiple risk factors leading to a massive deep venous thrombosis in a woman with a large uterine leiomyoma. 2. Case Presentation A 43-year-old woman presented with severe anaemia secondary to menorrhagia. She was symptomatic of dizziness, shortness of breath on exertion, and worsening fatigue. She reported having increasingly heavy and painful periods for 2 years, requiring 5-6 pads per day. She had regular 30-day cycles, and her periods lasted 5-6 days. She had previously tried oral contraceptive pills (OCP) but had experienced headaches with their use and hence had ceased them. She was a nonsmoker and had had one vaginal delivery 16 years ago. On examination, she was found to be tachycardic but normotensive. She had a soft nontender abdomen. Her uterus was enlarged and palpable and corresponded to 16 weeks' gestation. On vaginal examination, there was a mass felt in the pouch of Douglas. Her cervix appeared normal. Her haemoglobin was 55 g/L. Pelvic ultrasound showed an intramural posterior fundal fibroid measuring 9.3 × 9.1 × 10.8 cm. Endometrial thickness was 3 mm. After initial resuscitation with IV fluids and 3 units of packed red blood cells, she underwent a hysteroscopy, which showed the fibroid distorting the endometrial cavity, precluding endometrial biopsy and Mirena® device placement. As she was initially hesitant to have a hysterectomy, medical management was instituted with tranexamic acid and a different OCP than that used previously, with good effect. Fifteen months later, she presented again on day 2 of her periods after a syncopal episode at home. She had a postural drop in her blood pressure (99 mmHg to 76 mmHg systolic from lying to standing). She was admitted for observation and given tranexamic acid to stop her bleeding. Her haemoglobin dropped from 89 g/L to 65 g/L during that admission. She was again counselled regarding permanent methods of management of menorrhagia and consented to a hysterectomy. While awaiting her hysterectomy, she went on a long road trip and presented again with left leg swelling and tightness. She denied chest pain or shortness of breath. Circumferential measurements of her legs supported the diagnosis of a deep vein thrombosis (DVT) (Table 1). Doppler ultrasound of the left limb showed an extensive deep vein thrombosis involving the iliac, femoral, popliteal, posterior tibial, and gastrocnemius veins (Figure 1). The uterine fibroid was found to be compressing the left common iliac vein. She was commenced on warfarin for anticoagulation but presented again with menorrhagia and was found to be severely anaemic and overanticoagulated. Her haemoglobin was 54 g/L and INR 6.6. She was stabilized with multiple blood transfusions and reversal of anticoagulation. After consultation with a haematologist, her warfarin was ceased in favour of enoxaparin. She underwent fluoroscopy-guided inferior vena cava (IVC) filter insertion (Figure 2) prior to a total abdominal hysterectomy and bilateral salpingectomy. Significant endometriosis was an unexpected incidental intraoperative finding. Histology confirmed a benign leiomyoma measuring 14 × 9 cm with foci of red degeneration and minor adenomyosis (Figure 3). The uterus weighed 1490 g. The patient was restarted on enoxaparin 6 hours postoperatively and underwent removal of the IVC filter a few weeks after the hysterectomy. 3. Discussion In this case, the patient had multiple factors contributing to the development of her DVT. The large fibroid uterus compressed her left common iliac vein, causing venous stasis. She also had been on a long road trip, and this period of prolonged immobility resulted in further venous stasis. Her severe menorrhagia was treated with the oral contraceptive pill and tranexamic acid, both of which have a procoagulant effect. Women with large uterine fibroids may have asymptomatic DVTs that they and their clinicians may be unaware of. A hysterectomy to treat symptoms caused by the fibroid(s) could potentially release the occlusive effect of the fibroid uterus on deep pelvic veins. This may inadvertently lead to a pulmonary embolus immediately after hysterectomy, which can be fatal. Duplex ultrasound of the lower limbs should be considered if a woman presents with a large fibroid uterus. This could identify asymptomatic DVTs in such women. A prehysterectomy IVC filter would then reduce their risk of postoperative pulmonary embolism. Large fibroids usually result in menorrhagia. In many instances, this is managed medically with the oral contraceptive pill and/or tranexamic acid, while awaiting definitive surgery. However, the procoagulant effect of these drugs increases the risk of venous thromboembolism in such women [13, 14]. Medical management of menorrhagia with procoagulants (e.g., the oral contraceptive pill or tranexamic acid) should be avoided for management of menorrhagia in such women given their higher risk of developing VTE. Thromboprophylaxis must be offered to women with large leiomyomas who plan to undertake long trips [15]. Competing Interests The authors declare that there is no conflict of interests. Figure 1 Doppler US of left lower limb showing DVT in the left common femoral (a) and femoral veins (b). Figure 2 Venogram showing IVC filter. Figure 3 Cross section of uterus with large leiomyoma. Table 1 Calf and thigh circumferences of the patient's lower limbs.   Left Right Calf circumference (cm) 37.5 34 Thigh circumference (cm) 51.5 43.5 ==== Refs 1 Kumar D. R. Hanlin E. R. Glurich I. Mazza J. J. Yale S. H. Virchow's contribution to the understanding of thrombosis and cellular biology Clinical Medicine and Research 2010 8 3-4 168 172 10.3121/cmr.2009.866 2-s2.0-78650840507 20739582 2 Fletcher H. Wharfe G. Williams N. P. Gordon-Strachan G. Pedican M. Brooks A. Venous thromboembolism as a complication of uterine fibroids: a retrospective descriptive study Journal of Obstetrics and Gynaecology 2009 29 8 732 736 10.3109/01443610903165545 2-s2.0-70350107370 19821668 3 Mannucci P. M. Levi M. Prevention and treatment of major blood loss The New England Journal of Medicine 2007 356 22 2301 2311 10.1056/nejmra067742 2-s2.0-34249821653 17538089 4 Parker W. H. Etiology, symptomatology, and diagnosis of uterine myomas Fertility and Sterility 2007 87 4 725 736 10.1016/j.fertnstert.2007.01.093 2-s2.0-34147162657 17430732 5 Okolo S. Incidence, aetiology and epidemiology of uterine fibroids Best Practice and Research: Clinical Obstetrics and Gynaecology 2008 22 4 571 588 10.1016/j.bpobgyn.2008.04.002 2-s2.0-44949260958 18534913 6 Zimmermann A. Bernuit D. Gerlinger C. Schaefers M. Geppert K. Prevalence, symptoms and management of uterine fibroids: an international internet-based survey of 21,746 women BMC Women's Health 2012 12, article 6 10.1186/1472-6874-12-6 2-s2.0-84858744917 7 Akins P. T. Glenn S. Nemeth P. M. Derdeyn C. P. Carotid artery thrombus associated with severe iron-deficiency anemia and thrombocytosis Stroke 1996 27 5 1002 1005 10.1161/01.STR.27.5.1002 2-s2.0-0029881912 8623090 8 Voigt W. Jordan K. Sippel C. Amoury M. Schmoll H.-J. Wolf H. H. Severe thrombocytosis and anemia associated with celiac disease in a young female patient: a case report Journal of Medical Case Reports 2008 2, article 96 10.1186/1752-1947-2-96 2-s2.0-45149120160 9 Shiota M. Kotani Y. Umemoto M. Deep-vein thrombosis is associated with large uterine fibroids The Tohoku Journal of Experimental Medicine 2011 224 2 87 89 10.1620/tjem.224.87 2-s2.0-79957614244 21576892 10 Rosenfeld H. Byard R. W. Lower extremity deep venous thrombosis with fatal pulmonary thromboembolism caused by benign pelvic space-occupying lesions—an overview Journal of Forensic Sciences 2012 57 3 665 668 10.1111/j.1556-4029.2011.02047.x 2-s2.0-84860184249 22268621 11 Nishikawa H. Ideishi M. Nishimura T. Deep venous thrombosis and pulmonary thromboembolism associated with a huge uterine myoma: a case report Angiology 2000 51 2 161 166 10.1177/000331970005100210 2-s2.0-0033965960 10701725 12 Riat R. Chowdary P. Mavrides E. Magos A. Gatt A. Is there an Association between Thrombosis and Fibroids? A single centre experience and literature review International Journal of Laboratory Hematology 2013 35 1 e13 e16 10.1111/ijlh.12020 2-s2.0-84872232122 23113896 13 Vandenbroucke J. P. Rosing J. Bloemenkamp K. W. M. Oral contraceptives and the risk of venous thrombosis The New England Journal of Medicine 2001 344 20 1527 1535 10.1056/nejm200105173442007 2-s2.0-0035902193 11357157 14 Sundström A. Seaman H. Kieler H. Alfredsson L. The risk of venous thromboembolism associated with the use of tranexamic acid and other drugs used to treat menorrhagia: a case-control study using the General Practice Research Database BJOG: An International Journal of Obstetrics and Gynaecology 2009 116 1 91 97 10.1111/j.1471-0528.2008.01926.x 2-s2.0-57349119262 19016686 15 Philbrick J. T. Shumate R. Siadaty M. S. Becker D. M. Air travel and venous thromboembolism: a systematic review Journal of General Internal Medicine 2007 22 1 107 114 10.1007/s11606-006-0016-0 2-s2.0-34248598449 17351849

PMC005xxxxxx/PMC5002291.txt

==== Front Biomed Res IntBiomed Res IntBMRIBioMed Research International2314-61332314-6141Hindawi Publishing Corporation 10.1155/2016/2071945Research ArticleWear Scar Similarities between Retrieved and Simulator-Tested Polyethylene TKR Components: An Artificial Neural Network Approach http://orcid.org/0000-0002-9914-0059Orozco Villaseñor Diego A. 1 2 3 http://orcid.org/0000-0001-6169-3873Wimmer Markus A. 1 2 * 1Orthopedic Surgery, Rush University Medical Center, 1611 W. Harrison Street, Chicago, IL 60612, USA2Bioengineering, University of Illinois at Chicago, 851 S. Morgan Street, Chicago, IL 60607, USA3Departamento de Bioingeniería, Tecnológico de Monterrey, Campus Guadalajara, Guadalajara, JAL, Mexico*Markus A. Wimmer: markus_a_wimmer@rush.eduAcademic Editor: Padhraig O'Loughlin 2016 14 8 2016 2016 207194520 4 2016 22 6 2016 Copyright © 2016 D. A. Orozco Villaseñor and M. A. Wimmer.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.The aim of this study was to determine how representative wear scars of simulator-tested polyethylene (PE) inserts compare with retrieved PE inserts from total knee replacement (TKR). By means of a nonparametric self-organizing feature map (SOFM), wear scar images of 21 postmortem- and 54 revision-retrieved components were compared with six simulator-tested components that were tested either in displacement or in load control according to ISO protocols. The SOFM network was then trained with the wear scar images of postmortem-retrieved components since those are considered well-functioning at the time of retrieval. Based on this training process, eleven clusters were established, suggesting considerable variability among wear scars despite an uncomplicated loading history inside their hosts. The remaining components (revision-retrieved and simulator-tested) were then assigned to these established clusters. Six out of five simulator components were clustered together, suggesting that the network was able to identify similarities in loading history. However, the simulator-tested components ended up in a cluster at the fringe of the map containing only 10.8% of retrieved components. This may suggest that current ISO testing protocols were not fully representative of this TKR population, and protocols that better resemble patients' gait after TKR containing activities other than walking may be warranted. Fulbright15030157National Institutes of HealthR03 AR052039R01 AR059843 ==== Body 1. Introduction Wear performance evaluation has become an important preclinical tool for the assessment of materials and designs of total knee replacement (TKR) components. To date, the International Organization for Standardization (ISO) has established two wear testing protocols to evaluate the long-term wear performance of TKR components [1, 2]. Both ISO protocols aim at replicating load and motion characteristics of a natural knee during level walking, which is considered to be the most frequently performed physical activity of daily living [3]. As with any simulation tool, the ultimate goal of wear simulations is to recreate in vivo conditions as closely as possible. For knee wear simulation, this means recreating wear damage characteristics (wear rates, wear modes, wear patterns, damage appearances, particle sizes, and morphologies) that are similar to those generated in vivo. However, reproducing in vivo wear damage characteristics of the knee has proven to be very challenging because simulators generate tibial liner wear scars that are less variable in size and location compared to those observed in retrievals of the same design type [4, 5]. Several factors, such as the characteristics of the prosthesis (materials and designs), the patient (height, weight, joint loading during daily activities, and activity level), and the surgical technique (alignment and soft tissue balancing), influence the wear of a TKR polyethylene tibial liner. Discrepancies between simulated and in vivo worn components can be identified by comparing their wear scar characteristics, which are substantially influenced by the kinetics and kinematics of the knee joint. Hence, wear scars are useful indicators of the physiological load and motion spectrum applied to the tibial insert during daily physical activity. However, a detailed analysis of wear scars is very complex. The mathematical description of wear scar patterns is nonlinear and multidimensional, which makes it very difficult or even impossible to model these patterns using traditional mathematical or statistical methods. For instance, different geometric parameters, including area, perimeter, or centroid of a wear scar, could be used to form the basis for a specific model. However, even multiple geometric parameters may not sufficiently explain the overall wear scar generation process, which is why we propose to analyze in vivo and in vitro generated war scars as a whole using bitmap images. In this study, an artificial neural network (ANN) model based on image information is implemented as a data mining tool to differentiate wear scars that originate from different loading histories. ANNs have been successfully used for similar models because of their ability to handle nonlinear behavior, to learn from experimental data, and to generalize solutions [6–11]. From the pool of ANN models, the self-organizing feature map (SOFM) was selected for this study because it is an unsupervised neural network (i.e., no a priori knowledge of the data structure and classification is used). It is frequently used for the visualization of high dimensional data and for data mining and knowledge discovery [7–10, 12–14]. SOFMs are particularly useful because of their ability to map nonlinear statistical relationships between high dimensional data onto a convenient and easily comprehendible two-dimensional map. This type of mapping preserves the topology of the data, meaning that points within close proximity in the high dimensional space are mapped to neighboring map units in the output space. While this modeling technology has been used for image mapping since the early 2000s [15], to the best of our knowledge, it has not been used for applications in orthopedic tribology. The purpose of the present investigation was to create a clustering structure of wear scar images based on similarities between retrieved (revision and postmortem) and simulator-tested components of the same design type. Wear scars from the retrieved group were used to create a clustering structure, whereas the wear scars from simulator-tested components were then assigned to the existing clustering structure based on their similarities. Subsequently, data mining was performed to understand the similarities among wear scars clustered together as well as to explain the differences between wear scars of different clusters. Two hypotheses were tested: (1) wear scars from retrieved components will generate several clusters of wear scars because of the variability of wear scar size and location that characterizes retrieved components and (2) all simulator components, regardless of the testing standard used, will be clustered together, reflecting the comparability of the two ISO testing standards and their limitation in generalizing the greater variability observed in retrieved components of the same design type. 2. Materials and Methods An overview of the materials and methods used in this investigation is presented in Figure 1. With approval from the Institutional Review Board (#L03072801), twenty-one postmortem- and fifty-four revision-retrieved tibial liners were selected from the Retrieval Repository at Rush University Medical Center (Table 1). Before being included in the study, components were screened for missing demographic information and for signs of heavy delamination. All retrieved components were manufactured by a single company (Zimmer, Inc., Warsaw, IN, USA) and were of the posterior cruciate retaining MG-II design, a fixed bearing prosthesis with a flat tibial polyethylene plateau. Wear testing was performed using eight tibial liners, which were of the same design type and company as the retrieved components (MG-II, Zimmer, Inc., Warsaw, IN, USA). Testing components were randomized into two equal groups. In each group, three samples were tested for wear performance and one sample served as a loaded soak control. The tibial plateaus were machined from ultra-high molecular weight polyethylene (UHMWPE), gamma sterilized, and packaged in a nitrogen environment by the manufacturer. The boxes were opened immediately prior to testing. Wear performance tests were carried out in a four-station knee simulator (EndoLab, Rosenheim, Germany). The simulator met ISO standards [1, 2] and could be set up to run either in load control mode or in displacement control mode. The simulator motions were hydraulically actuated and closed-loop controlled. The difference in control mode refers to two degrees of freedom (anterior-posterior and internal-external, resp.) that were either load or displacement controlled, resulting in different implant articulations that were determined by the specific design aspects of the artificial joint. The wear tests were conducted prior to 2009 following the original ISO standards and have been published elsewhere [16]. Briefly, each simulator station was comprised of a temperature-controlled chamber that maintained the test lubricant at 37°C. The lubricant was based on a buffered mixture of bovine serum (Hyclone Inc., Logan, UT, USA) mixed with a physiological salt solution to achieve a final protein content of 30 g/L and a pH of 7.4. In order to sequester metal ions, 200 mg/L ethylenediaminetetraacetic acid (EDTA) was added. All chambers were closed and sealed during the entire test to minimize fluid evaporation and contamination. The simulator was connected to a computer equipped with a user interface for machine control, test supervision, and data acquisition. The first simulator group was tested in load control mode (LCM) and the second group was tested in displacement control mode (DCM). The LCM and DCM tests followed the same general protocol and testing parameters stated in the original 2002 and 2004 versions [1, 2]. Tests were conducted at 1.0 Hz cycle frequency and lasted for five million cycles (Mc). Load and displacement input represented one full walking cycle per test cycle and were taken from the respective ISO standards. The experiment was interrupted every 0.5 Mc to dismount, clean, and weigh the specimens according to the ISO standard [17]. Wear scars on the tibial UHMWPE plateaus that developed during the test were analyzed after test completion. Medial and lateral articulating surfaces were visually analyzed using a video-based microscope (SmartScope, OGP NY, USA). Wear scars were digitized by manually tracking their contours (i.e., the boundary between worn and unworn areas) on the liner surface (Figure 1(a)) [18]. Since the goal of this study was to compare wear scar patterns using images rather than discrete geometric parameters, black and white wear scar bitmap images (220 × 170 pixels) were generated for each component (Figure 1(b)). Each bitmap image contained medial and lateral wear scar shapes with black pixels representing worn areas and white pixels representing unworn areas. Each bitmap image was converted to a 220 × 170 matrix with “1” representing white pixels and “0” representing black pixels. Each matrix was then reshaped to a single row vector and used as input for the SOFM model. While the component border was not kept in the image, the length and height of the image were adjusted to match the component size. All components were normalized to an equal size and right implantation side. Components with unknown implantation side (~7%, Table 1) were normalized after side determination with ANN. Geometric wear scar parameters, including area, perimeter, centroid, bounding box, anterior/posterior stretch, medial/lateral stretch, moment of inertia, and multiple shape factors, were computed for each component (Figure 1(c)) and used for statistical analysis. The SOFM network was designed and trained using the Matlab SOM Toolbox 2.0 (Helsinki University of Technology, Finland). A sensitivity analysis was conducted to identify ideal training parameters generating best mapping results. The networks consisted of an input layer of 37,400 neurons (from image dimensions of 220 × 170 pixels = 37,400), a competitive layer, and an n × m neurons map or output layer (Figure 2). Five different networks with different map dimensions were generated. Map size and neighborhood radius were the only parameters tuned during the sensitivity analysis. The learning rate was linearly adjusted for all networks and the presentation of training samples was done in a random order. Training was performed using postmortem-retrieved components only. Subsequently, simulator- and revision-retrieved components were assigned to already existing clusters. No network learning occurred from the simulator wear scar patters. Training was done using the batch algorithm based on the Euclidean metric. Statistical analysis of the clustering structure was performed only from the map providing the smallest quantization error (which is a measure of “fit” between input and output mapping) and a well-defined cluster structure. The u-matrix method was used to visualize the distance of each map neuron to its neighbors. The shorter the distance between neurons, the smaller the difference between them [19, 20]. This method was used to visually uncover the clustering structure in the SOFM. Commonly, a two-dimensional color coded u-matrix is used to identify cluster boundaries. Component planes (another commonly used visualization tool) were not created because the type of input data used in this study would have produced 37,400 component planes (one for each dimension). Clustering robustness was evaluated by producing multiple versions of the map with the best mapping results. The goal of this process was to detect mapping irregularities caused by the inherent mapping error meaning that data from a high dimensional space mask a significantly smaller dimensional space. To detect clustering irregularities, three network versions were created and trained until they converged. The networks were created and analyzed by an independent investigator. The networks' map size, learning rate, and neighborhood radius were left unchanged. The only training parameters that differed between networks were the initial values of the map neurons and the presentation of the training samples, which were both randomly chosen. The clustering structure was visualized and compared between network versions. The map neurons assigned to each wear scar in each of the networks were recorded and used for comparison. Cohen's kappa analysis was carried out to investigate if each component was consistently clustered with the same group of components. SOFM mapping configurations were evaluated based on quantization errors. To test the interrater reliability of the network, intraclass correlation coefficients (ICC) were computed. An analysis of variance (ANOVA) was conducted to detect differences within and among clustered wear scar images. The geometric parameters, computed for medial and lateral wear scars separately, were used as output variables in the statistical analysis. The associations between two available input variables (“time in host” and “age at surgery”) with output variables were evaluated using regression analysis. Only clusters with available input on more than three retrieved components were included. The chance probability that five of six simulator components would land in a single cluster was estimated using the binomial distribution. The probability of “success” (i.e., landing in cluster “X”) was estimated from the proportion of revision and postmortem components that landed in that cluster. All statistical analyses were performed in SPSS 16.0 for Windows (SPSS Inc., Champaign, IL, USA). 3. Results A network with a map size of 12 × 10 and initial to final neighborhood radii of 4 to 1 was found to provide the lowest quantization error (q e = 11.14) and a well-defined clustering structure. The other network configurations evaluated were 20 × 10/4 to 1, 20 × 10/4 to 1, 10 × 10/4 to 1, 10 × 10/5 to 3.5, and 7 × 7/4 to 1. The 20 × 10 network had a lower quantization error (q e(20×10) = 10.9) than the network selected for the final analysis; however, its cluster boundaries were not easily identifiable. The remaining networks evaluated had higher quantization errors: q e(10×10/4 to 1) = 12.7; q e(10×10/5 to 3.5) = 15.3; and q e(7×7) = 17.1. The clustering robustness analysis showed substantial interrater reliability for the different SOFMs created with a kappa value of 0.69 (p < 0.001) and 95% CI (0.667, 0.712). Despite the random initial values of map neurons and the random presentation of the training samples, tibial inserts that were clustered together in the first round stayed mostly in the same cluster during the second round. On average 84% (SD ± 19%) of all components were consistently mapped with the same components. Using the u-matrix visualization method, eleven clusters became evident, each containing at least one postmortem-retrieved component and a maximum of 18 retrieved components (Figures 3 and 4). While 54 revision-retrieved components were assigned to nine of eleven clusters, all but one of the six simulator-tested components were placed in cluster 1. The chance probability that five or more of the simulator components would land in cluster 1 was estimated to be 1.6E − 4 using the binomial distribution. It is worth mentioning that cluster 1 contained only 10.8% of retrieved components and was one of the more isolated clusters at the fringe of the map. The geometric features of the wear scars are summarized in Table 2. There was no single geometric variable that could have explained the differences between all clusters. Thus, it was found that cluster 1 was not significantly different from the other clusters based on wear scar geometric parameters alone, although the SOFM network had established cluster 1 as one of the most dissimilar clusters. Interestingly, the largest number of significant differences was found in cluster 11. For simulator components only, medial and lateral wear scars were more anteriorly located and more symmetrical than for the retrieved components in the cluster. However, only the anterior location differed significantly from all other cluster-retrieved components (p < 0.05), whereas the wear scar symmetry did not. The associations between two available input variables (i.e., “time in host” and “age at surgery”) and geometric output variables differed between the various groups (Table 2). 4. Discussion In this study, the relationship between wear scar images of simulator-tested and retrieved TKR tibial components was investigated. A nontraditional qualitative modeling approach was used to project nonlinear relationships of a high dimensional data set (wear scar images) onto a two-dimensional map. The SOFM algorithm was used as a data mining and knowledge discovering tool and served as visual aid in the discovery of wear scar characteristics. After successfully training with wear scars from postmortem-retrieved components, eleven clusters were created. Purposefully, postmortem-retrieved inserts were used for this training purpose since they count as well-functioning at the time of retrieval and as such may be considered a “gold standard” for TKR wear simulation. As hypothesized, several clusters of wear scars were generated, mimicking the variability of wear scar patterns that characterizes retrieved components [4, 5]. Further, wear scars generated through mechanical simulation were clustered together, suggesting that the clustering process is meaningful in that wear scars of a similar loading history are recognized by the SOFM. It must be stressed that this cluster contained wear scars from both load and displacement control tested inserts, which showed distinct differences in wear scar size in an earlier study [17]. Hence, there must be other important wear scar features that render them similar. All but one of the simulator-tested components were clustered together. The simulator-tested component assigned to cluster 4 clearly differs visually from the other simulator components (see Figure 3). We were aware of this difference because one of the AP actuators of the simulator became faulty during one of the wear tests. However, this information was not used as input into the SOFM. The only data and information used as input into the network was the medial and lateral wear scar images from both retrieved and simulator-tested components, which were all presented to the network in a random order during the training process. Hence, it appears that the SOFM network is capable of identifying subtle differences in loading history. Based on the clustering results, the load and displacement control tested inserts account only for about 11% of the wear scar characteristics found in retrieved components. Cluster 1 is at the fringe of the cluster map and relatively isolated from other components (as indicated by the high ridge around it; see Figure 3). Ideally, the cluster containing the simulator components establishes itself in the center of the map to have shorter distances to all components and, thus, be more representative. The sole application of ISO gait cycles may not be sufficient in mimicking the greater variability of wear scar patterns observed on retrieved components. Ngai et al. reported that not only do the motion patterns of TKR patient differ from the motion pattern applied by the displacement [21] and/or load control [22] standard, but they are also highly variable between patients [23]. Also, these findings may indicate that it is important to consider other activities of daily living for knee wear testing. Both Benson et al. [24] and Cottrell et al. [25] found that the inclusion of one cycle of stair descent or ascent for every seventy cycles of level walking during wear testing produced more in vivo-like wear scars than those generated by walking alone. Thus, the variability of wear scars observed in retrieved components may not just be the result of different walking patterns but may reflect the range of physical activities performed by the patient, raising the need for a more representative TKR motion testing pattern. There are limitations to using the SOFM. The network does not identify variables that characterize each cluster and best discriminate between the clusters [7]. Hence, the user is left in ambiguity. In this study we were unable to explain wear scar clustering by geometric characteristics. Since the clustering created by the SOFM is a projection of a nonlinear and high dimensional input space, the clustering results may not be fully explained by traditional linear statistical models. Perhaps, future mathematical means may resolve this issue. Because of the nature of the clustered data of this study, the issue was amplified. Typically, cluster correlations created by a SOFM are performed using component planes; however, our data sets were based on pixel information and this analysis was not applicable. A second limitation was that the high dimensionality of the input data set affected the training time of the SOFM, ranging from four hours to almost a full day until convergence, depending on the map size. Smaller bitmap images or a different representation of the wear scar pattern may be used to limit the computational time spent on training the SOFM. Smaller bitmap images may also reduce the quantization error because this error depends directly on the dimensionality of the input space and the output map where a greater dimensionality reduction will result in a greater quantization error. On the other hand, a coarser, more pixilated wear scar may result in loss of sensitivity and a threshold has yet to be established. Finally, there were also limitations with the study design. The simulator tests were executed according to the original knee wear testing standards and should be repeated following the updated protocols. Our retrieval collection was small in size, with modest and partially incomplete patient information. This resulted in underrepresented clusters with few components and prevented a thorough data mining. Both “time in situ” and “patient age” are only auxiliary variables for prosthetic use and patient activity. Knowledge about the number of individual walking steps, the specific gait mechanics, and activity profile of each patient may have provided important clues in identifying associations and differences within and between clusters. 5. Conclusions In conclusion, an artificial neural network approach has been applied for the comparison of wear scar images of simulator and retrieved TKR tibial inserts. This modeling approach proved to be robust and repeatable. The model, which was based on the self-organizing feature map network, can be used to directly compare wear scars from simulator and retrieved tibial liners. The SOFM network analysis revealed that (1) wear scars from retrieved components are highly variable, generating multiple clusters, (2) wear scars generated through wear testing using two different ISO standards were clustered together and are, thus, deemed comparable, and (3) wear scars from simulator components were clustered away from the center of the map and, therefore, are not representative of the whole retrieval collection. In the future, we may check if a new multiactivity testing protocol is capable of generating wear scars that more closely resemble retrieved components. The SOFM model may also be used for data mining of very large retrieval cohorts and search for associations and differences beyond physical context. For example, the input could contain surgical factors and/or socioeconomic factors. In the summary, the SOFM established in this study provides a unique and versatile platform for future discovery analysis. Acknowledgments The authors are grateful to Mr. Spencer Fullam, who repeated the ANN analysis independently to test the repeatability of findings. The authors also want to thank Drs. Michel Laurent, Lou Fogg, and Marzieh Ardestani for their insightful discussions about content and statistics. Further, the authors acknowledge editorial support of Ms. Daniela Gollegger-Knafl. The study was financially supported by Fulbright student stipend (Diego A. Orozco Villaseñor), IIE no. 15030157, and two federal grants (Markus A. Wimmer) from the National Institutes of Health (NIH R03 AR052039 and R01 AR059843). Competing Interests The authors declare that they have no competing interests regarding the publication of this paper. Figure 1 Flow diagram providing the methodology applied in this investigation. The methodology was divided into three main sections: (1) data collection and preprocessing; (2) data processing; and (3) SOFM training. Figure 2 Self-organizing feature map (SOFM) neural network structure. In the competitive layer, input vectors are assigned to the neuron with the shortest Euclidean distance. Similar input vectors will be assigned to neighboring neurons. Figure 3 U-matrix visualization of the SOFM after training. Eleven wear pattern clusters were identified. Five out of six in vitro tested components were assigned to cluster “1”. ∗The number of revised (R), postmortem (P), and simulator (S) components and the total percentage (%) of components assigned to each group are noted in brackets [R, P, S, %]. Light map colors represent cluster areas (valleys), while darker colors represent cluster boundaries (hills). Figure 4 Eleven Clusters were established. Except for one, all simulator components fell in Cluster “1” together with six revision and three postmortem components. Table 1 Demographic information of liner hosts (revision and postmortem). Implant source (N) Gender (N) Side (N) In situ time (mo.) Cause of failure (N) Revisions (54) Females (22) Males (26) Unknown (6) Left (24) Right (23) Unknown (7) Range (1–108) Mean (26) Unknown (16) Infection (10) Maltracking (9) Loose (9) Instability (5) Synovitis (2) Fracture (1) Osteolysis (1) Failed liner (1) PE wear∗∗ (1) Unknown (15) Postmortem (21) Females (13) Males (8) Left (11) Right (10) Range (19–144) Mean (79) Autopsy (21) Simulator (6) Not applicable Left (6) 60 months∗ Not applicable Heavily delaminated (10) Females (5) Males (4) Unknown (1) Left (7) Right (3) Range (24–130) Mean (73) Unknown (3) Instability (2) Polyethylene wear (2) Tibial subsidence (2) Painful tibial component (1) Unknown (3) ∗1 million cycles representing 12 months of level walking. ∗∗PE = polyethylene. Table 2 Summary of geometric parameters for retrieved and simulator components. Bold values denote a significant (p < 0.05) and meaningful (R 2 > 0.4) association with input variables “time in host” and “age at surgery.” Mean (StDev) Medial Lateral Cluster number Area (mm2) Perimeter (mm) ML stretch (mm) AP stretch (mm) Area (mm2) Perimeter (mm) ML stretch (mm) AP stretch (mm) 1 391.84 (135.11) 79.18 (13.54) 23.59 (3.51) 21.87 (6.03) 460.96 (166.04) 84.25 (12.78) −12.77 (20.80) 25.64 (4.97) 2 498.21 (78.26) 83.56 (4.40) 24.39 (4.66) 26.15 (3.32) 566.35 (80.80) 89.47 (2.84) 12.25 (28.34) 27.06 (1.99) 3 712.27 (185.35) 100.02 (12.20) 26.87 (2.88) 32.92 (5.23) 754.24 (180.98) 102.73 (11.72) 0.93 (29.67) 33.26 (5.72) 4 416.07 (146.55) 78.78 (1.51) 23.12 (3.37) 23.68 (4.87) 421.97 (165.74) 79.01 (12.56) −3.67 (26.80) 22.86 (6.32) 5∗ 283.47 62.59 22.63 16.71 337.01 67.39 −20.51 21.38 6 418.87 (101.98) 77.46 (6.04) 21.78 (0.63) 24.22 (3.62) 412.37 (121.97) 76.86 (10.95) −23.54 (2.06) 22.74 (3.61) 7∗ 355.69 71.44 19.39 23.58 436.42 76.99 −26.19 21.71 8 179.84 (34.29) 54.08 (6.05) 17.21 (1.75) 14.83 (3.31) 143.83 (76.49) 46.72 (12.10) 3.17 (15.72) 13.87 (3.19) 9 374.52 (108.82) 75.28 (9.46) 23.33 (3.76) 21.05 (4.33) 392.90 (124.16) 74.99 (10.18) 3.62 (23.54) 21.97 (5.55) 10 363.73 (15.08) 73.69 (0.83) 22.83 (3.82) 21.21 (5.66) 308.54 (46.11) 68.10 (6.71) −7.98 (28.02) 16.16 (3.08) 11 129.36 (88.82) 45.12 (18.78) 13.76 (5.50) 13.53 (6.04) 241.70 (136.61) 58.06 (26.35) 8.61 (15.45) 16.62 (7.67) StDev = standard deviation, ML stretch = medial-lateral stretch, and AP stretch = anterior-posterior stretch. ∗StDev not available, n (cluster) = 1. ==== Refs 1 ISO Implants for surgery—wear of total knee-joint prostheses—part 1: loading and displacement parameters for wear-testing machines with load control and corresponding environmental conditions for test ISO 2009 142431:2002 2 ISO Implants for surgery—wear of total knee-joint prostheses—part 3: loading and displacement parameters for wear-testing machines with displacement control and corresponding environmental conditions for test ISO 2004 14243-3:2004 3 Seedhom B. B. Wallbridge N. C. Walking activities and wear of prostheses Annals of the Rheumatic Diseases 1985 44 12 838 843 10.1136/ard.44.12.838 2-s2.0-0022395089 4083940 4 Wimmer M. A. Paul P. Haman J. Differences in damage between revision and postmortem retrieved TKA implants ORS Transactions 2005 51 p. 1204 5 Harman M. K. Desjardins J. Benson L. Banks S. A. LaBerge M. Hodge W. A. Comparison of polyethylene tibial insert damage from in vivo function and in vitro wear simulation Journal of Orthopaedic Research 2009 27 4 540 548 10.1002/jor.20743 2-s2.0-63349103150 18932244 6 Bertis E. Mining pixels: the extraction and classification of astronomical sources Mining the Sky. ESO Astrophysics Symposia 2001 Berlin, Germany Springer 353 371 7 Castellani B. Castellani J. Data mining: qualitative analysis with health informatics data Qualitative Health Research 2003 13 7 1005 1018 10.1177/1049732303253523 2-s2.0-0142029070 14502965 8 Fornells A. Martorell J. M. Golobardes E. Garrell J. M. Vilasis X. Patterns out of cases using kohonen maps in breast cancer diagnosis International Journal of Neural Systems 2008 18 1 33 43 10.1142/s012906570800135x 2-s2.0-43949083024 18344221 9 Huang J. Shimizu H. Shioya S. Clustering gene expression pattern and extracting relationship in gene network based on artificial neural networks Journal of Bioscience and Bioengineering 2003 96 5 421 428 10.1016/s1389-1723(03)70126-1 2-s2.0-0348013431 16233550 10 Silver H. Shmoish M. Analysis of cognitive performance in schizophrenia patients and healthy individuals with unsupervised clustering models Psychiatry Research 2008 159 1-2 167 179 10.1016/j.psychres.2007.06.009 2-s2.0-42749089318 18440074 11 Yang Z. R. Chou K.-C. Mining biological data using self-organizing map Journal of Chemical Information and Computer Sciences 2003 43 6 1748 1753 10.1021/ci034138n 2-s2.0-0344687170 14632420 12 Glover C. J. Rabow A. A. Isgor Y. G. Shoemaker R. H. Covell D. G. Data mining of NCI's anticancer screening database reveals mitochondrial complex I inhibitors cytotoxic to leukemia cell lines Biochemical Pharmacology 2007 73 3 331 340 10.1016/j.bcp.2006.10.005 2-s2.0-33845605490 17109823 13 Yan S. Abidi S. S. Artes P. H. Analyzing sub-classifications of glaucoma via SOM based clustering of optic nerve images Studies in Health Technology and Informatics 2005 116 483 488 16160304 14 Kohonen T. Self-Organizing Maps 2001 3rd Berlin, Germany Springer 15 Endo M. Ueno M. Tanabe T. A clustering method using hierarchical self-organizing maps Journal of VLSI Signal Processing Systems for Signal, Image and Video Technology 2002 32 1-2 105 118 10.1023/a:1016371519687 2-s2.0-0036690072 16 Schwenke T. Orozco D. A. Schneider E. Wimmer M. A. Differences in wear between load and displacement control tested total knee replacements Wear 2009 267 5–8 757 762 10.1016/j.wear.2009.01.028 2-s2.0-65749094059 17 ISO Implants for surgery—wear of total knee-joint prostheses—part 2: methods of measurement ISO 2000 14243-2:2000 18 Knowlton C. B. Bhutani P. Wimmer M. A. Relationship of surface damage appearance and volumetric wear in retrieved TKR polyethylene liners Journal of Biomedical Materials Research Part B: Applied Biomaterials 2016 10.1002/jbm.b.33684 19 Vesanto J. Alhoniemi E. Clustering of the self-organizing map IEEE Transactions on Neural Networks 2000 11 3 586 600 10.1109/72.846731 2-s2.0-0034187784 18249787 20 Ultsch A. Siemon H. P. Kohonen's self organizing feature maps for exploratory data analysis Proceedings of the International Neural Network Conference 1990 305 308 21 Ngai V. Wimmer M. A. Kinematic evaluation of cruciate-retaining total knee replacement patients during level walking: a comparison with the displacement-controlled ISO standard Journal of Biomechanics 2009 42 14 2363 2368 10.1016/j.jbiomech.2009.06.030 2-s2.0-69949097082 19651410 22 Ngai V. Schwenke T. Wimmer M. A. In-vivo kinematics of knee prostheses patients during level walking compared with the ISO force-controlled simulator standard Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine 2009 223 7 889 896 10.1243/09544070JEIM549 2-s2.0-70349467986 23 Ngai V. Wimmer M. A. Variability of TKR knee kinematics and relationship with gait kinetics: implications for total knee wear BioMed Research International 2015 2015 6 284513 10.1155/2015/284513 2-s2.0-84926512741 24 Benson L. C. DesJardins J. D. Harman M. K. LaBerge M. Effect of stair descent loading on ultra-high molecular weight polyethylene wear in a force-controlled knee simulator Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine 2002 216 6 409 418 10.1243/095441102321032201 2-s2.0-0036050575 25 Cottrell J. M. Babalola O. Furman B. S. Wright T. M. Stair ascent kinematics affect UHMWPE wear and damage in total knee replacements Journal of Biomedical Materials Research—Part B Applied Biomaterials 2006 78 1 15 19 10.1002/jbm.b.30451 2-s2.0-33745713093

PMC005xxxxxx/PMC5002292.txt

==== Front Int J DentInt J DentIJDInternational Journal of Dentistry1687-87281687-8736Hindawi Publishing Corporation 10.1155/2016/8169356Clinical StudyComparison of Piezosurgery and Conventional Rotary Instruments for Removal of Impacted Mandibular Third Molars: A Randomized Controlled Clinical and Radiographic Trial http://orcid.org/0000-0002-7588-8468Arakji Hani 1 * Shokry Mohamed 2 3 Aboelsaad Nayer 2 4 1Oral Surgical Sciences Department, Division of Oral and Maxillofacial Surgery, Faculty of Dentistry, Beirut Arab University, Riad El Solh, P.O. Box 11-5020, Beirut 11072809, Lebanon2Oral Surgical Sciences Department, Faculty of Dentistry, Beirut Arab University, Riad El Solh, P.O. Box 11-5020, Beirut 11072809, Lebanon3Oral and Maxillofacial Surgery Department, Faculty of Dentistry, Alexandria University, Alexandria, Egypt4Oral Medicine and Periodontology Department, Faculty of Dentistry, Mansoura University, Mansoura 35516, Egypt*Hani Arakji: haniarakji@gmail.comAcademic Editor: Tommaso Lombardi 2016 14 8 2016 2016 81693568 3 2016 23 6 2016 11 7 2016 Copyright © 2016 Hani Arakji et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.The purpose of this study was to test the effect of the surgical removal of impacted mandibular third molars using piezosurgery versus the conventional surgical technique on postoperative sequelae and bone healing. Material and Methods. This study was carried out as a randomized controlled clinical trial: split mouth design. Twenty patients with bilateral mandibular third molar mesioangular impaction class II position B indicated for surgical extraction were treated randomly using either the piezosurgery or the conventional bur technique on each site. Duration of the procedure, postoperative edema, trismus, pain, healing, and bone density and quantity were evaluated up to 6 months postoperatively. Results. Test and control sites were compared using paired t-test. There was statistical significance in reduction of pain and swelling in test sites, where the time of the procedure was statistically increased in test site. For bone quantity and quality, statistical difference was found where test site showed better results. Conclusion. Piezosurgery technique improves quality of patient's life in form of decrease of postoperative pain, trismus, and swelling. Furthermore, it enhances bone quality within the extraction socket and bone quantity along the distal aspect of the mandibular second molar. ==== Body 1. Introduction Approximately 20% of the population has impacted teeth, where mandibular and maxillary third molars are the most common [1]. The highest incidence of impaction has been shown in mandibular wisdom teeth which may lead to pathologies like pericoronitis, periodontitis, second molars tooth-crown resorption, pain, cysts or odontogenic tumors, and primary or secondary crowding of the dentition. Early removal of these teeth to prevent the abovementioned problems is widely approved [2]. Surgical removal of these teeth is usually correlated with postoperative pain, swelling, and trismus whereas complications such as infection, dry socket, trigeminal nerve injuries, and, rarely, fracture of the mandible are less common to occur [1, 3]. Hard tissue cutting is a common procedure in the dental fields, especially during maxillofacial, oral, and periodontal surgeries. Traditionally, rotating instruments like burs have been used for osseous surgery. However, bone overheating and damage to adjacent tissues are disadvantages that are related to the use of these methods [4]. Piezosurgery is a novel technique that has been introduced as a valuable alternative to overcome the disadvantages associated with the conventional rotating bone cutting instruments [4]. It is performed by means of a device that uses microvibration at a frequency capable of cutting bone. Its mechanism of action is based on the ability of certain ceramics and crystals to deform when an electric current is passed across them, resulting in microvibration at ultrasonic frequency [5]. A frequency of 25–30 KHz, from a nitride-hardened or diamond-coated insert, allows for selective cut of bone tissue [6]. Since its approval for commercial use in 2002, it has been successfully utilized for many surgical procedures, such as maxillary sinus lifting, autologous bone graft harvesting, bone splitting, lateralization of the inferior alveolar nerve, and orthognathic and neurologic surgeries [7–9]. Goyal et al. (2012) [10] compared piezosurgery with the conventional rotary surgical technique and found that pain, swelling, trismus, and healing were significantly decreased in the piezosurgery site. Moreover, a systematic review and meta-analysis done by Jiang et al. 2015 [11] compared piezosurgery and conventional rotary osteotomy techniques in third molar extraction. They concluded that although the patients undergoing piezosurgery experienced longer surgery time, they developed less swelling, less pain, and less postoperative trismus. Also, similar findings were concluded in the meta-analysis conducted by Al-Moraissi et al. 2015 [12] in which there was a significant reduction in postoperative sequelae (facial swelling, pain, and trismus) with the piezoelectric surgical technique in third molar extraction, whereas their results showed that the duration of surgery and operating time for third molar extraction were significantly shorter with conventional rotary instruments compared to the piezoelectric surgical technique. Despite all the studies that had been conducted similar to this present work, they did not tackle the piezosurgery as a recent bone cutting method on bone quality and bone quantity on the resultant extraction site; therefore this study was performed in order to evaluate the bone healing within the extraction socket after removal of the impacted mandibular third molar using piezosurgery together with its effect on the postoperative sequelae: pain, trismus, and swelling. 2. Materials and Methods 2.1. Study Design and Setting This study was carried out as a randomized controlled clinical trial: split mouth design. The estimated sample size was calculated according to http://epitools.ausvet.com.au/, by taking the means of mouth opening from a previous similar study conducted by Goyal et al. [10] where mean for test site = 2.54 ± (0.93) and mean for control site = 3.91 ± (0.99), where the variance was calculated to be 0.92, assuming a confidence level of 95% and a study power of 80%. The calculated sample size was 16 male patients (32 operating sites). 20% was added to the sample size from the start of the study to eliminate the probability of drop-out through the treatment protocol. Therefore, twenty male patients (40 operating sites) who required removal of bilateral impacted lower third molars were conveniently recruited from the outpatient Department of Oral Surgical Sciences, Faculty of Dentistry, Beirut Arab University, Beirut, Lebanon. They have been selected to fulfill certain inclusion and exclusion criteria. The inclusion criteria were male patients with age ranging from 18 to 35 years having bilateral mandibular mesioangular impacted third molars (Pell and Gregory class II, position B) [13]. On the other hand, the exclusion criteria were heavy smokers (≥25 cigarettes) [14], uncontrolled systemic conditions, pathologies, and infection related to the site of surgery. Sites were randomly selected by tossing a coin, where the face was the test site where piezosurgery (Mectron Dental, Italy) was performed and the back was the control site where the conventional rotary instruments have been used. All work was conducted in accordance with the Declaration of Helsinki (1964), where all patients were informed about the whole procedure and signed a detailed consent form. The study had been started after the approval of the Institutional Review Board (IRB) of Beirut Arab University, code 2015H-025-D-M-0085. 2.2. Preoperative Phase A detailed medical and dental history and panoramic and intraoral periapical radiographs of the surgical site were taken. The patients were instructed to rinse their mouth by chlorhexidine mouthwash 0.12%, 30 minutes before the operation. 2.3. Surgical Phase All operations were done by the same surgeon under local anesthesia consisting of 2% lidocaine hydrochloride with 1 : 80,000 adrenaline (Lignospan Special, Septodont, UK). Both sites were prepared with 5% povidone-iodine solution, and a conventional extended buccal incision was made. A mucoperiosteal flap was reflected with a periosteal (Molt number 9) elevator to expose the impacted tooth and surrounding bone. For control site, a number 6 carbide round bur (DENTSPLY, USA) mounted on a straight handpiece was used at 35,000 rpm for guttering at the buccal and distal aspect of the tooth. A straight fissure bur was used to section the tooth. At all times cutting of bone and tooth was accompanied by copious irrigation with cooled saline solution. For test site, the OT7 and OT2 (Mectron Inserts, Italy) cutting inserts of the piezosurgery were used for bone guttering around the impacted tooth (Figure 1). The frequency was adjusted between 28 and 36 kHz and the microvibration amplitude between 30 and 60 μm/s. Sectioning of the teeth has been performed in the same manner as the control site. In both sites after removal of the tooth the extraction socket was debrided, irrigated with 0.9% normal saline, and closed with 3/0 black silk sutures. 2.4. Postsurgical Phase The duration of the operation was calculated in each case from the start of the incision till the termination of the suturing. Patients were instructed to apply cold fomentation for 10 min/hour for the first 6 postsurgical hours. Amoxicillin with clavulanic acid 625 mg (Augmentin by GlaxoSmithKline), 3 times daily for 5 days, and ibuprofen 400 mg (Bruffen Abbott) twice daily for 3 days were prescribed. Sutures were removed after 7 days. 2.5. Follow-Up Phase 2.5.1. Clinical Variables Pain, trismus, and swelling were evaluated on days 1, 7, and 14 postoperatively. Also healing of the flap and color of the overlying mucosa were checked. Postoperative pain was assessed using a visual analogue scale (VAS) [2]. Trismus was evaluated by measuring the interincisal distance between the incisal edge of the upper and lower central incisors using a caliper at maximum mouth opening (cm). Furthermore, postoperative swelling was measured using a tape by taking the mean of the distance between the lateral corner of the eye and the angle of the mandible, tragus and the outer corner of the mouth, and tragus and soft tissue pogonion [2]. 2.5.2. Radiographical Variables Standardized periapical X-rays using the XCP (RINN, DENTSPLY, USA) sensor holder were taken on the same day of surgery as a baseline, 3 months and 6 months postoperatively, in order to measure bone density using ImageJ software (image processing and analysis in Java V. 1.48). A standardized sized square (33×33 pixels) was inserted in the center of the extracted socket which is determined by identification of the intersected point between 2 straight lines: a horizontal line extending from the anterior border of the ramus to a midpoint on the distal aspect of the mandibular second molar and a vertical line extending from the alveolar bone crest to the roof of the inferior alveolar canal. The bone density within this square was measured by selecting Region of Interest (ROI), from tools, and then the given data was analyzed in terms of pixels (Figures 2(a)–2(c)). The same square was drawn for each X-ray for all patients where the bone density was measured [15]. CBCT (CS 9300, Carestream, USA) were taken immediately, 3 and 6 months postoperatively (Figures 3(a)–3(c)), for measurement of marginal bone height along the distal aspect of the mandibular second molar through drawing a straight line extending from the cementoenamel junction (CEJ) of the distal aspect of the mandibular second molar to the alveolar bone crest [16]. Test and control sites were compared regarding the study clinical and radiographic variables using paired t-test. Significance level was set at the 5% level. Statistical analysis was performed using SPSS version 20.0. 3. Results Twenty male patients who had bilateral impacted mandibular third molars extracted were included in the study. Their age ranged between 19 and 32 years. All 20 patients completed the 6-month follow-up period with no drop-out from the sample. The mean ± (SD) time of surgery was 17.60 ± (2.95) min in control site, whereas it was 28.50 ± (3.57) min in the test site. When comparing both sites regarding the operation time there was a statistical difference (P = 0.0001). All patients were thoroughly clinically evaluated starting from the first postoperative day till the 14th postoperative day. They showed an eventful soft tissue healing with absence of any signs of infection. Table 1 shows the comparison of pain sensation measured by VAS score between test and control at different follow-up periods. Significant differences existed in mean VAS scores after 1, 7, and 14 days (P ≤ 0.0001, <0.0001, and 0.001). After 1, 7, and 14 days, mean VAS scores in the test site were lower than that in the control site (mean in test = 3.60, 1.10, and 0.10 compared to mean in control = 6.70, 3.30, and 1.00). Moreover, Table 2 shows the comparison of trismus between test and control at different follow-up periods. Significant differences existed between mean measurements, indicating mouth opening at baseline and after 1, 7, and 14 days (P ≤ 0.0001, <0.0001, <0.0001, and 0.002). At baseline and after 1, 7, and 14 days, mouth opening in the control sites was less than at test sites (mean in control = 4.50, 2.74, 3.49, and 4.49 compared to mean in test = 4.78, 3.85, 4.53, and 4.77). Also, Table 3 shows the comparison of swelling between test and control at different follow-up periods. Significant differences existed between mean measurements indicating swelling at 1, 7, and 14 days postoperatively (P ≤ 0.0001, <0.0001, and 0.03). At 1, 7, and 14 days postoperatively, swelling was greater at control sites than at test sites (mean in control = 12.32, 11.78, and 11.30 compared to mean in test = 11.55, 11.29, and 11.20). Radiographically, the mean bone density in test site immediately and after 3 months and 6 months postoperatively, respectively, was 55.70 ± (3.60), 69.80 ± (8.19), and 84.45 ± (4.73), where the control recorded 54.00 ± (3.87), 62.75 ± (5.19), and 74.87 ± (4.03) in the same interval period. The results showed statistical difference between the two sites where piezosurgery site showed improved bone quality (P ≤ 0.0001) (Figure 4). On the other hand, bone loss has been observed along the distal aspect of the second molar within the two sites; greater amount of bone loss was statistically noticed in the control site when it was compared to the test site (P ≤ 0.0001). Immediately after the operation and after 3 and 6 months, mean bone loss in the control site was greater than that in the test site (mean in control = 5.30, 4.41, and 4.03 compared to mean in test = 4.01, 3.23, and 2.91) (Figure 5). 4. Discussion Recently, after painstaking research and the application of advanced principles of physics, newer instruments have been introduced to reduce the difficulty and morbidity in third molar surgery. One such innovation is piezosurgery or the application of piezoelectric, ultrasonic vibrations to make precise and safe osteotomies [17]. This study was carried out as an experimental, randomized, controlled clinical trial: split mouth design; this type of study is especially selected as it has the distinct advantage of removing the patients compliance bias from the estimated treatment effect as described by Zhu et al. [18]. To standardize our results, it was conducted on twenty male patients having their age ranging from 19 to 32 years, in order to remove the gender factor that may play a role in postoperative complications due to hormonal changes in females. They had mesioangular class II position B bony impacted mandibular third molar, according to Pell and Gregory [13]. This type of impaction was selected as it is the most commonly found and it was in agreement with a study conducted by Goyal et al. [10] and Piersanti et al. [17] where they chose the same impacted mandibular third molar category in their study. There was no drop-out from the selected sample and this may be attributed to the well-educated level of the selected patients and their commitment to their treatment in addition to the availability of the social media which makes the follow-up communication with the patients easier. The duration of the procedure in each site was calculated in terms of minutes starting from the establishment of the flap till the end of suturing. The mean duration of the operation was longer in the piezosurgery site than in the control site. This is in agreement with a similar study performed by Goyal et al. [10]. The mean recorded pain score was significantly lesser in the study site than in the control site. This finding is parallel to the results obtained by Goyal et al. [10], Mantovani et al. [19], and Piersanti et al. [17]. They reported in their studies a significant difference in pain score using the same scale, and all agreed that the site where the impacted mandibular third molar resides using piezosurgery has less postoperative pain. Furthermore swelling was evaluated in this study. Better improvement was noticed within the test site and this is in accordance with studies done by Pappalardo and Guarnieri [20], Mantovani et al. [19], Piersanti et al. [17], and Mozatti et al. [21] where they compared the postoperative outcomes between piezosurgery and conventional rotary surgery in removing mandibular third molars. These results run along the same line of findings of a meta-analysis study conducted by Jiang et al. [11] where 7 studies were included in their analysis. The aim of their study was to compare piezosurgery with rotary osteotomy techniques, with regard to surgery time and the severity of postoperative sequelae, including pain, swelling, and trismus. Their meta-analysis indicates that although patients undergoing piezosurgery experienced longer surgery time, they had less postoperative swelling, indicating that piezosurgery is a promising alternative technique for extraction of impacted third molars. Radiographically, bone density was assessed by the aid of standardized periapical radiographs. While the literature supports the usefulness of CBCT scans for the determination of radiographic bone density [22–25], there are other studies stating that the grey levels in CBCT scans are not accurate when compared with CT. In a 2006 presentation, Armstrong [26] concluded that “Hounsfield units sampled from identical anatomic areas with CBCT and Medical CT (MDCT) are not identical.” A study carried out by Katsumata et al. [27] found that the grey levels in a CBCT image for bone varied from 1500 to over 3000. They concluded that “the ability to assess the density or quality of bone is limited and because the grey level range is so variable the derived density provided less than meaningful data.” Due to all the previous factors, we selected the standardized digital radiography for assessing the bone density in an accurate manner. Standardization of the periapical X-rays was done through silicone based bite block for each patient for purpose of repeatability of the position of the sensor; also the angulation of the cone was standardized using the RINN XCP paralleling device. It was showed that there is a greater increase in bone density occurring within the test site from immediately postoperative period to 6 months after surgery. These results are in accordance with Vercelotti et al. [28] where they compared piezosurgery with carbide burs in ostectomy and osteoplasty and proved that there is better bone healing in terms of quantity and quality when using piezosurgery in osseous surgeries. Moreover, Rullo et al. [29] analyzed the bone histology and found well-defined histological differences between the bone samples collected with the bur and the ultrasonic device. They reported that more integrity of the bony structure, well-designed osteotomy lines, and no evidence of bone heat osteonecrosis characterized the bone samples harvested with the piezoelectric device. The alveolar bone loss was assessed radiographically using CBCT. There was a greater loss in bone height along the distal aspect of the mandibular second molar in the control site than in test site from baseline to 6 months after surgery. This difference was statistically significant. Rahnama et al. [30] stated that the ultrasound vibration stimulates cells' metabolism. Moreover, the lack of necrosis in the cut area accelerates bone regeneration. Soft tissue damage is not noticed. Furthermore, Labanca et al. [31] have made a review on piezosurgery and found that it has less damage to osteocytes and this can explain the decreased bone loss within the test site compared to the control one. Taking into consideration the aforementioned observations and despite the presence of controversies about the effect of surgical removal of impacted mandibular third molar on periodontal health distal to the adjacent second molar [32], the current study has shown a decreased bone loss along the distal aspect of the mandibular second molar when using piezosurgery, which enhances the periodontal health condition along the distal aspect of the second molar reducing the need for performing further periodontal procedures. Similarly, Tsai et al. [33] stated that piezoelectric surgical instruments might promote faster wound healing compared to rotary instruments over a short-term observation period. The main disadvantage of piezosurgery noticed so far besides expense and the risk of breakage of the surgical tips is the increased operating time as a result of the slow rate of cutting. The time of surgery can be improved by the operator's experience. Increasing the sample size with longer duration of follow-up and taking bone specimen for histological examination from the surgical site can add valuable findings to the previous results. 5. Conclusion Within the limitation of this study, it can be concluded that piezosurgery reduces postoperative pain, trismus, and swelling and enhances the postsurgical quality of patient's life. Also, it may play an important role in increasing bone density within the extraction socket and decreasing the amount of bone loss along the distal aspect of the mandibular second molar. Competing Interests The authors declare that there are no competing interests regarding the publication of this paper. Figure 1 (a) Piezosurgery device and (b) bone guttering around impacted mandibular third molar using piezosurgery (test site). Figure 2 (a) Immediate, (b) 3-month, and (c) 6-month standardized periapical X-ray measuring bone density using ImageJ software (test site). Figure 3 (a) Immediate, (b) 3-month, and (c) 6-month CBCT showing bone height measurement along the distal aspect of the second molar (test site). Figure 4 Mean and standard deviation for bone density (pixels). ∗: statistically significant at P ≤ 0.05. Figure 5 Mean and standard deviation for bone loss (millimeters). ∗: statistically significant at P ≤ 0.05. Table 1 Comparison of pain sensations between study and control at different follow-up periods.   Mean (SD) P value of paired t-test Study Control After 1 day 3.60 (1.71) 6.70 (0.95) <0.0001∗ After 7 days 1.10 (0.74) 3.30 (0.95) <0.0001∗ After 14 days 0.10 (0.32) 1.00 (0.67) 0.001∗ P value of paired t-test <0.0001∗ <0.0001∗   ∗Statistically significant at P ≤ 0.05. Table 2 Comparison of trismus (mouth opening) between test and control at different follow-up periods (in millimeters).   Mean (SD) P value of paired t-test Test Control Baseline 4.78 (0.14) 4.50 (0.11) <0.0001∗ After 1 day 3.85 (0.07) 2.74 (0.13) <0.0001∗ After 7 days 4.53 (0.08) 3.49 (0.09) <0.0001∗ After 14 days 4.77 (0.22) 4.49 (0.11) 0.002∗ P value of paired t-test 0.61 0.81   ∗Statistically significant at P ≤ 0.05. Table 3 Comparison of swellings between test and control at different follow-up periods (in millimeters).   Mean (SD) P value of paired t-test Test Control Baseline 11.21 (0.07) 11.27 (0.05) 0.001∗ After 1 day 11.55 (0.08) 12.32 (0.04) <0.0001∗ After 7 days 11.29 (0.08) 11.78 (0.12) <0.0001∗ After 14 days 11.20 (0.04) 11.30 (0.15) 0.03∗ P value of paired t-test 0.43 0.46   ∗Statistically significant at P ≤ 0.05. ==== Refs 1 Juodzbalys G. Daugela P. Mandibular third molar impaction: review of literature and a proposal of a classification Journal of Oral & Maxillofacial Research 2013 4 2, article e1 10.5037/jomr.2013.4201 2 Mansuri S. Mujeeb A. Hussain S. A. Hussain M. A. Z. Mandibular third molar impactions in male adults: relationship of Operative time and Types of impaction on inflammatory complications Journal of International Oral Health 2014 6 2 9 15 24876696 3 de Santana-Santos T. de Souza-Santos J.-A. Martins-Filho P.-R. da Silva L.-C. de Oliveira e Silva E.-D. Gomes A.-C. Prediction of postoperative facial swelling, pain and trismus following third molar surgery based on preoperative variables Medicina Oral, Patología Oral y Cirugía Bucal 2013 18 1 e65 e70 10.4317/medoral.1803 2-s2.0-84871959702 4 Bui C. H. Seldin E. B. Dodson T. B. Types, frequencies, and risk factors for complications after third molar extraction Journal of Oral and Maxillofacial Surgery 2003 61 12 1379 1389 10.1016/j.joms.2003.04.001 2-s2.0-0345714875 14663801 5 Leclercq P. Zenati C. Dohan D. M. Ultrasonic bone cut part 2: state-of-the-art specific clinical applications Journal of Oral and Maxillofacial Surgery 2008 66 1 183 188 10.1016/j.joms.2005.12.050 2-s2.0-37049034904 18083438 6 Barone A. Marconcini S. Giacomelli L. Rispoli L. Calvo J. L. Covani U. A randomized clinical evaluation of ultrasound bone surgery versus traditional rotary instruments in lower third molar extraction Journal of Oral and Maxillofacial Surgery 2010 68 2 330 336 10.1016/j.joms.2009.03.053 2-s2.0-74249101998 20116704 7 Happe A. Use of a piezoelectric surgical device to harvest bone grafts from the mandibular ramus: report of 40 cases International Journal of Periodontics and Restorative Dentistry 2007 27 3 241 249 2-s2.0-34250161634 17694947 8 Danza M. Guidi R. Carinci F. Comparison between implants inserted into piezo split and unsplit alveolar crests Journal of Oral and Maxillofacial Surgery 2009 67 11 2460 2465 10.1016/j.joms.2009.04.041 2-s2.0-70349859628 19837318 9 Landes C. A. Stübinger S. Rieger J. Williger B. Ha T. K. L. Sader R. Critical evaluation of piezoelectric osteotomy in orthognathic surgery: operative technique, blood loss, time requirement, nerve and vessel integrity Journal of Oral and Maxillofacial Surgery 2008 66 4 657 674 10.1016/j.joms.2007.06.633 2-s2.0-40749115567 18355589 10 Goyal M. Marya K. Jhamb A. Comparative evaluation of surgical outcome after removal of impacted mandibular third molars using a Piezotome or a conventional handpiece: A Prospective Study British Journal of Oral and Maxillofacial Surgery 2012 50 6 556 561 10.1016/j.bjoms.2011.10.010 2-s2.0-84864608082 22088359 11 Jiang Q. Qiu Y. Yang C. Yang J. Chen M. Zhang Z. Piezoelectric versus conventional rotary techniques for impacted third molar extraction: a meta-analysis of randomized controlled trials Medicine 2015 94 41 e1685 10.1097/md.0000000000001685 2-s2.0-84944457724 12 Al-Moraissi E. A. Elmansi Y. A. Al-Sharaee Y. A. Alrmali A. E. Alkhutari A. S. Does the piezoelectric surgical technique produce fewer postoperative sequelae after lower third molar surgery than conventional rotary instruments? A systematic review and meta analysis International Journal of Oral and Maxillofacial Surgery 2016 45 3 383 391 10.1016/j.ijom.2015.10.005 2-s2.0-84957843794 26572830 13 Pell G. Gregory G. Impacted mandibular third molar third molars: classification and modified technique for removal Dental Digest 1933 39 9 330 338 14 CDC Decrease in smoking prevalence—Minnesota, 1999-2010 Morbidity and Mortality Weekly Report 2011 60 5 138 141 21307824 15 Hassan K. S. Marei H. F. Alagl A. S. Does grafting of third molar extraction sockets enhance periodontal measures in 30- to 35-year-old patients? Journal of Oral and Maxillofacial Surgery 2012 70 4 757 764 10.1016/j.joms.2011.09.010 2-s2.0-84859111229 22177808 16 Misch K. A. Yi E. S. Sarment D. P. Accuracy of cone beam computed tomography for periodontal defect measurements Journal of Periodontology 2006 77 7 1261 1266 10.1902/jop.2006.050367 2-s2.0-33748518930 16805691 17 Piersanti L. Dilorenzo M. Monaco G. Marchetti C. Piezosurgery or conventional rotatory instruments for inferior third molar extractions? Journal of Oral and Maxillofacial Surgery 2014 72 9 1647 1652 10.1016/j.joms.2014.04.032 2-s2.0-84906245768 25109581 18 Zhu H. Zhang S. Ahn C. Sample size considerations for split-mouth design Statistical Methods in Medical Research 2015 24 4 399 487 10.1177/0962280214520734 26206561 19 Mantovani E. Arduino P. G. Schierano G. A split-mouth randomized clinical trial to evaluate the performance of piezosurgery compared with traditional technique in lower wisdom tooth removal Journal of Oral and Maxillofacial Surgery 2014 72 10 1890 1897 10.1016/j.joms.2014.05.002 2-s2.0-84909619829 25234524 20 Pappalardo S. Guarnieri R. Randomized clinical study comparing piezosurgery and conventional rotatory surgery in mandibular cyst enucleation Journal of Cranio-Maxillofacial Surgery 2014 42 5 e80 e85 10.1016/j.jcms.2013.06.013 2-s2.0-84902126986 23932541 21 Mozzati M. Gallesio G. Russo A. Staiti G. Mortellaro C. Third-molar extraction with ultrasound bone surgery: A Case-Control Study Journal of Craniofacial Surgery 2014 25 3 856 859 10.1097/scs.0000000000000825 2-s2.0-84900854301 24799108 22 Chun Y. S. Lim W. H. Bone density at interradicular sites: Implications for orthodontic mini-implant placement Orthodontics and Craniofacial Research 2009 12 1 25 32 10.1111/j.1601-6343.2008.01434.x 2-s2.0-64749105981 19154272 23 Noujeim M. Prihoda T. J. Langlais R. Nummikoski P. Evaluation of high-resolution cone beam computed tomography in the detection of simulated interradicular bone lesions Dentomaxillofacial Radiology 2009 38 3 156 162 10.1259/dmfr/61676894 2-s2.0-62649163779 19225086 24 Miracle A. C. Mukherji S. K. Conebeam CT of the head and neck, part 2: clinical applications American Journal of Neuroradiology 2009 30 7 1285 1292 10.3174/ajnr.a1654 2-s2.0-68549104047 19461061 25 Scarfe W. C. Farman A. G. Sukovic P. Clinical applications of cone-beam computed tomography in dental practice Journal of the Canadian Dental Association 2006 72 1 75 80 2-s2.0-33644639722 16480609 26 Armstrong R. T. Acceptability of cone beam CT vs. multi-detector CT for 3D anatomic model construction Journal of Oral and Maxillofacial Surgery 2006 64 9, supplement p. 37 10.1016/j.joms.2006.06.086 27 Katsumata A. Hirukawa A. Okumura S. Relationship between density variability and imaging volume size in cone-beam computerized tomographic scanning of the maxillofacial region: an in vitro study Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology and Endodontology 2009 107 3 420 425 10.1016/j.tripleo.2008.05.049 2-s2.0-59649097468 28 Vercellotti T. Nevins M. L. Kim D. M. Osseous response following resective therapy with piezosurgery International Journal of Periodontics and Restorative Dentistry 2005 25 6 543 549 2-s2.0-27944475997 16353529 29 Rullo R. Addabbo F. Papaccio G. D'Aquino R. Festa V. M. Piezoelectric device vs. conventional rotative instruments in impacted third molar surgery: relationships between surgical difficulty and postoperative pain with histological evaluations Journal of Cranio-Maxillofacial Surgery 2013 41 2 e33 e38 10.1016/j.jcms.2012.07.007 2-s2.0-84872678066 22890087 30 Rahnama M. Czupkałło Ł. Czajkowski L. Grasza J. Wallner J. The use of piezosurgery as an alternative method of minimally invasive surgery in the authors' experience Videosurgery and Other Miniinvasive Techniques 2013 8 4 321 326 10.5114/wiitm.2011.35144 2-s2.0-84892527599 24501602 31 Labanca M. Azzola F. Vinci R. Rodella L. F. Piezoelectric surgery: twenty years of use British Journal of Oral and Maxillofacial Surgery 2008 46 4 265 269 10.1016/j.bjoms.2007.12.007 2-s2.0-43049108429 18342999 32 Kaveri G. S. Prakash S. Third molars: a threat to periodontal health?? Journal of Maxillofacial & Oral Surgery 2012 11 2 220 223 10.1007/s12663-011-0286-x 23730073 33 Tsai S.-J. Chen Y.-L. Chang H.-H. Shyu Y.-C. Lin C.-P. Effect of piezoelectric instruments on healing propensity of alveolar sockets following mandibular third molar extraction Journal of Dental Sciences 2012 7 3 296 300 10.1016/j.jds.2012.07.001 2-s2.0-84869204297

PMC005xxxxxx/PMC5002293.txt

==== Front Case Rep HematolCase Rep HematolCRIHEMCase Reports in Hematology2090-65602090-6579Hindawi Publishing Corporation 10.1155/2016/4342820Case ReportSimultaneous Manifestation of Chronic Myelomonocytic Leukemia and Multiple Myeloma during Treatment by Prednisolone and Eltrombopag for Immune-Mediated Thrombocytopenic Purpura http://orcid.org/0000-0002-7143-4813Hagihara Masao * Inoue Morihiro Kodama Kenichiro Uchida Tomoyuki Hua Jian Department of Hematology, Eiju General Hospital, 2-23-16 Higashi-Ueno, Taito-ku, Tokyo 110-8645, Japan*Masao Hagihara: hagihara@eijuhp.comAcademic Editor: Kazunori Nakase 2016 14 8 2016 2016 434282030 3 2016 9 6 2016 10 7 2016 Copyright © 2016 Masao Hagihara et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.An 80-year-old man was admitted to our hospital because of severe thrombocytopenia. He was diagnosed with idiopathic thrombocytopenia, and prednisolone together with eltrombopag was started, leading to significant improvement of platelet counts. Four years later, there was a prominent increase of peripheral blood monocytes, which was accompanied by recurrence of thrombocytopenia. Bone marrow aspirates and serum electrophoresis revealed coexistence of chronic myelomonocytic leukemia (CMML) and multiple myeloma (MM). The patient received lenalidomide plus dexamethasone therapy but died due to exacerbation of the disorder. It was supposed that thrombocytopenia was secondarily caused by CMML and MM developed at a later period. ==== Body 1. Introduction Immune-mediated thrombocytopenic purpura (ITP) is a disorder which occurs spontaneously as idiopathic thrombocytopenic purpura or develops secondarily from various disorders, such as infection or malignancies through mechanisms of immune alteration [1]. Myelodysplastic syndrome (MDS) is one such underlying disorder that manifests isolated thrombocytopenia, although the incidence is quite low [2]. Chronic myelomonocytic leukemia (CMML) overlaps pathophysiologically with MDS and is categorized as MDS/MPN (myeloproliferative neoplasm) in the WHO 2008 classification. Patients with multiple myeloma (MM) or monoclonal gammopathy of undetermined significance (MGUS) have a significantly high risk of developing MDS, CMML, or acute myeloid leukemia (AML) compared with healthy controls. Furthermore, there have been several case reports that demonstrated the simultaneous occurrence of MM/MGUS and myeloid malignancy. In contrast, there has been no report of MDS or CMML preceding the development of MM. In the present case, it was concluded that CMML likely existed as an underlying disorder for the development of ITP, which had been in remission under successful treatment with eltrombopag, a recombinant thrombopoietin receptor (TPO-R). Four years later, CMML exacerbated simultaneously with the appearance of MM. 2. Case Report An 80-year-old man was referred to our hospital because of severe thrombocytopenia (Table 1). Bone marrow (BM) study showed an increased number of megakaryocytes. Neither significant increase of monocytes nor plasma cells were observed. As dysplastic change was not evident, he was diagnosed with ITP. At admission, prednisolone was started at 0.5 mg/kg, which did not result in significant improvement of platelet counts (1.7 × 104/mm3 to 2.3 × 104/mm3) for 1 month, and eltrombopag was added at 12.5 mg/day as a starting dose and then gradually increased to 37.5 mg/day during 4 months. Platelet counts continued to be stable (at around 10 × 104/mm3) through this treatment (Table 2). Four years later, there was an increase of peripheral blood monocytes, which was then accompanied by decreased platelet counts and severe anemia (Table 3). At this period, bone marrow examination was again performed and revealed a significant increase in monocytes as well as plasma cells (Figure 1), as determined based on the κ chain expression on CD38 gating by flow cytometry (Figure 2). Peripheral blood (PB) monocytes exceeded 1000/mm3 and IgG-κ typed M-protein was detected by an analysis of serum electrophoresis. Therefore, it was diagnosed that CMML and MM were simultaneously present. As anemia together with deterioration of renal insufficiency was considered to be MM-related symptoms, treatment by lenalidomide plus dexamethasone (starting doses of 25 mg/day and 20 mg/week, resp.) was started on day 1 of a second admission. On day 7, although both PB-monocytes and serum IgG had significantly decreased, the treatment was discontinued because of disseminated intravascular coagulation, possibly due to tumor lysis syndrome. On day 35, the patient was complicated with septic shock. Thereafter, his general condition rapidly worsened and he eventually died due to deterioration of CMML. 3. Discussion It has been recognized that MDS or AML (MDS/AML) can develop in patients with MM [3]. The use of anticancer drugs, especially melphalan, has been considered as the main factor for such elevated risk [4, 5]. Additionally, MM patients have a 2.4-fold to 8.01-fold increased risk of MDS/AML even without such cytotoxic treatment [6, 7]. On the contrary, simultaneous occurrence of MDS/AML and MM is quite a rare event. In the past, 27 cases of such coexistence of plasma cell dysplasia and myeloid malignancy, including 16 cases with MDS, have been reported [8–20]. Only 2 cases were diagnosed as CMML, while 6 out of 8 AML cases were diagnosed as myelomonocytoid type (AML-M4 or AML-M5 in FAB classification). It has been demonstrated that myeloma cells express multiple hematopoietic markers, including myelomonocytic antigens [21], and occasionally express biphenotypic features not only of myelomonocytic but also of erythroid, megakaryocytic, and T-lymphoid lineages [22]. It was also reported that IgG was produced in supernatants from CD14 positive monocytes in such a CMML/MM case. In addition, a rearranged band of IgH gene from CD14+ peripheral monocytes was identical to that from bone marrow containing plasma cells [8]. According to these findings, it has been speculated that the neoplastic transformation deriving from a common progenitor results in the occurrence of such hybrid hematological disorders. In the present case, myeloma cells expressed CD19 antigen, which generally remains in polyclonal but is absent in monoclonal plasma cells and lacked CD138 antigen, which is commonly expressed in myeloma cells [23]. In addition, CD33 antigen, which is expressed in myelomonocytoid but not in plasma cells, was significantly expressed. From these results, it was highly likely that CMML and myeloma cells in the present case originated from these immature, poorly differentiated progenitors or such atypical antigen expression might simply suggest the malignant, but not reactive (benign), properties of the increasing plasma cells, irrespective of the common progenitors. At disease onset, neither proliferation of plasma cells nor M-peak by analysis of serum protein fractionation was observed. In contrast, the number of PB-monocytes gradually increased and finally exceeded 1000/mm3 during the 4-year period of prednisolone plus eltrombopag treatment. Therefore, it was concluded that CMML existed as an underlying disorder when the patient was firstly diagnosed with ITP, and the present case is the first report of myeloid malignancy proceeding the occurrence of MM. It has been widely recognized that MDS with CMML is complicated with various types of autoimmune disorders, such as Sweet's disease [2, 24]. It has also been reported that immune-mediated thrombocytopenia can occur in CMML [2, 24]. Hadjadj et al. reported that ITP occurred in the early stage of CMML with no further progression of the disorder. In their report, a favorable response to treatment, including corticosteroid or TPO-R agonist, was obtained. Ineffective platelet production secondary to disordered maturation or proliferation of megakaryocytes has been considered as the other main mechanism of thrombocytopenia in MDS [25]. Eltrombopag has been shown to be useful to treat thrombocytopenia in MDS patients [26]. Platelet counts were improved or remained stable in 9 out of 12 cases, despite azacitidine treatment [27]. A meta-analysis study also demonstrated that romiplostim, another TPO-R agonist, significantly decreases the bleeding or platelet transfusion rate [28]. In the present case, eltrombopag was shown to be effective in maintaining platelet counts for as long as 4 years before diagnosis of CMML/MM. The safety of long-term treatment by TPO-R agonist has recently been reported and only 2 patients suffered from lymphoid malignancies of diffuse large B-cell and Hodgkin lymphoma [29]. The relative risk of AML progression with romiplostin was 1.36; however, this result was judged as having a high risk of bias [28]. In addition, eltrombopag inhibits proliferation of leukemic cells through a TPO-R independent mechanism in vitro [27, 30, 31]. From these reports, it was concluded that eltrombopag was not involved in either the progression of CMML or the development of MM. Lenalidomide, an immunomodulatory drug, is widely known as one of the key drugs for initial treatment of MM and also has been approved for low or low-intermediate risk MDS with del(5q) [32, 33]. The drug is also effective, even if to a lesser extent in non-del(5q) MDS. In an MDS-002 trial, the drug demonstrated transfusion independence in one quarter of MDS patients not harboring del(5q) [34]. It also shows a clinical response in CMML patients (25% of overall response), when combined with metronomic melphalan through an antiangiogenic effect [35]. We selected this drug to treat both MM and CMML. However, only a temporal improvement of these disorders was obtained, and eventually, the treatment was discontinued because of decreased performance status due to complications of the severely infectious disease. Other effective treatments with azacitidine, an analog of pyrimidine nucleoside, have been used in various myeloid malignancies, including CMML [36]. Clinical trials with combination therapy of lenalidomide and azacitidine have been reported. In one study, an overall response rate of 72%, including 44% with complete response, was obtained [37]. Such combination therapy might be a promising option as long as the general conditions of the patient were favorable enough to continue the chemotherapy. Competing Interests Authors have no conflict of interests. Figure 1 Bone marrow aspirate on second admission (Giemsa staining). A significant increase of monocytes (arrowhead) as well as plasma cells (arrow) was observed in bone marrow aspirates ((a), ×400 in magnification). Myelodysplastic change was also detected, showing as hypogranulated neutrocytes with Pseudo-Pelger nuclei ((b), ×1000 in magnification) or micromegakaryocytes ((c), ×1000 in magnification). Figure 2 Flow cytometry of bone marrow. CD38 positive cells were positively stained with CD19, CD33, and κ antigens and negatively with CD138, CD20, and λ antigens. Table 1 Laboratory data on first admission. WBC 3800/mm3  Band + seg 15.5%  Baso 0.5%  Eosino 0%  Mono 28%  Lymph 26% RBC 439 × 106/mm3 Hb 13.3 g/dL MCV 92.7 MCH 30.3 Platelets 3.6 × 104/mm3 TP 7.1 g/dL BUN 13.5 mg/dL Cr 0.89 mg/dL Na 140 mEq/L K 4.7 mEq/L AST 17 U/L ALT 12 U/L LDH 114 U/L IgG 1729 mg/dL IgA 404 mg/dL IgM 46 mg/dL PAIgG 118 ng/107 cells PAIgG: platelet-agglutinated IgG. Table 2 The number of PB-monocytes and platelets during 4 years.   X-4 years X-3 years X-2 years X-1 years X-8 months X years WBC (/mm3) 3800 7900 7500 6600 11000 17500 Monocyte (%) 28 24.5 27.5 26.7 32 51.5 Monocytes number (/mm3) 1064 1935 2025 1780 3550 8925 Platelet (×104/mm3) 3.6 5.3 5.6 6.9 9.2 4.5 Table 3 Laboratory data on second admission. WBC 17500/mm3  Blast 0.5%  Myelo 5.5%  Band 4.0%  Seg 32.5%  Mono 51.5%  Lymph 4.5% RBC 252 × 106/mm3 Hb 7.7 g/dL Platelets 4.5 × 104/mm3 TP 7.6 g/dL Alb 2.9 mg/dL BUN 64.2 mg/dL Cr 2.85 mg/dL UA 16.5 mg/dL Na 144 mEq/L K 4.7 mEq/L Cl 112 mEq/L Ca 8.1 mEq/L AST 12 IU/L ALT 6 IU/L LDH 248 IU/L ALP 106 IU/L IgG 3152 mg/dL IgA 82 mg/dL IgM 29 mg/dL β-2MG 12.9 mg/L Free light chain κ 2100 mg/L Free light chain λ 5.8 mg/L κ/λ ratio 362 ==== Refs 1 Cooper N. Bussel J. The pathogenesis of immune thrombocytopaenic purpura British Journal of Haematology 2006 133 4 364 374 10.1111/j.1365-2141.2006.06024.x 2-s2.0-33645814192 16643442 2 Saif M. W. Hopkins J. L. Gore S. D. Autoimmune phenomena in patients with myelodysplastic syndromes and chronic myelomonocytic leukemia Leukemia & Lymphoma 2002 43 11 2083 2092 10.1080/1042819021000016186 2-s2.0-0036838704 12533032 3 Kyle R. A. Pierre R. V. Bayrd E. D. Multiple myeloma and acute myelomonocytic leukemia The New England Journal of Medicine 1970 283 21 1121 1125 10.1056/nejm197011192832101 2-s2.0-0014955268 5273282 4 Govindarajan R. Jagannath S. Flick J. T. Preceding standard therapy is the likely cause of MDS after autotransplants for multiple myeloma British Journal of Haematology 1996 95 2 349 353 10.1046/j.1365-2141.1996.d01-1891.x 2-s2.0-0029804411 8904891 5 Cuzick J. Erskine S. Edelman D. Galton D. A. G. A comparison of the incidence of the myelodysplastic syndrome and acute myeloid leukaemia following melphalan and cyclophosphamide treatment for myelomatosis. A report to the Medical Research Council's working party on leukaemia in adults British Journal of Cancer 1987 55 5 523 529 10.1038/bjc.1987.107 2-s2.0-0023198180 3300761 6 Roeker L. E. Larson D. R. Kyle R. A. Kumar S. Dispenzieri A. Rajkumar S. V. Risk of acute leukemia and myelodysplastic syndromes in patients with monoclonal gammopathy of undetermined significance (MGUS): a population-based study of 17 315 patients Leukemia 2013 27 6 1391 1393 10.1038/leu.2013.34 2-s2.0-84878909164 23380709 7 Mailankody S. Pfeiffer R. M. Kristinsson S. Y. Risk of acute myeloid leukemia and myelodysplastic syndromes after multiple myeloma and its precursor disease (MGUS) Blood 2011 118 15 4086 4092 10.1182/blood-2011-05-355743 2-s2.0-80054106794 21795746 8 Akashi K. Harada M. Shibuya T. Simultaneous occurrence of myelomonocytic leukemia and multiple myeloma: involvement of common leukemic progenitors and their developmental abnormality of ‘lineage infidelity’ Journal of Cellular Physiology 1991 148 3 446 456 10.1002/jcp.1041480317 2-s2.0-0026006116 1655817 9 Economopoulos T. Pappa V. Panani A. Myelopathies during the course of multiple myeloma Haematologica 1991 76 4 289 292 2-s2.0-0026041927 1794735 10 Rios Tamayo R. Sole F. Gascon F. Simultaneous occurrence of the 5q-syndrome and multiple myeloma Clinical and Laboratory Haematology 2000 22 1 49 51 10.1046/j.1365-2257.2000.00274.x 2-s2.0-0010452173 10762305 11 Kastanas G. A. Papaioannou C. A. Nathena H. A. Multiple myeloma and acute leukemia Acta Haematologica 1979 62 2 78 80 2-s2.0-0018649023 119404 12 Raz I. Polliack A. Coexistence of myelomonocytic leukemia and monoclonal gammopathy or myeloma. Simultaneous presentation in three patients Cancer 1984 53 1 83 85 10.1002/1097-0142(19840101)53:1<83::aid-cncr2820530115>3.0.co;2-q 2-s2.0-0021333291 6360331 13 Parker A. C. A case of acute myelomonocytic leukaemia associated with myelomatosis Scandinavian Journal of Haematology 1973 11 4 257 260 2-s2.0-0015760176 4521505 14 Cleary B. Binder R. A. Kales A. N. Veltri B. J. Simultaneous presentation of acute myelomonocytic leukemia and multiple myeloma Cancer 1978 41 4 1381 1386 10.1002/1097-0142(197804)41:4<1381::aid-cncr2820410423>3.0.co;2-a 2-s2.0-0017817926 273468 15 Taddeini L. Schrader W. Concomitant myelomonocytic leukemia and multiple myeloma Minnesota Medicine 1972 55 5 446 448 2-s2.0-0015336620 4503608 16 Parapia L. Abbott C. R. Masters G. Roberts B. E. Simultaneous presentation of acute myelomonocytic leukaemia and multiple myeloma Acta Haematologica 1982 68 2 153 156 2-s2.0-0019973087 6812372 17 Shibata K. Shimamoto Y. Nakazato S. Matsuzaki M. Tadano J. Refractory anaemia with ringed sideroblasts concurrent with multiple myeloma—a brief review of the recent literature Haematologia 1997 28 4 199 205 2-s2.0-0031309120 9408763 18 Mufti G. J. Hamblin T. J. Clein G. P. Race C. Coexistent myelodysplasia and plasma cell neoplasia British Journal of Haematology 1983 54 1 91 96 10.1111/j.1365-2141.1983.tb02070.x 2-s2.0-0020579728 6849839 19 Copplestone J. A. Mufti G. J. Hamblin T. J. Oscier D. G. Immunological abnormalities in myelodysplastic syndromes. II. Coexistent lymphoid or plasma cell neoplasms: a report of 20 cases unrelated to chemotherapy British Journal of Haematology 1986 63 1 149 159 10.1111/j.1365-2141.1986.tb07505.x 2-s2.0-0022619224 3707860 20 Annino L. Martino P. Barsotti P. Serra P. Marinozzi V. Mandelli F. Multiple myeloma and acute myelomonocytic leukemia: simultaneous occurrence without previous chemotherapy Acta Haematologica 1980 64 4 195 200 2-s2.0-0019174544 6781197 21 Grogan T. M. Durie B. G. M. Spier C. M. Richter L. Vela E. Myelomonocytic antigen positive multiple myeloma Blood 1989 73 3 763 769 2-s2.0-0024583777 2644987 22 Epstein J. Xiao H. He X.-Y. Markers of multiple hematopoietic-cell lineages in multiple myeloma The New England Journal of Medicine 1990 322 10 664 668 10.1056/nejm199003083221005 2-s2.0-0025318069 2137556 23 Rosado F. G. Morice W. G. He R. Howard M. T. Timm M. Mcphail E. D. Immunophenotypic features by multiparameter flow cytometry can help distinguish low grade B-cell lymphomas with plasmacytic differentiation from plasma cell proliferative disorders with an unrelated clonal B-cell process British Journal of Haematology 2015 169 3 368 376 10.1111/bjh.13303 2-s2.0-84927693718 25644063 24 Hadjadj J. Michel M. Chauveheid M.-P. Godeau B. Papo T. Sacre K. Immune thrombocytopenia in chronic myelomonocytic leukemia European Journal of Haematology 2014 93 6 521 526 10.1111/ejh.12393 2-s2.0-84911375681 24913055 25 Kalina U. Hofmann W.-K. Koschmieder S. Alteration of c-mpl-mediated signal transduction in CD34+ cells from patients with myelodysplastic syndromes Experimental Hematology 2000 28 10 1158 1163 10.1016/s0301-472x(00)00527-0 2-s2.0-0033797720 11027834 26 Brierley C. K. Steensma D. P. Thrombopoiesis-stimulating agents and myelodysplastic syndromes British Journal of Haematology 2015 169 3 309 323 10.1111/bjh.13285 2-s2.0-84927641846 25659186 27 Svensson T. Chowdhury O. Garelius H. A pilot phase I dose finding safety study of the thrombopoietin-receptor agonist, eltrombopag, in patients with myelodysplastic syndrome treated with azacitidine European Journal of Haematology 2014 93 5 439 445 10.1111/ejh.12383 2-s2.0-84911978389 24853277 28 Prica A. Sholzberg M. Buckstein R. Safety and efficacy of thrombopoietin-receptor agonists in myelodysplastic syndromes: a systematic review and meta-analysis of randomized controlled trials British Journal of Haematology 2014 167 5 626 638 10.1111/bjh.13088 2-s2.0-84925221726 25155450 29 Saleh M. N. Bussel J. B. Cheng G. Safety and efficacy of eltrombopag for treatment of chronic immune thrombocytopenia: results of the long-term, open-label EXTEND study Blood 2013 121 3 537 545 10.1182/blood-2012-04-425512 2-s2.0-84872582501 23169778 30 Will B. Kawahara M. Luciano J. P. Effect of the nonpeptide thrombopoietin receptor agonist Eltrombopag on bone marrow cells from patients with acute myeloid leukemia and myelodysplastic syndrome Blood 2009 114 18 3899 3908 10.1182/blood-2009-04-219493 2-s2.0-70449726874 19710504 31 Erickson-Miller C. L. Kirchner J. Aivado M. May R. Payne P. Chadderton A. Reduced proliferation of non-megakaryocytic acute myelogenous leukemia and other leukemia and lymphoma cell lines in response to eltrombopag Leukemia Research 2010 34 9 1224 1231 10.1016/j.leukres.2010.02.005 2-s2.0-77955271313 20202683 32 Harada H. Watanabe M. Suzuki K. Lenalidomide is active in Japanese patients with symptomatic anemia in low- or intermediate-1 risk myelodysplastic syndromes with a deletion 5q abnormality International Journal of Hematology 2009 90 3 353 360 10.1007/s12185-009-0400-8 2-s2.0-76449116206 19705057 33 Fenaux P. Giagounidis A. Selleslag D. A randomized phase 3 study of lenalidomide versus placebo in RBC transfusion-dependent patients with Low-/Intermediate-1-risk myelodysplastic syndromes with del5q Blood 2011 118 14 3765 3776 10.1182/blood-2011-01-330126 2-s2.0-80053621748 21753188 34 Raza A. Reeves J. A. Feldman E. J. Phase 2 study of lenalidomide in transfusion-dependent, low-risk, and intermediate-1 risk myelodysplastic syndromes with karyotypes other than deletion 5q Blood 2008 111 1 86 93 10.1182/blood-2007-01-068833 17893227 35 Buckstein R. Kerbel R. Cheung M. Lenalidomide and metronomic melphalan for CMML and higher risk MDS: a phase 2 clinical study with biomarkers of angiogenesis Leukemia Research 2014 38 7 756 763 10.1016/j.leukres.2014.03.022 2-s2.0-84902115955 24819395 36 Adès L. Sekeres M. A. Wolfromm A. Predictive factors of response and survival among chronic myelomonocytic leukemia patients treated with azacitidine Leukemia Research 2013 37 6 609 613 10.1016/j.leukres.2013.01.004 2-s2.0-84877084272 23415110 37 Sekeres M. A. Tiu R. V. Komrokji R. Phase 2 study of the lenalidomide and azacitidine combination in patients with higher-risk myelodysplastic syndromes Blood 2012 120 25 4945 4951 10.1182/blood-2012-06-434639 2-s2.0-84869853588 22915641

PMC005xxxxxx/PMC5002295.txt

==== Front Int J Reprod MedInt J Reprod MedIJRMEDInternational Journal of Reproductive Medicine2356-71042314-5757Hindawi Publishing Corporation 10.1155/2016/3480629Clinical StudyCould Metformin Manage Gestational Diabetes Mellitus instead of Insulin? http://orcid.org/0000-0003-0056-938XSaleh Hend S. * Abdelsalam Walid A. Mowafy Hala E. Abd ElHameid Azza A. Obstetrics and Gynecology Department, Faculty of Medicine, Zagazig University, Zagazig 44519, Egypt*Hend S. Saleh: drhendsaleh@yahoo.comAcademic Editor: Samir Hamamah 2016 14 8 2016 2016 34806299 4 2016 19 6 2016 12 7 2016 Copyright © 2016 Hend S. Saleh et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Gestational diabetes mellitus (GDM) complicates a significant number of pregnancies. Blood glucose control improves perinatal outcomes. Medical nutrition therapy is the foundation in management. Aim of This Study. To evaluate efficacy of metformin in comparison to insulin for managing GDM. Methods. In prospective randomized comparative study, 150 antenatal women whose pregnancies had been complicated by GDM and did not respond to diet alone were recruited from antenatal clinics at Obstetrics Department in Zagazig University Hospitals from November 2012 to December 2014. They were divided randomly into two groups, 75 patients in each, and were subjected to either insulin or metformin medication. Outcomes were comparing the effects of both medications on maternal glycemic control, antenatal complications, and neonatal outcome. Results. No significant difference in controlling high blood sugar in GDM with the use of metformin or insulin (P = 0.95, 0.15). Maternal complications in both groups had no significant difference and fetal outcomes were as well similar except the fact that the hypoglycemia occurred more in insulin group with P value 0.01. Conclusion. Glycaemic control in GDM can be achieved by using metformin orally without increasing risk of maternal hypoglycemia with satisfying neonatal outcome. ==== Body 1. Introduction Gestational diabetes mellitus (GDM) is a condition with any level of glucose intolerance which began or was detected for first time during pregnancy despite type of management; it may also relate to situations that continue after pregnancy. It affects approximately 7% of pregnancies with an incidence of more than 200,000 cases per year [1]. 40–60% of gestational diabetes mellitus (GDM) cases have chance of developing diabetes mellitus over the 5–10 years after pregnancy [2]. Older and more obese pregnant women have the highest incidence of GDM. It is associated with numerous undesirable outcomes over the short and long term for both mother and neonate [3]. Incidence of preeclampsia and rate of cesarean section increased in GDM as some of the short term complications. Developing type 2 diabetes (T2D) after pregnancy is one of the long term bad maternal outcomes [4]. Extreme mother-to-fetus glucose transmit is an augmented hazard for congenital defects, neonatal death, and still birth. The hyperglycemic environment intrauterine influences children later in life [5]. Macrosomia is an extraimportant complication which is a risk factor for instrumental delivery, shoulder dystocia, and cesarean section during delivery. Neonatal hypoglycemia directly after birth is one of the most risky complications, putting neonate in danger [6]. The first screening test for GDM was advised in 1973, in the form of the 1-h 50 gm oral glucose tolerance test [7]. Some guidelines recommended screening common screening to all pregnant women to improve pregnancy outcomes. Others excluded low risk patients who were <25 years old with normal body weight, no history of abnormal glucose metabolism, no first-degree relatives with diabetes, and no history of poor obstetric outcomes [8]. During first antenatal visit, pregnant women with high risk for GDM should be screened for it immediately. If negative, they should be retested between 24 and 28 weeks of gestation. Average risk pregnant women (neither high nor low risk) should be screened between 24 and 28 weeks of gestation [1]. The World Health Organization (WHO) recommends using a 75-gm glucose tolerance test for screening and diagnosis. The doorstep values are a fasting glucose concentration of more than 126 mg/dL (7.0 mmol/L) and/or a 2-h glucose concentration of more than 140 mg/dL (7.8 mmol/L) [9]. Recently, trials have exhibited that efficient management of hyperglycemia in women with GDM is the main principle to prevent hyperinsulinemia and fetal macrosomia [10]. We diagnosed GDM by the American Diabetes Association (ADA) criteria, depending on 75-gm glucose load then checking the fasting serum glucose concentration, 1-h glucose concentration, and 2-h glucose concentration [11]. The glucose threshold values are 95 mg/dL (5.3 mmol/L), 180 mg/dL (10.0 mmol/L), and 155 mg/dL (8.6 mmol/L), respectively. Two or more abnormal values are required for diagnosis. Some studies have shown that a single abnormal value is significantly associated with amplified neonatal hazards [12]. When the World Health Organization (WHO) advised using a 75-gm glucose tolerance test for screening and diagnosis of GDP with the threshold values of a fasting glucose concentration of more than 126 mg/dL (7.0 mmol/L) and/or a 2-h glucose concentration of more than 140 mg/dL (7.8 mmol/L), about twice as a lot of patients will be positive diagnosis [9]. The main management started by dietary and exercise counseling, but about 20–60% of GDM patients often require pharmacological treatment, which has conventionally been insulin [13]. Dietary adjustment is frequently called medical nutrition therapy. Evidence shows that it is efficient in glycaemic control and improving pregnancy and neonatal outcomes [14]. American Diabetes Association (ADA) suggests exercise programs in moderate level for those who have no medical or obstetrical complications, in the form of 3 or more times per week for 30 min [11]. The American College of Obstetrics and Gynecology (ACOG) recommended that GDM patients should keep up the following capillary blood glucose values: preprandial glucose <95 mg/dL (5.3 mmol/L), 1-h postprandial glucose 130–140 mg/dL (7.8 mmol/L), and 2-h postprandial glucose <120 mg/dL (6.7 mmol/L) [15]. Others recommend maintaining fasting glucose levels of <90–99 mg/dL (5.0–5.5 mmol/L), 1-h postprandial glucose levels of <140 mg/dL (7.8 mmol/L), and 2-h postprandial glucose levels of <120–127 mg/dL (6.7–7.1 mmol/L) [16]. There is agreement that measuring postprandial glucose levels is more important compared to preprandial levels since the former associates better with definite neonatal risks like hypoglycemia, macrosomia, and shoulder dystocia [17]. The pharmacological interference is in the form of either subcutaneous insulin which has been considered the standard for management of GDM or oral hypoglycaemic agents (metformin and glyburide) [18]. Insulin regimens frequently consist of intermediate acting insulin such as isophane and short acting agents such as regular recombinant insulin (Humulin R) [19]. Adjustments of its doses depend on the patient's body mass index, glucose levels, and lifestyle. Insulin therapy has several disadvantages including multiple daily injections, the risk of hypoglycemia, and maternal weight gain [20]. Health education for dose adjustment of insulin is essential to provide confident safe self-administration of insulin. Currently, considerable costs of health education on the safe use of insulin in addition to the cost of the drug itself are chased. Observably, oral therapy if safe and effective could be more satisfactory and desired [3]. So, it is good idea to use oral hypoglycemic agents in controlling blood sugar. Hypothetically, metformin is an alternative to insulin in the treatment of hyperglycemia during pregnancy. It reduces hyperglycemia by suppressing hepatic glucose output so it reduces hepatic gluconeogenesis and it is intensifying insulin sensitivity therefore enhancing peripheral glucose uptake [21]. It has been found to have a rate of maternofetal transfer of 10–16%. Before, it had not been widely used in GDM but, nowadays, growing studies focus on investigating the effectiveness and safety of metformin in such cases. These Studies were either case-control, observational trials or randomized controlled trials (RCTs). Still its use in pregnancy is controversy [22]. The aim of this study is to compare efficacy and safety of metformin to those of insulin on glycemic control and maternal and neonatal outcomes in GDM to reach end conclusion about the possibility of replacing insulin by metformin in pregnancy. 2. Methods It is a prospective randomized comparative study. One hundred and fifty antenatal women whose pregnancies had been complicated by GDM and did not respond to diet modifications or nutritional instructions alone in 3 weeks were recruited from antenatal clinics at Obstetrics Department in Zagazig University Hospitals from November 2012 to December 2014. GDM was diagnosed at 26–34 weeks using WHO criteria: fasting plasma glucose ≥7.0 mmol/L or 2-h value >7.8 mmol/L following a 2-h 75 g OGTT [23]. Exclusion criteria were type 1 and type 2 diabetes and anyone who was already on insulin treatment, recognized fetal anomaly by ultrasound investigation, the fact that mother had hypersensitivity or intolerance to metformin intake like gastrointestinal side effects, liver or kidney diseases, and any obstetric high risk conditions. After the study protocol was approved by the Research Ethics Committee of the Zagazig University Hospitals, the research course was completely explained to the participants receiving their verbal and written informed consents. They were divided randomly into two groups by permuted block randomization. Each group had 75 pregnant mothers. Group 1 received metformin orally initially at dose of 500 mg/day with meals which slowly increased up to 3000 mg in divided doses as tolerated by the patient and till controlled glycemic profile was realized. If the target was not achieved or tolerance was not achieved then insulin was commenced. Group 2 received insulin as a combination of short acting (Actrapid) and intermediate acting (Mixtard) human insulin as twice daily injections before breakfast and before dinner to face the three meals and three snacks per day depending on individual patient requirement, in order to achieve the desired glycemic goals. 24-hour total insulin dose was estimated using 0.6 units/kg body weight in 1st trimester, 0.7 units/kg body weight in 2nd trimester, 0.8 units/kg body weight from 28 to 32 weeks of gestation, 0.9 units/kg body weight from 32 to 36 weeks of gestation, and 1 unit/kg body weight from 36 weeks onwards. Monitoring at home was done by estimating blood glucose levels. Fasting and 2-hour postprandial blood sugar had been measured after the three main meals. The target of management was to maintain fasting blood sugar (FBS) at <100 mg/dL (5.5 mmol/lit) and postprandial blood sugar (PPBS) levels at <120 mg/dL (7 mmol/lit). Glycemic profile, fasting blood sugar (FBS), and postprandial blood sugar (2 hr PPBS) were done weekly for all cases. Dose modifications of drugs were made at each antenatal visit weekly till delivery. Usual obstetric care was offered at the antenatal clinics including ultrasound examination which was done at first visit (dating scan) and then at 16–19 weeks (anomaly scan) and then monthly after 28 weeks as fetal well-being scan. HbAIC was done at entrance of study and at around 37 weeks of pregnancy. Mode and time of delivery were decided around 38 weeks of pregnancy. Maternal outcome in the form of glycemic control, medical complications, and mode of delivery were documented. Neonatal outcomes were recorded and all were statistically analyzed. The recorded data was evaluated using SPSS12.0. Mean with SD was reported for all continuous variables and was expressed as mean ± standard deviation (SD). Qualitative analysis was done using Student's t-test. Two-sample independent Student's t-test and Mann-Whitney test were used for continuous data. For quantitative analysis chi-square test Fisher Exact test, and Mann-Whitney test were used. Statistical significance was considered at P value of <0.05. 3. Results This prospective comparative study is to compare the usefulness of metformin versus human insulin in patients with GDM. A total of 150 patients with GDM were registered in the study. They met the inclusion criteria and were randomized to treatment with metformin or insulin. 137 participants completed the study and their data was finally analyzed, with 67 patients in metformin group and 70 patients in insulin group. The design and subject course through the study are exemplified in Figure 1. The demographic characteristics of metformin and insulin groups at the time of diagnosis of GDM were similar (Table 1). Fasting and 2-hour postprandial blood glucose levels were statistically analogous in two groups. Glycemic targets were achieved and maintained throughout pregnancy in the intention variety with no statistical difference in both groups (Table 2). GDM was diagnosed around the period of 26–34 weeks of gestation in our participants. Preeclampsia developed in 13 patients of the metformin group and in 12 patients of insulin group. Seven patients developed preterm labour in metformin group versus 5 patients in insulin group. Eight patients in metformin group developed polyhydramnios whereas only 6 patients in insulin group showed polyhydramnios on growth scan. Urinary tract infection was found in 4 patients in metformin groups versus 3 in insulin group. No significant difference was found between both groups according to medical disorders which developed during antenatal period (Table 3). As for mode of delivery, statistically, no significant differences were found between both groups as 40.2% of metformin group underwent elective cesarean section versus 42.8% in insulin group. 20.8% of metformin group underwent emergency cesarean section versus 22.8% in insulin group. 34.3% of metformin group underwent spontaneous vaginal delivery versus only 28.5% in the insulin group. Assisted vaginal delivery using ventouse was done in cases of metformin group and cases in the insulin group (Table 4) (Figures 2, 3, and 4). Neonatal outcomes were presented in Tables 5(a) and 5(b). Occurrence of transient tachypnea, respiratory distress, neonatal jaundice, need for phototherapy, or admission to neonatal intensive care unit in both groups was comparable with no significant difference. Hypoglycemia developed in 7 babies of metformin group and 15 cases in insulin group with P value 0.01 which is statistically significant. No birth trauma happened in any baby of any group, Table 5(a). There was no significant difference between both groups with regard to mean gestational age at birth, Apgar score at 5, estimated fetal weight, or presence of congenital anomalies Table 5(b). 4. Discussion Gestational diabetes mellitus (GDM) has been described as any extent of glucose intolerance with first detection throughout pregnancy and, depending on the diagnostic tests in employment, it complicated 1–14% of all pregnancies in current years [24]. It is one of the most common medical complications of pregnancy which is related to numerous adverse results to mother and raised risks of prenatal morbidity. So, the management of GDM seeks to diminish such risk of unfavorable neonatal and pregnancy complications [25]. Management is based on self-monitoring of blood glucose concentrations, diet, and physical exercise. When these measurements cannot control blood glucose levels in pregnant women, pharmacological therapy is needed to be added in addition [26]. Using of any medication during pregnancy is limited by its safety which depends on crossing the placenta and if it has effect on the fetus or not. A lot of drugs frequently used in pregnancy cross the placenta and may not exert effects on the fetus [27]. For a lot of years, conventionally, the first-line pharmacological management of GDM has been insulin with no fetal or neonatal obstacles [28]. The drawbacks of insulin are as follows; it needs health education, needs many daily subcutaneous injections, and requires dose modification depending on body mass index of patient, occurrence of hypoglycemia, and gaining weight in mother [14]. So, secure and valuable oral therapy would be more suitable and preferred by patients [29]. The controversy of using oral hypoglycemic agents like glyburide and metformin in pregnancy is related to concerns about their safety for the developing fetus [30]. The American College of Obstetricians and Gynecologists (ACOG) does not support or recommend against the use of oral antidiabetic agents in pregnancy [15]. But The United Kingdom National Institute for Health and Clinical Excellence (NICE) recommends metformin use before and during pregnancy and supports metformin and glyburide as choices for handling of gestational diabetes [2]. Most studies discussed the amount of transplacental passage of glyburide; Kraemer et al. recognized active transport of glyburide from the fetal circulation to the maternal one that may guard the fetus from contact with the drug [31]. Langer et al. could not detect any glyburide in the cord blood at delivery despite its presence in maternal serum [32]. Other studies indicate that fetal concentrations of glyburide may be 1% to 2% of maternal concentration [33]. Moore et al. had compared neonatal outcomes of pregnant women with GDM that were managed by metformin to those with glyburide in randomized study but found no significant difference [34]. In the current study, we preferred using metformin as one of the oral hypoglycemic medications. The occurrence of unfavorable outcomes either in pregnancy or in neonate were not raised in those who were managed with metformin compared with those who were managed with insulin except the fact that the neonatal hypoglycemia happened more in insulin group. The results of our study to high extent were similar to studies by Coetzee and Jackson in 1970. They were the first researchers who studied metformin during pregnancy in women with insulin-independent diabetes. Their study had also two groups of patients; one received metformin and the second one received insulin. The maternal and perinatal outcome were the same for both [35]. The studies on this issue have been continuing from Coetzee and Jackson until Lim in 1997 [36], who was the first one that described that GDM can be managed efficiently and securely with oral hypoglycemic drugs with no distinction in pregnancy outcomes. Then, in 2000, Hellmuth and colleagues [37] presented a cohort study of type 2 DM pregnant women on metformin in opposition to glyburide versus insulin. Their results proposed apprehension for the use of metformin because of the raised rate of preeclampsia (32% metformin versus 7% glyburide versus 10% insulin) and intrauterine fetal death (8% versus 0% versus 2.3%, correspondingly). Conversely, this study has become controversial with reviewers arguing that women in the study were not sound matched. Those women who received the metformin were morbidly obese and started to use the medication later on in the pregnancy. Consequently, the women were essentially at threat for poor pregnancy outcomes not related to metformin [38]. Rowan et al. 2008 [39] had randomized Australian study performed on women with gestational diabetes between 20 and 33 weeks of pregnancy getting metformin or insulin. There was no difference in efficacy between both groups in controlling glucose levels. Infants of metformin group had a lower rate of hypoglycemia compared with infants of insulin group with no more neonatal outcomes. Glueck and his colleagues proved that, in many studies, metformin in pregnancy was not associated with increased incidence of medical disorders in pregnancy and mainly preeclampsia or hypoglycemia also associated with less spontaneous abortion, fetal anomalies, and neonatal complications [40]. The results of this current study were comparable to the findings of Glueck et al. as we also found that metformin intake during pregnancy was not associated with increasing rate of preeclampsia or neonatal complications. Rai et al. [22] in their prospective observational study were comparing metformin to insulin for patients with GDM and type 2 DM (T2DM) in pregnancy. They found that glycemic control was better with metformin after 1 week of therapy and also throughout gestation compared to insulin and also found no major complications or perinatal deaths related to metformin uptake. They proved that metformin is clinically efficient, inexpensive, and a harmless alternative to insulin therapy in pregnant diabetic women. These days, more studies center on investigating the effectiveness and safety of metformin when used during pregnancy in managing GDM. Some are observational studies and others are case-control trials [41]. The randomized controlled trials (RCTs) are present but with little samples which are deficient in the authority to represent valid conclusion about the use of metformin for managing GDM. Our study had the same drawback that although it is randomized one, the sample size was small. 5. Conclusion Metformin is effective and safe in gestational diabetes mellitus (GDM) and could be used instead of insulin in such cases. Ethical Approval The study was approved by the institutional ethics committee. Competing Interests The authors declare that they have no conflict of interests. Figure 1 Study in flow chart. Figure 2 The figure shows pattern of blood sugar in both groups before and after treatment (fasting blood sugar (FBS) and 2 h postprandial blood sugar (2 h PPBS)). Figure 3 The figure shows % of modes of deliveries in metformin group. Figure 4 The figure shows % of modes of deliveries in insulin group. Table 1 Demographic profile of metformin and insulin groups. Parameter Metformin group N: 67 Insulin group N: 70 P value Age (years) 31 ± 3.42 29.8 ± 2.18 0.398 Parity 3.05 ± 1.61 3.24 ± 1.72 0.253 Family history        (i) Diabetes 40% 42% 0.911  (ii) Hypertension 31% 34% 0.897  (iii) Preeclampsia 29% 27% 0.963 Mean gestational age of diagnosis of GDM 27.28 ± 3.458 29.31 ± 3.12 0.348 BMI-early pregnancy (kg/m2) 30.52 ± 3.17 31.58 ± 30.12 0.614 BMI-late pregnancy (kg/m2) 34.28 ± 2.17 37.11 ± 3.87 0.016 Table 2 Fasting blood sugar (FBS) and mean 2-hr postprandial blood sugar at start and throughout treatment (mg/dL). Parameter Metformin group Insulin group P value Mean fasting (FBS) (mg/dL) at starting of treatment 136.09 ± 39.85 137.56 ± 41.10 0.869 Mean 2-hr postprandial blood sugar at starting of treatment 198.32 ± 214.67 196.52 ± 15.45 0.451 Mean fasting (FBS) (mg/dL) throughout treatment 93.25 ± 13.7 94.33 ± 11.1 0.953 Mean 2-hr postprandial blood sugar throughout treatment 116.52 ± 3.53 117.12 ± 3.45 0.158 Table 3 Maternal complications in study groups. Maternal complication   Metformin group   Insulin group P value   N: 67   N: 70 % N % N Preeclampsia 19.4% 13 17.1% 12 0.273 Preterm 10.4% 7 7.1% 5 0.039 Polyhydramnios 11.9% 8 8.5% 6 0.710 Urinary tract infection 5.9% 4 4.2% 3 0.801 Table 4 Mode of delivery between metformin and insulin groups. Mode of delivery   Metformin group   Insulin P value   N: 67   N: 70 % N % N Elective LSCS 40.2% 27 42.8% 30 0.61 Emergency LSCS 20.8% 14 22.8% 16 0.37 Assisted vaginal delivery (vacuum extraction) 4.4% 3 5.7% 4 0.21 Spontaneo-us vaginal delivery 34.3% 23 28.5% 20 0.14 Table 5 Neonatal outcomes. (a) Variable   Metformin   Insulin P value   N: 67   N: 70 % N % N Hypoglycemia 10.4% 7 21.4% 15 0.01 Transient tachypnea 2.9% 2 4.2% 3 0.67 Respiratory distress 1.4% 1 2.8% 2 0.85 Neonatal jaundice 19.4% 14 15.7% 11 0.31 Phototherapy 19.4% 14 15.7% 11 0.31 Neonatal intensive care unit admission 14.9% 10 17.1% 12 0.51 Birth trauma 0% 0%   (b) Variable   Metformin   Insulin P value   N: 67   N: 70 % N % N Apgar score at 5 minutes < 7 1.5% 1 1.4% 1 0.59 Gestational age at birth 38.7 ± 1.1 38.9 ± 1.4 0.06 Estimated weight > 90th percentile 14.9% 10 15.7% 11 0.89 Estimated weight < 10th percentile 5.9% 4 7.1% 5 0.31 Congenital anomalies 1.5% 1 2.8% 2 0.91 ==== Refs 1 Expert Committee on the Diagnosis and Classification of Diabetes Mellitus Report of the expert committee on the diagnosis and classification of diabetes mellitus Diabetes Care 2003 26 supplement 1 S5 S20 12502614 2 National Diabetes Information Clearinghouse (NIDC) Diabetes Overview 2008 National Diabetes Information Clearinghouse. NIH Publication 3 Maymone A. C. Baillargeon J.-P. Ménard J. Ardilouze J.-L. Oral hypoglycemic agents for gestational diabetes mellitus? Expert Opinion on Drug Safety 2011 10 2 227 238 10.1517/14740338.2011.521740 2-s2.0-79952327338 21210750 4 Ben-Haroush A. Yogev Y. Hod M. Epidemiology of gestational diabetes mellitus and its association with type 2 diabetes Diabetic Medicine 2004 21 2 103 113 10.1046/j.1464-5491.2003.00985.x 2-s2.0-1242292305 14984444 5 Metzger B. E. Gabbe S. G. Persson B. International association of diabetes and pregnancy study groups recommendations on the diagnosis and classification of hyperglycemia in pregnancy Diabetes Care 2010 33 3 676 682 10.2337/dc09-1848 20190296 6 Esakoff T. F. Cheng Y. W. Sparks T. N. Caughey A. B. The association between birthweight 4000 g or greater and perinatal outcomes in patients with and without gestational diabetes mellitus American Journal of Obstetrics and Gynecology 2009 200 6 672 e4 10.1016/j.ajog.2009.02.035 2-s2.0-67349182890 19376489 7 O'Sullivan J. B. Mahan C. M. Charles D. Dandrow R. V. Screening criteria for high risk gestational diabetic patients American Journal of Obstetrics and Gynecology 1973 116 7 895 900 2-s2.0-0015789981 4718216 8 Cosson E. Benchimol M. Carbillon L. Universal rather than selective screening for gestational diabetes mellitus may improve fetal outcomes Diabetes and Metabolism 2006 32 2 140 146 10.1016/S1262-3636(07)70260-4 2-s2.0-33646007672 16735962 9 Alberti K. G. M. M. Zimmet P. Z. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus. Provisional report of a WHO consultation Diabetic Medicine 1998 15 7 539 553 2-s2.0-0031851293 9686693 10 Langer O. A spectrum of glucose thresholds may effectively prevent complications in the pregnant diabetic patient Seminars in Perinatology 2002 26 3 196 205 10.1053/sper.2002.33962 2-s2.0-0035989897 12099309 11 American Diabetes Association Gestational diabetes mellitus Diabetes Care 2003 26 supplement 1 S103 S105 10.2337/diacare.26.2007.s103 12502631 12 Lindsay M. K. Graves W. Klein L. The relationship of one abnormal glucose tolerance test value and pregnancy complications Obstetrics and Gynecology 1989 73 1 103 106 2-s2.0-0024545365 2909030 13 Langer O. From educated guess to accepted practice: the use of oral antidiabetic agents in pregnancy Clinical Obstetrics and Gynecology 2007 50 4 959 971 10.1097/grf.0b013e31815a55f3 2-s2.0-35848965693 17982339 14 de Lima H. T. Lopes Rosado E. Ribeiro Neves P. A. Monteiro Machado R. C. de Oliveira L. M. Saunders C. Systematic review; nutritional therapy in gestational diabetes mellitus Nutricion Hospitalaria 2013 28 6 1806 1814 10.3305/nh.2013.28.6.6892 2-s2.0-84891690322 24506355 15 American College of Obstetricians and Gynecologists Committee on Practice Bulletins—Obstetrics ACOG practice bulletin. Clinical management guidelines for obstetrician-gynecologists. Number 30, September 2001 (replaces Technical Bulletin Number 200, December 1994). Gestational diabetes Obstetrics & Gynecology 2001 98 3 525 538 11547793 16 Metzger B. E. Buchanan T. A. Coustan D. R. Summary and recommendations of the Fifth International Workshop-Conference on Gestational Diabetes Mellitus Diabetes Care 2007 30 2 S251 S260 10.2337/dc07-s225 2-s2.0-34447129042 17596481 17 De Veciana M. Major C. A. Morgan M. A. Postprandial versus preprandial blood glucose monitoring in women with gestational diabetes mellitus requiring insulin therapy The New England Journal of Medicine 1995 333 19 1237 1241 10.1056/nejm199511093331901 2-s2.0-0028880530 7565999 18 Magon N. Seshiah V. Gestational diabetes mellitus: non-insulin management Indian Journal of Endocrinology and Metabolism 2011 15 4 284 293 10.4103/2230-8210.85580 22028999 19 Magon N. Seshiah V. Gestational diabetes mellitus: insulinic management Journal of Obstetrics and Gynecology of India 2014 64 2 82 90 10.1007/s13224-014-0525-4 2-s2.0-84904671027 24757334 20 Simmons D. Metformin treatment for type 2 diabetes in pregnancy? Best Practice and Research: Clinical Endocrinology and Metabolism 2010 24 4 625 634 10.1016/j.beem.2010.05.002 2-s2.0-79952117035 20832741 21 Kirpichnikov D. McFarlane S. I. Sowers J. R. Metformin: an update Annals of Internal Medicine 2002 137 1 25 33 10.7326/0003-4819-137-1-200207020-00009 2-s2.0-0037008045 12093242 22 Rai L. Meenakshi D. Kamath A. Metformin—a convenient alternative to insulin for Indian women with diabetes in pregnancy Indian Journal of Medical Sciences 2009 63 11 491 497 10.4103/0019-5359.58878 2-s2.0-76549102255 20075550 23 IADPSG Consensus Panel International association of diabetes and pregnancy study groups recommendations on the diagnosis and classification of hyperglycemia in pregnancy Diabetes Care 2010 33 3 676 682 10.2337/dc09-1848 20190296 24 American Diabetes Association Diagnosis and classification of diabetes mellitus Diabetes Care 2010 33 supplement 1 S62 S69 10.2337/dc10-S062 20042775 25 Horvath K. Koch K. Jeitler K. Effects of treatment in women with gestational diabetes mellitus: systematic review and meta-analysis British Medical Journal 2010 340 7750 p. 796 c1395 10.1136/bmj.c1395 2-s2.0-77950636044 26 Crowther C. A. Hiller J. E. Moss J. R. McPhee A. J. Jeffries W. S. Robinson J. S. Effect of treatment of gestational diabetes mellitus on pregnancy outcomes The New England Journal of Medicine 2005 352 24 2477 2486 10.1056/nejmoa042973 2-s2.0-20444442419 15951574 27 American Diabetes Association Diagnosis and classification of diabetes mellitus Diabetes Care 2006 30, supplement 1 S42 S47 10.2337/dc07-s042 28 Alwan N. Tuffnell D. J. West J. Treatments for gestational diabetes Cochrane Database of Systematic Reviews 2009 3 CD003395 2-s2.0-68549133262 29 Verma A. Boney C. M. Tucker R. Vohr B. R. Insulin resistance syndrome in women with prior history of gestational diabetes mellitus Journal of Clinical Endocrinology and Metabolism 2002 87 7 3227 3235 10.1210/jc.87.7.3227 2-s2.0-0036324763 12107230 30 Metzger B. E. Lowe L. P. Dyer A. R. Hyperglycemia and adverse pregnancy outcomes The New England Journal of Medicine 2008 358 19 1991 2002 18463375 31 Kraemer J. Klein J. Lubetsky A. Koren G. Perfusion studies of glyburide transfer across the human placenta: implications for fetal safety American Journal of Obstetrics and Gynecology 2006 195 1 270 274 10.1016/j.ajog.2005.12.005 2-s2.0-33745253793 16579925 32 Langer O. Conway D. L. Berkus M. D. Xenakis E. M.-J. Gonzales O. A comparison of glyburide and insulin in women with gestational diabetes mellitus The New England Journal of Medicine 2000 343 16 1134 1138 10.1056/nejm200010193431601 2-s2.0-0034687440 11036118 33 Nanovskaya T. N. Nekhayeva I. Hankins G. D. V. Ahmed M. S. Effect of human serum albumin on transplacental transfer of glyburide Biochemical Pharmacology 2006 72 5 632 639 10.1016/j.bcp.2006.05.019 2-s2.0-33746767522 16828060 34 Moore L. E. Clokey D. Rappaport V. J. Curet L. B. Metformin compared with glyburide in gestational diabetes: a randomized controlled trial Obstetrics and Gynecology 2010 115 1 55 59 10.1097/aog.0b013e3181c52132 2-s2.0-74049144107 20027034 35 Coetzee E. J. Jackson W. P. Themanagement of non insulin DM inpregnancy Diabetes Research and Clinical Practice 1986 1 281 287 3939118 36 Lim J. M. H. Tayob Y. O'Brien P. M. S. Shaw R. W. A comparison between the pregnancy outcome of women with gestation diabetes treated with glibenclamide and those treated with insulin Medical Journal of Malaysia 1997 52 4 377 381 2-s2.0-0031396624 10968114 37 Hellmuth E. Damm P. Mølsted-Pedersen L. Oral hypoglycaemic agents in 118 diabetic pregnancies Diabetic Medicine 2000 17 7 507 511 10.1046/j.1464-5491.2000.00314.x 2-s2.0-0033848295 10972579 38 Norman R. J. Wang J. X. Hague W. Should we continue or stop insulin-sensitizing drugs during pregnancy? Current Opinion in Obstetrics and Gynecology 2004 16 3 245 250 10.1097/00001703-200406000-00007 2-s2.0-2542593109 15129054 39 Rowan J. A. Hague W. M. Gao W. Battin M. R. Moore M. P. Metformin versus insulin for the treatment of gestational diabetes The New England Journal of Medicine 2008 358 19 2003 2015 10.1056/nejmoa0707193 2-s2.0-43249092553 18463376 40 Glueck C. J. Bornovali S. Pranikoff J. Goldenberg N. Dharashivkar S. Wang P. Metformin, pre-eclampsia, and pregnancy outcomes in women with polycystic ovary syndrome Diabetic Medicine 2004 21 8 829 836 10.1111/j.1464-5491.2004.01251.x 2-s2.0-3843127654 15270785 41 Goh J. E. L. Sadler L. Rowan J. Metformin for gestational diabetes in routine clinical practice Diabetic Medicine 2011 28 9 1082 1087 10.1111/j.1464-5491.2011.03361.x 2-s2.0-80051604651 21679232

PMC005xxxxxx/PMC5002296.txt

==== Front Adv Prev MedAdv Prev MedAPMAdvances in Preventive Medicine2090-34802090-3499Hindawi Publishing Corporation 10.1155/2016/5494821Research ArticlePrevalence of Musculoskeletal Symptoms among Dental Health Workers, Southern Thailand http://orcid.org/0000-0001-7621-2008Decharat Somsiri 1 * Phethuayluk Piriyalux 2 Maneelok Supandee 1 1Department of Industrial Hygiene and Health Science, Faculty of Health and Sports Science, Thaksin University, 222 Moo 2 Papayom District, Phatthalung Province 93210, Thailand2Department of Public Health, Faculty of Health and Sports Science, Thaksin University, 222 Moo 2 Papayom District, Phatthalung Province 93210, Thailand*Somsiri Decharat: somsiri_9@hotmail.comAcademic Editor: Masaru Shimada 2016 14 8 2016 2016 54948218 3 2016 17 5 2016 Copyright © 2016 Somsiri Decharat et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Objectives. The objective of this study was to describe the socioeconomic situation of dental health work and work characteristics and to evaluate the prevalence of musculoskeletal symptoms among dental health workers. Material and Methods. A cross-sectional study was conducted with 124 dental health workers and 124 persons in the reference group, matched to dental health workers by gender, were recruited from the workers who worked at the same 17 community hospitals in Nakhon Si Thammarat province, Thailand. Information was collected by using questionnaire. Data analysis comprised descriptive and analytical components. Results and Discussion. 75.8% were female and 24.2% were male dental health workers. 91.9% of subjects had worked >5 years. Most subjects worked for >8 hours per day and worked >6 days per week, at 63.7% and 53.2%, respectively. 100% of subjects worked in public institutions, and 68% also worked in both public and private institutions. Most subjects (52.4%) did not exercise. Daily activity, gender, duration of work, hours worked per day, days worked per week, and physical activity were significantly associated with musculoskeletal symptoms at <0.001. Conclusion. The prevention and reduction of MSDs among dentists should include improving their education in dental ergonomics. Faculty of Health and Sports Science, Thaksin University ==== Body 1. Introduction Dentistry is recognized to be a demanding occupation. In Thailand, the numbers of dentist increased from 16,137 dentists in 2010 to more than 20,000 in 2014 [1]. In addition, health effects data of the dentists in Thailand was weak. While many studies reported the work related musculoskeletal disorders (WMSDs) in the dental personnel across different countries [2–5], the phenomenon knowledge of dentists in Thailand is lacking. Musculoskeletal symptom (MSS) may be defined as pain commonly experienced by dentists in the course of their career [6]. Studies show that the circumstance of musculoskeletal symptoms among dentists varies. Biller, 1994 [7], presented that 65% of dentists complained of back pain in his study. Similarly, Finsen et al. 1998 [8] reported 65% MSS among Danish dentists and 78% of dentists in Thailand [9]. Part-time Thai dentists were found to have a higher proportion of musculoskeletal problems than their full-time counterparts. The number of years since graduation was also significantly associated with musculoskeletal pain in these Thai dentists. Of 264 dental personnel, the report of MSD pain impacting daily activity was 76.1%. Nithithamthada and Puntumetakul, 2013 [10], reported work related MSDs as 71.2%. The objectives in this study were to describe the socioeconomic situation of dental health work and work characteristics and to evaluate the prevalence of musculoskeletal symptoms among dental health workers. 2. Material and Methods The investigation was a cross-sectional descriptive study. This study was a descriptive comparative survey of two groups of healthcare workers, during May and September 2013. 2.1. Study Population and Samples The study population comprised a subject group and a reference group. The subjects were obtained by a purposive sampling, all of 17 community hospitals in Nakhon Si Thammarat province in South of Thailand. The samples size in this study was calculated by estimating from the total number of dental health workers cases. Thus, the subjects enrolled were 124 dental health workers (30 males and 94 females including 16 dentists, 70 dental hygienists, and 38 dental assistants) who worked at 17 community hospitals in Nakhon Si Thammarat province in South of Thailand. 124 reference groups (e.g., 35 general officers, 25 medical record officers, 37 office cleaners, and 27 security officers), matched to dental health workers by gender, were recruited from the workers who worked at the same 17 community hospitals in Nakhon Si Thammarat province in South of Thailand. The inclusion criteria of the subject group were dental health workers aged between 20 and 60 years who had at least one-year experience. They agreed to participate in the study and provided written informed consent. This study was approved by the Ethics Committee of the Institute of Research and Development, Thaksin University. Permission to conduct the study was obtained from the community hospitals owners. All of participants received a clear explanation of the purpose of this study and agreed to participate using signed consent forms. 2.2. Sample Collection The questionnaire collected information on the following variables: general information, work characteristics, occupational life style (e.g., duration of work, hours worked per day, and days worked per week), and presence or absence of musculoskeletal symptoms in the previous 12 months (dependent variables). In addition, sociodemographic variables were age, weight, height, and gender. Behavioral variables were prophylaxis and physical activity. Interviews were conducted after work by trained interviewers. Data was confirmed by using direct observation. 2.3. Statistical Analysis Data analysis comprised descriptive and analytical components. In the descriptive component, workers' personal and occupational characteristics are defined and their musculoskeletal symptoms described. The analytical component assessed the relationships between musculoskeletal symptom prevalence and the independent variables described above. All independent and dependent variables were categorical, so chi-square tests were used to assess these relationships. Data analysis was conducted using SPSS for Windows. 3. Results From 17 purposively selected hospitals, 124 subjects and 124 references were interviewed. The mean age of the dental health workers was 37.3 years and the mean age of the reference group was 39.4 years. Duration of work by dental health workers was 15.5 years and 16.8 years for the reference group. Hours worked per day and days worked per week by the dental health workers were 12.3 hours and 6.3 days, respectively. All of the subjects worked in public institutions, and 68% also worked in a private capacity. Of all respondents in the research study, 75.8% were female and 24.2% were male dental health workers. 91.9% of subjects had worked >5 years. Most subjects worked for >8 hours per day and worked >6 days per week, at 63.7% and 53.2%, respectively. 100% of subjects worked in public institutions, and 68% also worked in both public and private institutions. Most subjects (52.4%) did not exercise. The proportion of musculoskeletal symptoms that interfered with a dental health worker's daily activities is given in Table 1. The 3 most common symptoms that had interfered with a dental health worker's daily activities during the previous 12 months were at the shoulders (27.4%), the neck (23.4%), and the lower back (22.9%), respectively. The neck pain, the lower back, the shoulders, the hands/wrists, and the knees that interfered with daily activity were significantly more likely to be reported by dental health workers (P < 0.001, P < 0.001, P < 0.001, P < 0.001, and P < 0.001, resp.). The occurrence of symptoms in the shoulders, the neck, and the lower back was statistically significantly higher among dental health workers than the reference group. The occurrence of symptoms in the shoulders (90.0%), the neck (83.3%), the lower back (50.0%), and the knees (16.7%) was statistically significantly higher among male dental health workers than female dental health workers (P < 0.001, P < 0.001, P < 0.001, and 0.023, resp.). The working position was significantly associated with a higher prevalence of the shoulders, the neck, the lower back, and the knees symptoms as is shown in Table 2. The occurrence of symptoms in the shoulders (50.%), the neck (41.2%), the lower back (33.3%), the knees (25.4%), and the ankles/feet (9.7%) was statistically significantly higher among dental health workers who worked >5 years compared to dental health workers who worked ≤5 years (P < 0.001, P < 0.001, P < 0.001, P < 0.001, and P < 0.001, resp.). Thus, the duration of work as >5 years was significantly associated with a higher occurrence of symptoms in the shoulders, neck, lower back, knees, and the ankles/feet. The occurrence of symptoms in the shoulders (67.1%), the neck (59.5%), the lower back (43.0%), the upper back (30.4%), the knees (34.5%), the hands/wrists (27.9%), the elbows (19.0%), and the ankles/feet (13.9%) was statistically significantly higher among dental health workers who worked >8 hours compared to dental health workers who worked ≤8 hours (P < 0.001, P < 0.001, P < 0.001, P < 0.001, P < 0.001, P < 0.001, P < 0.001, and P < 0.001, resp.). Thus, >8 hours worked per day appears to be significantly associated with a higher occurrence of symptoms as shown in Table 3. The incidence of symptoms in the shoulders (86.4%), the neck (67.8%), the lower back (47.0%), and the hands/wrists (7.58%) was statistically significantly higher among dental health workers who worked >6 days per week compared to dental health workers who worked ≤6 days (P < 0.001, P < 0.001, P < 0.001, and P < 0.001, resp.). Thus, the days worked per week of >6 days were significantly associated with a higher incidence of symptoms in the shoulders, the neck, the lower back, the knees, and the ankles/feet symptoms. The occurrence of symptoms in the shoulders (72.3%), the neck (69.2%), the lower back (43.0%), the knees (43.1%), the upper back (32.3%), and the ankles/feet (7.7%) was statistically significantly higher among dental health workers who did no exercise compared to dental health workers who did exercise (P < 0.001, P < 0.001, P < 0.001, P < 0.001, P < 0.001, and P < 0.001, resp.). Thus, the prophylaxis of exercise shows significant association with a reduction of symptoms as is shown in Table 4. 4. Discussion The objective of this research was to evaluate musculoskeletal symptoms among dental health workers. Pain in neck, lower back, shoulders, hands/wrists, hips/thighs, and knees was shown in this study that was significantly more prevalent among dental health workers than other nondental health workers. This finding was consistent with Åkesson et al., 2000 [11], who reported that female dental hygienists and female dentists showed higher prevalence of neck, shoulders, and hands/wrists, compared with their referents during the past 1 year. In addition, this finding was consistent with many studies [8], which compared the prevalence of such symptoms to subjects working in a different environment, such as farmers, pharmacists, and office employees. These symptoms presented more often among dentists. Shoulder complaints were more prevalent than other bodily locations. The symptoms are shown for 3 selected musculoskeletal areas: shoulders, neck, and lower back. The findings for shoulder complaints, neck pain, and lower back pain in this study were supported by Finsen et al. [8], who studied by using a questionnaire that shown to musculoskeletal symptoms among dentists, and variation in dental work. 65% and 59% of dentists replied to neck/shoulder and low back, respectively. B. Valachi and K. Valachi, 2003 [6], reported that approximately 80% of dentists in America complained of neck, shoulder, and lower back pain. After this, lower back pain was the 2nd most prevalent musculoskeletal area of complaint as presented by many studies [12–14]: 46% reported in a Greek research [15] and 53.7% reported in an Australian research [16]. More than 25% of all subjects reported back pain; it was as severe chronic back pain [12]. In addition, hands/wrist symptoms were significantly more prevalent among dental health workers than other nondental health workers. This finding was consistent with many studies [12, 17–19], which reported hand/wrist complaints among dentists and especially dental hygienists as high. Hand/wrist complaints follow low back disorders [12, 18] and result in a significantly higher chronic situation than any other complaint [12]. Gender was statistically significantly associated with the occurrence of neck, lower back, shoulder, and knee pain (P < 0.001, P < 0.001, P = 0.023, and P < 0.001, resp.). The findings for neck, lower back, and shoulder pain in this study was similar to Alexopoulos et al., 2004 [12], who reported that gender was a significant factor for neck pain and that the female gender was more significantly related to chronic back and shoulder pain than male. In addition, all complaints of chronic pain increased with age. In this study, the occurrence of the MSDs was statistically significantly higher among male dental health workers than female dental health workers. This finding was not consistent with many studies [20–22], which reported that the results showed consistency with previous studies that female dental students showed a higher prevalence of WMSD symptoms than males. Time working included duration of work being hours worked per day and days worked per week that showed a statistically significant correlation with the occurrence of musculoskeletal symptoms. This finding was also supported by Szymanska [17], who reported on illness of the musculoskeletal system among dentists in Poland. This study showed that a long working time in the course of a day was significantly associated with ergonomics. In addition, years of work consequently increase the number of illnesses of the musculoskeletal system, while years of work or experience working time has been significantly associated with increasing disorders of the musculoskeletal system. It is generally accepted that poor practice such as poor posture when working in a long time can be shown to cause chronic fatigue, discomfort, and pain, even if the soft tissues are not structurally altered. This finding was also supported by Al Wazzan et al. [23], who reported that the disorders of MSDs related with the number of years of dental practice, although the same symptoms were reported in both younger and older dentists. In addition, experiencing pain disorder was significantly earlier with the dentists working in a standing position. This research showed that MSDs occurred within 3 years after working for approximately 10 years of practice and working over 8 hours a day. Exercise showed a statistically significant reduction in the occurrence of musculoskeletal symptoms. This finding was also supported by Lehto et al. [24], who suggested that most of the dentists did not find as much time as they would like for exercise, a key strategy for reducing dentists' musculoskeletal symptoms. This finding was also supported by Newell and Kumar [25], who reported that using proper body posture, incorporating regular rest breaks, maintaining good general health, performing exercises, and positioning during work can reduce the risk of disorders of MSDs. The study is based on the self-reporting by dentists. However, the study allowed for a general evaluation of the musculoskeletal symptoms among dentists. In this study, the sample size was limited, with only 124 subjects and 124 references being recruited from 17 community hospitals in Nakhon Si Thammarat province in South of Thailand. Therefore, the results in this study may not be readily comparable beyond this sample group in Thailand. 5. Conclusion In conclusion, musculoskeletal problems among dentists appear to increase with the amount of work they do, and targeted physical exercise appears to counter these problems to some degree. The prevention and reduction of MSDs among dentists should include improving their education in dental ergonomics and increasing their awareness regarding the importance of work related risk factors. Acknowledgments The authors would like to thank dental health workers at the community hospitals in Nakhon Si Thammarat Province, Thailand. This study was supported by grant from the Faculty of Health and Sports Science, Thaksin University. Ethical Approval This study was approved by the Ethical Committee of Thaksin University Review Board. Consent All of the participants received a clear explanation of the purpose of this study and agreed to participate using signed consent forms. Competing Interests No potential competing interests relevant to this paper were reported. Table 1 Occurrence (percent) of musculoskeletal symptoms that interfered with daily activities experienced by dental health workers in the previous 1 year by bodily location and position. Musculoskeletal symptoms Count (%) (n = 248) Dental health workers (n = 124) Reference group (n = 124) P value Trunk          Neck 58 (23.4) 48 (38.7) 10 (8.1) <0.001∗  Upper back 49 (19.8) 26 (21.0) 23 (18.6) 0.115  Lower back 57 (22.9) 39 (31.5) 18 (14.5) <0.001∗ Arms          Shoulders 68 (27.4) 59 (47.5) 9 (7.3) <0.001∗  Elbows 28 (11.3) 15 (12.1) 13 (10.5) 0.210  Hands/wrists 35 (14.1) 24 (19.4) 11 (8.9) <0.001∗ Lower body          Hips/thighs 25 (10.1) 5 (4.0) 20 (16.1) <0.001∗  Knees 32 (12.9) 30 (24.2) 2 (1.6) <0.001∗  Ankles/feet 19 (7.7) 11 (8.9) 8 (6.5) 0.250 ∗Significantly associated at P value of <0.05. Table 2 Prevalence (percent) of musculoskeletal symptoms that interfered with daily activities experienced by dental health workers in the previous 1 year, by bodily location and gender. Musculoskeletal symptoms Count (%) (n = 124) Gender (n, %) P value Male (n = 30) Female (n = 94) Trunk          Neck 48 (38.7) 25 (83.3) 23 (24.5) <0.001∗  Upper back 26 (21.0) 5 (16.7) 21 (22.3) 0.102  Lower back 39 (31.5) 15 (50.0) 24 (25.5) <0.001∗ Arms          Shoulders 59 (47.5) 27 (90.0) 32 (34.1) <0.001∗  Elbows 15 (12.1) 2 (6.7) 13 (13.8)    Hands/wrists 24 (19.4) 5 (16.7) 19 (20.2) 0.115 Lower body          Hips/thighs 5 (4.0) 2 (6.7) 3 (3.2) 0.059  Knees 30 (24.2) 5 (16.7) 25 (26.7) 0.023∗  Ankles/feet 11 (8.9) 3 (10.0) 8 (8.5) 0.254 ∗Significantly associated at P value of <0.05. Table 3 Occurrence (percent) of musculoskeletal symptoms that interfered with daily activities experienced by dental health workers in the previous 1 year by bodily location and duration of work (yrs). Musculoskeletal symptoms Count (%) (n = 124) Duration of work (yrs) (n, %) P value Hours worked per day (n, %) P value ≤5 (n = 10) >5 (n = 114) ≤8 (n = 45) >8 (n = 79) Trunk                Neck 48 (38.7) 1 (10.0) 47 (41.2) <0.001∗ 1 (2.2) 47 (59.5) <0.001∗  Upper back 26 (21.0) 2 (20.0) 24 (21.1) 0.257 2 (4.5) 24 (30.4) <0.001∗  Lower back 39 (31.5) 1 (10.0) 38 (33.3) <0.001∗ 5 (11.1) 34 (43.0) <0.001∗ Arms                Shoulders 59 (47.5) 2 (20.0) 57 (50.0) <0.001∗ 6 (13.3) 53 (67.1) <0.001∗  Elbows 15 (12.1) 2 (20.0) 13 (11.4) 0.057 0 15 (19.0) <0.001∗  Hands/wrists 24 (19.4) 2 (20.0) 12 (10.5) 0.051 2 (4.5) 22 (27.9) <0.001∗ Lower body                Hips/thighs 5 (4.0) 0 5 (4.4) 0.551 0 5 (6.3) 0.053  Knees 30 (24.2) 1 (10.0) 29 (25.4) <0.001∗ 2 (4.5) 28 (34.5) <0.001∗  Ankles/feet 11 (8.9) 0 11 (9.7) <0.001∗ 0 11 (13.9) <0.001∗ ∗Significantly associated at P value of <0.05. Table 4 Incidence (percent) of musculoskeletal symptoms that interfered with daily activities experienced by dental health workers in the previous 1 months by bodily location and days worked per week. Musculoskeletal symptoms Count (%) (n = 124) Days worked per week P value Exercise (n = 59) No exercise (n = 65) P value ≤6 (n = 58) >6 (n = 66) Trunk                Neck 48 (38.7) 2 (3.5) 46 (67.8) <0.001∗ 3 (5.1) 45 (69.2) <0.001∗  Upper back 26 (21.0) 12 (20.7) 14 (21.2) 0.524 5 (8.4) 21 (32.3) <0.001∗  Lower back 39 (31.5) 8 (13.8) 31 (47.0) <0.001∗ 17 (28.8) 22 (33.9) 0.057 Arms                Shoulders 59 (47.5) 2 (3.5) 57 (86.4) <0.001∗ 12 (20.3) 47 (72.3) <0.001∗  Elbows 15 (12.1) 7 (12.1) 8 (12.1) 0.624 7 (11.9) 8 (12.3) 0.078  Hands/wrists 24 (19.4) 11 (19.0) 13 (19.7) 0.824 12 (20.3) 12 (18.5) 0.056 Lower body                Hips/thighs 5 (4.0) 0 5 (7.58) <0.001∗ 0 5 (7.7) <0.001∗  Knees 30 (24.2) 11 (19.0) 19 (28.8) 0.042 2 (3.4) 28 (43.1) <0.001∗  Ankles/feet 11 (8.9) 5 (8.6) 6 (9.1) 0.512 0 11 (16.9) <0.001∗ ∗Significantly associated at P value of <0.05. ==== Refs 1 Bureau of Dental Healt Personnel Report Oral Health in 2551 2014 http://www.anamai.ecgates.com/ 2 Ayers K. M. S. Thomson W. M. Newton J. T. Morgaine K. C. Rich A. M. Self-reported occupational health of general dental practitioners Occupational Medicine 2009 59 3 142 148 10.1093/occmed/kqp004 2-s2.0-65549135305 19223433 3 Kierklo A. Kobus A. Jaworska M. Botuliński B. Work-related musculoskeletal disorders among dentists—a questionnaire survey Annals of Agricultural and Environmental Medicine 2011 18 1 79 84 2-s2.0-79960371457 21736272 4 Leggat P. A. Kedjarune U. Smith D. R. Occupational health problems in modern dentistry: a review Industrial Health 2007 45 5 611 621 10.2486/indhealth.45.611 2-s2.0-36749072747 18057804 5 Samotoi A. Moffat S. M. Thomson W. M. Musculoskeletal symptoms in New Zealand dental therapists: prevalence and associated disability New Zealand Dental Journal 2008 104 2 49 53 2-s2.0-51549090661 18672829 6 Valachi B. Valachi K. Mechanisms leading to musculoskeletal disorders in dentistry Journal of the American Dental Association 2003 134 10 1344 1350 10.14219/jada.archive.2003.0048 2-s2.0-0347415643 14620013 7 Biller F. E. Occupational hazards in dental practice Oral Hygiene 1994 36 1994 1999 8 Finsen L. Christensen H. Bakke M. Musculoskeletal disorders among dentists and variation in dental work Applied Ergonomics 1998 29 2 119 125 10.1016/S0003-6870(97)00017-3 2-s2.0-0032056030 9763237 9 Chowanadisai S. Kukiattrakoon B. Yapong B. Kedjarune U. Leggat P. A. Occupational health problems of dentists in southern Thailand International Dental Journal 2000 50 1 36 40 10.1111/j.1875-595X.2000.tb00544.x 2-s2.0-0034135930 10945178 10 Nithithamthada R. A. Puntumetakul C. Prevalence of musculoskeletal disorders among dental personal in Khon Kaen province, Thailand Occupational and Environmental Medicine 2013 70 supplement 1 A1 A149 24171196 11 Åkesson I. Schutz A. Horstmann V. Moritz U. Musculoskeletal symptoms among dental personnel;—lack of association with mercury and selenium status, overweight and smoking Swedish Dental Journal 2000 24 1-2 24 38 2-s2.0-0033654934 12 Alexopoulos E. C. Stathi I.-C. Charizani F. Prevalence of musculoskeletal disorders in dentists BMC Musculoskeletal Disorders 2004 5, article 16 10.1186/1471-2474-5-16 2-s2.0-13444309408 13 Myers H. L. Myers L. B. ‘It's difficult being a dentist’: stress and health in the general dental practitioner British Dental Journal 2004 197 2 89 93 10.1038/sj.bdj.4811476 2-s2.0-4143079133 15272347 14 Ratzon N. Z. Yaros T. Mizlik A. Kanner T. Musculoskeletal symptoms among dentists in relation to work posture Work 2000 15 3 153 158 2-s2.0-0033791194 12441484 15 Simning A. Van Wijngaarden E. Literature review of cancer mortality and incidence among dentists Occupational and Environmental Medicine 2007 64 7 432 438 10.1136/oem.2006.029223 2-s2.0-34347398705 17259166 16 Leggat P. A. Smith D. R. Musculoskeletal disorders self-reported by dentists in Queensland, Australia Australian Dental Journal 2006 51 4 324 327 10.1111/j.1834-7819.2006.tb00451.x 2-s2.0-33846146657 17256307 17 Szymanska J. Disorders of the musculoskeletal system among dentists from the aspect of egronomics and prophylaxis Annals of Agricultural and Environmental Medicine 2002 9 169 173 12498585 18 Lund A. E. How do you rate your general health? The Journal of the American Dental Association 2002 133 11 p. 1478 2-s2.0-0036836244 19 Lalumandier J. A. McPhee S. D. Prevalence and risk factors of hand problems and carpal tunnel syndrome among dental hygienists Journal of Dental Hygiene 2001 75 2 130 134 2-s2.0-0035293588 11475758 20 Shehab D. Al-Jarallah K. Moussa M. A. A. Adham N. Prevalence of low back pain among physical therapists in Kuwait Medical Principles and Practice 2003 12 4 224 230 10.1159/000072288 2-s2.0-0345600225 12966194 21 Adegoke B. O. A. Akodu A. K. Oyeyemi A. L. Work-related musculoskeletal disorders among Nigerian Physiotherapists BMC Musculoskeletal Disorders 2008 9 1, article 112 10.1186/1471-2474-9-112 2-s2.0-51649107621 22 Khan S. A. Chew K. Y. Effect of working characteristics and taught ergonomics on the prevalence of musculoskeletal disorders amongst dental students BMC Musculoskeletal Disorders 2013 14, article 118 10.1186/1471-2474-14-118 2-s2.0-84875546788 23 Al Wazzan K. A. Almas K. Al Shethri S. E. Al-Qahtani M. Q. Back & neck problems among dentists and dental auxiliaries The Journal of Contemporary Dental Practice 2001 2 3 17 30 2-s2.0-0035882802 12167924 24 Lehto T. U. Helenius H. Y. Alaranta H. T. Musculoskeletal symptoms of dentists assessed by a multidisciplinary approach Community Dentistry and Oral Epidemiology 1991 19 1 38 44 10.1111/j.1600-0528.1991.tb00103.x 2-s2.0-0026110116 2019089 25 Newell T. M. Kumar S. Prevalence of musculoskeletal disorders among orthodontists in Alberta International Journal of Industrial Ergonomics 2004 33 2 99 107 10.1016/j.ergon.2003.06.003 2-s2.0-0347479346

PMC005xxxxxx/PMC5002297.txt

==== Front Oxid Med Cell LongevOxid Med Cell LongevOMCLOxidative Medicine and Cellular Longevity1942-09001942-0994Hindawi Publishing Corporation 10.1155/2016/8474303Research ArticleParkin Protects against Oxygen-Glucose Deprivation/Reperfusion Insult by Promoting Drp1 Degradation Tang Jiayu 1 2 Hu Zhiping 1 Tan Jieqiong 3 Yang Sonlin 2 http://orcid.org/0000-0002-8276-6194Zeng Liuwang 1 * 1Department of Neurology, Second Xiangya Hospital, Central South University, Changsha, Hunan 410011, China2Department of Neurology, The Second People's Hospital of Hunan Province, Changsha, Hunan 410007, China3National Key Laboratory of Medical Genetics, Central South University, Changsha, Hunan 410078, China*Liuwang Zeng: xyneurology@hotmail.comAcademic Editor: Maria G. Isaguliants 2016 14 8 2016 2016 847430327 2 2016 13 6 2016 11 7 2016 Copyright © 2016 Jiayu Tang et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Ischemic stroke results in severe brain damage and remains one of the leading causes of death and disability worldwide. Effective neuroprotective therapies are needed to reduce brain damage resulting from ischemic stroke. Mitochondria are crucial for cellular energy production and homeostasis. Modulation of mitochondrial function mediates neuroprotection against ischemic brain damage. Dynamin-related protein 1 (Drp1) and parkin play a key role in regulating mitochondrial dynamics. They are potential therapeutic targets for neuroprotection in ischemic stroke. Protective effects of parkin-Drp1 pathway on mitochondria were assessed in a cellular ischemia-reperfusion injury model. Mouse neuroblastoma Neuro2a (N2a) cells were subjected to oxygen-glucose deprivation/reperfusion (OGDR) insult. OGDR induces mitochondrial fragmentation. The expression of Drp1 protein is increased after OGDR insult, while the parkin protein level is decreased. The altered protein level of Drp1 after OGDR injury is mediated by parkin through ubiquitin proteasome system (UPS). Drp1 depletion protects against OGDR induced mitochondrial damage and apoptosis. Meanwhile, parkin overexpression protects against OGDR induced apoptosis and mitochondrial dysfunction, which is attenuated by increased expression of Drp1. Our data demonstrate that parkin protects against OGDR insult through promoting degradation of Drp1. This neuroprotective potential of parkin-Drp1 pathway against OGDR insult will pave the way for developing novel neuroprotective agents for cerebral ischemia-reperfusion related disorders. National Natural Science Foundation of China812010193150083281471335 ==== Body 1. Introduction Mitochondria, the power house of the cell, participate in many essential cellular functions, including energy production, ion homeostasis, inflammation, apoptotic cell death, and calcium signaling. Change in mitochondrial mass and function has been linked with multiple diseases, including cerebral ischemia. Mitochondrial dysfunction is the most fundamental mechanism of cell damage in cerebral ischemia-reperfusion injury, which involves multiple independently fatal terminal pathways in the mitochondria. Modulation of mitochondrial function mediates neuroprotection against ischemic brain damage. Mitochondria are promising targets for stroke therapy [1, 2]. Mitochondrial homeostasis depends on their biogenesis and degradation. Parkin and dynamin-related protein 1 (Drp1) play a pivotal role in mitochondrial fission and clearance [3]. Parkin, the ubiquitin E3 ligase, has been shown to control the biogenesis and degradation of mitochondria. Parkin has also been suggested to ubiquitinate mitochondrial proteins, such as Drp1, to promote autophagy of damaged mitochondria [4]. Drp1 is required for mitochondrial division in mammalian cells. Changes in Drp1 expression directly influence cellular metabolism and ultimately cell fate. Drp1 is required for functionally active mitochondria, and supplementing with ATP can restore the defects induced by Drp1 suppression [5]. Drp1 is activated after cardiac arrest and the inhibition of Drp1 is protective against cerebral ischemic injury [6]. Parkin and Drp1 are novel therapeutic targets for cytoprotection. Therefore, on the basis of previous findings, we presumed that parkin and Drp1 would exert neuroprotective effect on cerebral ischemia/reperfusion that occurred in stroke. To address this, we employed oxygen-glucose deprivation and reperfusion (OGDR) model, which had been widely used in cultured neurons and brain slices to simulate brain ischemia. We found that Drp1 depletion protects against OGDR induced mitochondrial damage and apoptosis. Meanwhile, overexpression of parkin protects against OGDR induced apoptosis and mitochondrial dysfunction, which is blocked by upregulation of Drp1. Thus, parkin-Drp1 pathway represents a novel therapeutic target for treatment of a myriad of disorders related to cerebral ischemia-reperfusion injury. 2. Materials and Methods 2.1. Cells Culture and Transfection Mouse N2a neuroblastoma cells were purchased from American Type Culture Collection (ATCC). N2a neuroblastoma cells were used and maintained in Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% FBS (Gibco BRL), 100 U/mL penicillin, and 100 μg/mL streptomycin, at 37°C in a moist atmosphere containing 5% CO2. For overexpression of parkin and Drp1, cells were transfected with the indicated plasmids (pcDNA3.1-parkin-myc and pcDNA3.1-Drp1-HA) using Lipofectamine 2000 (Invitrogen) according to manufacturer's instructions. 2.2. Oxygen-Glucose Deprivation and Reperfusion (OGDR) To mimic ischemic-like conditions in vitro, cell cultures were exposed to oxygen-glucose deprivation (OGD) for 4 hours and then returned to 95% air, 5% CO2, and glucose-containing medium for different recovery time as before. First, mouse N2a neuroblastoma cells were transferred into a temperature controlled (37°C) anaerobic chamber (Forma Scientific) containing a gas mixture composed of 5% CO2 and 95% N2. The culture medium was replaced with deoxygenated glucose-free Hanks' Balanced Salt Solution (Invitrogen) and cells were maintained in the hypoxic chamber for 4 hours. After OGD, N2a cells were maintained in DMEM supplemented with 10% FBS under normoxic culture conditions for 0, 4, and 12 hours. 2.3. Immunofluorescence Staining Mouse N2a neuroblastoma cells were subjected to OGD for 4 hours followed by reperfusion for 0, 4, and 12 hours. The cells were then fixed with 4% paraformaldehyde for 30 min and washed three times with PBS, pH 7.4. The cells were incubated with a primary rabbit anti-Tom20 antibody (1 : 200; SC-11415, Santa Cruz Biotechnology) overnight at 4°C. On the following day, the cells were incubated with fluorescein-conjugated anti-rabbit IgG (1 : 400; FI-1000, Vector Laboratories) for 1 h. N2a cells were counterstained with 1 μg/mL 4′,6-diamidino-2-phenylindole (DAPI) (Vector Laboratories, Burlingame, CA, USA) to visualize nuclear morphology. Slides were washed, wet-mounted, and observed using a confocal microscope (Zeiss LSM Meta710, 63x/1.4 objective). Mitochondria fragmentation was assessed visually and quantified as percentage of control. Cells that displayed a network of filamentous mitochondria were classified as normal. Cells with fragmented or partially reelongated mitochondria were classified as fragmented [7]. 2.4. Western Blot Analysis Total protein was isolated from the N2a cells using 2x SDS sample buffer (63 mM Tris-HCl, 10% glycerol, and 2% SDS). Samples (20–40 μg of protein) were electrophoresed onto a 10–15% SDS/polyacrylamide gel (SDS/PAGE) and transferred to PVDF membranes. The membranes were blocked in TBS-Tween buffer containing 20 mM Tris-HCl, 5% nonfat milk, 150 mM NaCl, and 0.05% Tween-20 (pH 7.5) for 1 hour at room temperature. Thereafter, the blot was incubated with primary parkin mouse monoclonal antibody (1 : 1000; P-6248, Sigma-Aldrich), Drp1 rabbit monoclonal antibody (1 : 1000, #8570, cell signaling), Myc-tag mouse monoclonal antibody (1 : 1000; ab18185, Abcam), and actin mouse monoclonal antibody (1 : 10,000; SC-47778, Santa Cruz) for 1-2 hours at room temperature. The membrane was washed with TBST 3 times at 10-minute intervals, incubated with the secondary antibody (1 : 5000; anti-rabbit or anti-mouse IgG conjugated with horseradish peroxidase; Jackson ImmunoResearch Laboratories) at room temperature for 1 hour, and then washed 3 times each at 10-minute interval with TBST and 2 times each for 10 minutes with TBS. Band was visualized via an enhanced chemiluminescence kit (ECL) according to the manufacturer's suggested protocol (GE Health). Membranes were then exposed to X-ray film. 2.5. Quantitative Real-Time PCR Total RNA was isolated from N2a cells by using TriZol (Invitrogen). Reverse transcription was performed by using the Reverse Transcription Kit (Promega). Equal amounts of total RNA (500 ng) were reverse-transcribed. Primers for Drp1 and actin were used as follows: Drp1 forward primer, 5′-ACAGGAGAAGAAAATGGAGTTTGAAGCAG-3′, Drp1 reverse primer, 5′-AACAAATCCTAGCACCACGCAT-3′, actin forward primer 5′-AGGCACCAGGGCGTGAT-3′, and actin reverse primer, 5′-GCCCACATAGGAATCCTTCTGAC-3′. Quantitative real-time PCRs were then conducted at 95°C for 10 s, followed by 40 cycles of 95°C for 5 seconds and 60°C for 30 s. The relative RNA levels were normalized to endogenous actin expression for each sample. The PCR experiments were repeated 3 times, each using separate sets of cultures. 2.6. siRNA-Mediated Parkin/Drp1 Knockdown Drp1 siRNA and control siRNA were used for knocking down protein expression (sequence: siDrp1: 5′-AAGCAGAAGAAUGGGGUAAAUdTdT-3′ and siparkin: 5′-UUCCAAACCGGAUGAGUGGdTdT-3′). The N2a neuroblastoma cells were transiently transfected with siRNA against Drp1 or parkin synthesized by GenePharma using Lipofectamine 2000 reagent (Invitrogen) before OGD. The scrambled sequence was used as a negative control. Lipofectamine 2000 (Invitrogen, Germany) and siRNA were dissolved separately in Optimem I (Invitrogen). After standing at room temperature for 10 min, each RNA solution was added with the Lipofectamine 2000 and mixed for 20 min gently allowing formation of siRNA liposomes at room temperature. Then the mixture was added to the DMEM (antibiotic-free) to a final concentration of 60 nmol and 2 μL/mL Lipofectamine. Protein expression was determined after 48 hours of transfection. The OGD was performed 48 h after transfection. 2.7. Measurement of Apoptosis Mouse N2a neuroblastoma cells were subjected to OGD for 4 hours followed by reperfusion for 12 hours. Apoptosis was detected by Annexin V FITC Apoptosis Detection Kit (Sigma). Briefly, N2a cells were collected and washed twice with PBS. 500 μL binding buffer suspension was then added to the treated cells. After that, 5 μL Annexin V-FITC and 10 μL propidium iodide were added to each group and cultures were incubated at 37°C for 5~15 min in dark. Flow cytometer (BD Biosciences) was used to detect the percent cells with apoptosis and flowJo software was used for flow cytometry analysis. 2.8. Cytochrome c Oxidase (COX) Activity Assays Mitochondria of N2a cells were isolated using the animal cells/tissues mitochondria extraction kit (Genmed Scientifics, Cat. GSM10006). Cytochrome c oxidase activity was detected using cytochrome c oxidase assay kit (Sciencell, Cat. 8278). All procedures followed the manufacturer's instructions. Briefly, mitochondria were suspended in 50 μL of isolation buffer containing 250 mM sucrose, 20 mM HEPES, pH 7.2, and 1 mM EDTA. Suspensions were added to a cuvette containing 0.95 mL of 1x assay buffer (10 mM Tris-HCl, pH 7.0, and 120 mM KCl), and the reaction volume was brought to 1.05 mL with 1x enzyme dilution buffer (10 mM Tris-HCl, pH 7.0). The reaction was then started by the addition of 50 μL of cytochrome c substrate solution; the change in absorbance of cytochrome c at 550 nm was measured using a DU640 spectrophotometer. The reading was recorded every 5 sec during the first 3 minutes. Background levels were measured without cell suspensions. All measurements were repeated at least three times. All normalized control of cytochrome c oxidase activity values will be 1 (or 100%). 2.9. Determination of Mitochondrial ATP Synthase Activity Mitochondrial ATP synthase activity was detected as follows. Mitochondria of N2a cells were isolated using the animal cells/tissues mitochondria extraction kit (Genmed Scientifics, Cat. GSM10006). Ten micrograms of mitochondria was used to detect ATP synthase activity by using the mitochondrial complex V activity quantitative analysis kit (Genmed Scientifics, Cat. GMS50083). All procedures followed the manufacturer's instructions. Briefly, the ADP generated by complex V reacts with phosphoenolpyruvic acid (1 mM) to pyruvate in the presence of pyruvate kinase (1.5 U/mL). Pyruvate oxidizes with NADH (250 μM) to lactate and NAD+ in the presence of lactate dehydrogenase (2.5 U/mL). This reaction was monitored spectrophotometrically at 340 nm for 8 min and used to determine complex V activity. All measurements were repeated at least three times. All normalized control of complex V activities values will be 1 (or 100%). 2.10. Statistical Analysis Quantitative data were expressed as mean ± SEM based on at least 3 separate experiments of triplicate samples. Differences among groups were statistically analyzed by one-way analysis of variance followed by Bonferroni's post hoc test. Comparison between two experimental groups was based on a two-tailed t-test. Differences between the mean values were considered significant if p < 0.05. 3. Results 3.1. OGDR Induces Mitochondrial Fragmentation in N2a Cells To explore whether mitochondrial fragmentation occurs in N2a cells upon OGDR insult, we used immunofluorescent staining to evaluate its temporal profiles (Figure 1). The increase of mitochondrial fragmentation in a time-dependent manner was found during the different time points of OGDR. As demonstrated in Figure 1(a), most of the cells displayed tubular and long mitochondria in normal conditions, indicating a balance between mitochondrial fusion and fission. After 4 h of OGD treatment, most of the cells still showed tubular and long mitochondria. However, after 4 and 12 h reperfusion following 4 h of OGD, the morphology of mitochondria changed to debris-like structures scattered in the cytoplasm. The increase of N2a cells with fragmented mitochondria began as early as 4 h reperfusion following 4 h OGD exposure and was further enhanced after 12 h reperfusion (Figure 1(b)). 3.2. Expression Pattern of Drp1 and Parkin Protein in N2a Cells upon OGDR Insult To elucidate the functional significance of Drp1 and parkin in mouse N2a cells upon OGDR challenge, we first determined the expression profile of Drp1 and parkin protein at different reperfusion time points following OGD insults. Western blot analysis showed similar expression of Drp1 between normal mouse N2a cells and N2a cells after 4 h of OGD. However, expression of Drp1 was strongly increased in mouse N2a cells after 4 and 12 h reperfusion following 4 h of OGD (Figures 1(c) and 1(d)). Meanwhile, western blot analysis demonstrated that the protein level of parkin was decreased in mouse N2a cells after OGDR insult, which was most significant after 12 h reperfusion following 4 h of OGD (Figures 1(c) and 1(e)). 3.3. OGDR Regulates the Expression of Drp1 Protein via Inhibition of Ubiquitin Proteasome System We next examined whether OGDR regulated Drp1 at the mRNA level in mouse N2a cells. Quantitative real-time PCR demonstrated that Drp1 transcription was unchanged in normal mouse N2a cells and N2a cells subjected to 0, 4, and 12 h reperfusion following 4 h of OGD, indicating that OGDR regulating the expression of Drp1 is not transcriptional (Figure 2(a)). We next used the proteasome inhibitors MG132 (20 μM for 4 h) to determine whether the OGDR-mediated increase in Drp1 protein expression was due to inhibition of ubiquitin proteasome system (UPS). We found that the expression of Drp1 protein was significantly increased after treatment with the MG132 (Figure 2(b)). Taken together, these data demonstrate that OGDR induces Drp1 expression through the proteasome-dependent proteolytic machinery. 3.4. Parkin Promotes Drp1 Degradation after OGDR Insult Parkin, an E3 ubiquitin ligase, can regulate the ubiquitination and proteasome-dependent degradation of its substrates. The Drp1 protein level has been reported to be regulated by the E3 ubiquitin ligase, parkin [8, 9]. We further investigated the regulatory effect of parkin on Drp1 expression after OGDR insult. siRNA-mediated knockdown of parkin expression was performed in mouse N2a cells. We found that Drp1 protein expression was significantly increased after knockdown of parkin (Figures 2(c) and 2(d)), suggesting that parkin induces Drp1 degradation after OGDR injury. 3.5. Knockdown of Drp1 Protects against OGDR Induced Mitochondrial Damage and Apoptosis To further analyze the role of Drp1 in ischemic injury, we examined the effect of Drp1 knockdown with a specific siRNA on OGDR induced mitochondrial damage and apoptosis in N2a cells. After being transfected with siRNA against Drp1, N2a cells were treated with OGD 4 h plus 12 h reperfusion. The RNA interference of Drp1 significantly reduced Drp1 protein levels in N2a cells (Figure 3(b)). Results showed that, in N2a cell without OGDR exposure, N2a cell displayed typical tubular and long mitochondria (Figure 3(a)). When N2a cell transfected with control siRNA is exposed to OGD 4 h plus 12 h reperfusion, lots of mitochondria changed to debris-like structures scattered in the cytoplasm. However, transfection with Drp1 siRNA significantly attenuated OGDR induced mitochondrial fragmentation (Figure 3(a)). Quantitative analysis demonstrated that there were a higher proportion of longer and tubular mitochondria in the cells treated with Drp1 siRNA compared with the vehicle-treated cells (Figure 3(c)). Additionally, the proportion of apoptotic cells was increased after OGD 4 h plus 12 h reperfusion. However, N2a cells transfected with Drp1 siRNA displayed a significant decrease in the number of apoptotic cells after OGDR exposure (Figures 3(d) and 3(e)). Meanwhile, COX activity and mitochondrial ATP synthase activity were also measured in N2a cells to provide an assessment of mitochondrial function. The COX activity was decreased after OGDR injury. However, Drp1 depletion significantly attenuated OGDR induced decrease in activity of cytochrome c oxidase (Figure 3(f)). In addition, mitochondrial ATP synthase activity was measured to investigate the effect of Drp1 depletion on mitochondrial ATP generation in N2a cells. Mitochondrial ATP synthase activity (Figure 3(g)) was inhibited by OGDR treatment. However, ATP synthase activity was improved after being transfected with Drp1 siRNA. Together, these data strongly suggest that OGDR induced mitochondrial damage and apoptosis were markedly reversed by Drp1 knockdown. 3.6. Parkin Protects against OGDR Induced Apoptosis and Mitochondrial Dysfunction by Promoting Drp1 Degradation We have demonstrated that parkin knockdown induced Drp1 expression after OGDR injury. To assess the role of parkin in the regulation of apoptosis and mitochondrial function after OGDR insult, we increased parkin expression in mouse N2a cells with specific expression plasmids for parkin with a Myc-tag. Parkin overexpression resulted in decreased expression of Drp1 protein (Figure 4(a)). Strikingly, enhanced expression of parkin was able to suppress OGDR induced apoptosis. However, parkin overexpression mediated protection against OGDR induced apoptosis was attenuated by increased expression of Drp1 (Figures 4(b) and 4(c)). Additionally, our results also demonstrated that the decreased activities of cytochrome c oxidase and mitochondrial ATP synthase after OGDR insult were improved after enhanced expression of parkin, which was abrogated by Drp1 overexpression (Figures 4(d) and 4(e), resp.). Therefore, our data suggest that the neuroprotective effect of parkin in OGDR insult is mediated by promoting Drp1 degradation. 4. Discussion Ischemic stroke remains one of the leading causes of death and disability worldwide. Cerebral ischemia and subsequent reperfusion result in cellular organelles damage, which plays a pivotal role in the development of tissue injury after acute ischemic stroke. Ischemia-reperfusion insult generally leads to mitochondrial dysfunction and fragmentation [10, 11]. Consistent with previous finding, we also found mitochondrial fragmentation in N2a cells after 4 h and 12 h reperfusion following 4 h OGD exposure. Our data suggest that mitochondria are sensitive to OGDR insult. OGDR induced mitochondrial fragmentation would lead to mitochondrial dysfunction, such as failure in the energy production for cell homeostasis and cell survival in N2a cells. Mitochondrial dysfunction after OGDR injury is one of the major contributors to ischemia-reperfusion damage. Many agents confer neuroprotective effects by targeting mitochondria [12, 13]. Therefore, developing novel therapeutic strategies targeting mitochondrial pathway will open a new avenue for treating cerebral ischemia. Moreover, OGDR insult not only induces mitochondrial fragmentation, but also regulates Drp1 expression. We found that the protein level of Drp1 was strongly upregulated after ODGR injury, while its transcription was unchanged, indicating that OGDR regulates the expression of Drp1 via a nontranscriptional mechanism. Mitochondrial morphology depends on the balance between two opposing processes, mitochondrial fission and fusion. Mitochondrial fission is mediated predominantly by a dynamin-related GTPase, Drp1 [14]. Thus, induced expression of Drp1 after OGDR injury could influence mitochondrial fission and functional activity in N2a cells. Drp1 upregulation would also influence cellular metabolism and ultimately cell fate after OGDR insult. In neurons, Drp1 controls the migration and neuronal differentiation of subventricular zone-derived neural progenitor cells [5]. Drp1-mediated mitochondrial fission plays a major role in neuronal cell death associated with acute ischemic brain damage [15]. Therefore, the changed expression of Drp1 may play a key role in the pathophysiology of OGDR insult. In our study, we also demonstrated that OGDR insult inhibits parkin protein expression in mouse N2a cells, which was more remarkable after prolonged reperfusion period following 4 h of OGD. Parkin is a widely expressed 465 amino acid protein that possesses E3 ubiquitin-protein ligase activity [16]. Parkin plays a critical role in the pathogenesis of Parkinson's disease (PD) [17]. Loss-of-function mutations in the parkin gene have been revealed to be the most common cause of autosomal recessive, early-onset PD [18, 19]. For ischemic stroke, expression of parkin was also decreased in transient focal cerebral ischemia [20], which is in line with our results. Parkin exerts various roles in neurons and is widely thought to be protective. Decreased expression of parkin would contribute to mitochondria injury and tissue damage after OGDR insult. Mutations in parkin would result in ubiquitin proteasome system (UPS) dysfunction [21]. Parkin protein depletion after cerebral ischemia could increase the aggregation of ubiquitylated proteins and the sensitivity of neurons to endoplasmic reticulum (ER) dysfunction [20]. OGDR induced parkin inhibition would cause dysfunction of UPS and influence degradation of ubiquitylated proteins. Decreased parkin protein expression may be of great importance in the pathological processes culminating in neuronal cell injury after OGDR insult and ischemic stroke therapy. Additionally, we found that expression of Drp1 protein was significantly enhanced after proteasome inhibitor MG132 treatment, indicating that OGDR-mediated increase in Drp1 protein expression is due to the inhibition of UPS. The UPS plays an important role in protein quality control, growth and inflammatory cell signaling, and neuron excitability. It has emerged as a drug target for diverse diseases characterized by altered proteostasis [22]. Thus, UPS might also be involved in OGDR insult induced cerebral damage. Meanwhile, we demonstrated that Drp1 protein expression was increased after parkin depletion after OGDR injury. Drp1 is a substrate of parkin. Parkin interacts with and subsequently ubiquitinates Drp1, for promoting its proteasome-dependent degradation [8]. Activation of the UPS for proteolysis of multiple outer membrane proteins is supposed to be a major function of parkin in dysfunctional mitochondria, which is a critical step in parkin-mediated mitophagy [23]. Therefore, parkin is involved in elimination of Drp1 after OGDR injury. Parkin uses the UPS to degrade and ship off Drp1 and consequently contribute to regulation of cellular signaling pathways activated by OGDR insult. In this study, we also demonstrated that knockdown of Drp1 by siRNA protects against OGDR induced mitochondrial damage and apoptosis. Drp1 interacts with mutant proteins of neurodegenerative diseases and then fragments mitochondria excessively, ultimately causing neuronal damage [24]. Therefore, Drp1 might also contribute to mitochondrial dysfunction and cell damage in N2a cells through interaction with abnormal protein after OGDR injury. Inhibition of Drp1 has been demonstrated to confer protection in various stress situations. It attenuates mitochondrial fragmentation in cisplatin-induced nephrotoxicity [25], protects the heart against ischemia/reperfusion injury [26], increases retinal ganglion cell survival in acute ischemic mouse retina  [27], provides neuroprotection against glutamate toxicity ischemic brain damage [15, 28], determines the fate of cells after irradiation [29], and protects against chlorpyrifos- (CPF-) induced cytotoxicity [30]. In line with these reports, our data suggest that Drp1 is critical for the morphology and function of mitochondria in N2a cells after OGDR insult. Drp1 inhibition exerts protective effects and is a potential therapeutic target in OGDR injury. Moreover, our data suggests that overexpression of parkin protects against OGDR induced apoptosis and mitochondrial dysfunction, indicating that parkin is protective in OGDR insult. Parkin protects against misfolded SOD1 toxicity by promoting its aggresome formation and autophagic clearance in amyotrophic lateral sclerosis (ALS) [31]. Parkin has also been shown to exert cytoprotective action in PD and AD by increasing the removal of cellular Abeta through a proteasome-dependent pathway or protecting against the toxicity associated with mutant alpha-synuclein [32, 33]. Therefore, parkin might also confer neuroprotection in OGDR insult by promoting degradation and clearance of toxic proteins through the UPS pathway. Drp1 is a substrate of parkin. We have demonstrated that parkin promotes Drp1 degradation after OGDR insult. Parkin interacts with and subsequently ubiquitinates Drp1, thus promoting its proteasome-dependent degradation [8]. We also found that Drp1 depletion protects against OGDR induced mitochondrial dysfunction and apoptosis. Based on these results, we presume that Drp1 is the downstream effector of parkin. We overexpressed parkin and Drp1 in N2a cells. We found that overexpression of parkin decreased Drp1 expression and cytoprotective action of parkin was abolished by increased expression of Drp1. These results suggest that the neuroprotective effect of parkin in OGDR insult is mediated by decreasing Drp1 expression. Therefore, targeting parkin-Drp1 pathway may provide a novel therapeutic strategy for cerebral ischemia-reperfusion. In conclusion, our study demonstrates that mitochondria are fragmented, and expression of Drp1 and parkin protein is altered after OGDR insult in mouse N2a cells. The regulated protein level of Drp1 after OGDR insult is mediated by parkin through the ubiquitin proteasome system. Knockdown of Drp1 protects against OGDR induced mitochondrial damage and apoptosis. Moreover, parkin protects against OGDR induced apoptosis and mitochondrial dysfunction by promoting degradation of Drp1, indicating its cytoprotective role in cerebral ischemia-reperfusion injury. Our results showed the beneficial effects of knockdown of Drp1 or overexpression of parkin against OGDR insult, which would pave the way for its potential clinical application. Further efforts may lead to the development of novel therapies for disorders whose etiology is based upon cerebral ischemia-reperfusion injury by targeting parkin-Drp1 pathway. Acknowledgments This paper has been supported by National Natural Science Foundation of China (Grants nos. 81201019, 31500832, and 81471335). Competing Interests The authors confirm that this paper content has no conflict of interests. Figure 1 OGDR affects mitochondrial morphology and the protein levels of Drp1 and parkin. (a) Mitochondrial morphology was analyzed with Tom20 staining by confocal microscopy in N2a cells. The confocal images and the enlarged section of the confocal images are displayed. (b) The average percentage of N2a cells with fragmented mitochondria was measured. (c) Western blot analysis and (d, e) quantitative analysis of Drp1 and parkin expressions in N2a cells upon OGDR insult. Actin was used as a loading control. Data are presented as the mean ± SEM. Asterisks indicate statistically significant difference compared with the control; ∗ p < 0.05, ∗∗ p < 0.01, and ∗∗∗ p < 0.001. Figure 2 OGDR affects Drp1 expression through ubiquitin proteasome system. (a) mRNA level of Drp1 in mouse N2a cells upon OGDR insult was assessed by quantitative real-time PCR. Actin mRNA was used as an internal control. (b) Mouse N2a cells were incubated in the absence or presence of MG132, a reversible proteasome inhibitor. Cell lysates were analyzed by western blot with antibodies against mitochondrial fission-specific Drp1. Actin was used as a loading control, and all blots are representative of three independent experiments. (c) N2a cells were transfected with parkin siRNA or control siRNA (Ctrl siRNA), and western blot was performed to examine the expression of parkin and Drp1 protein. (d) Drp1 protein level was quantified according to the results of three independent blots. Data represent the mean ± SEM. Asterisks indicate a statistically significant difference compared with the control. ∗ p < 0.05. Figure 3 The effect of Drp1 knockdown on mitochondria morphology, function, and apoptosis in N2a cells exposed to OGDR insult. After being transfected with siRNA against Drp1, cells were treated with OGD 4 h plus 12 h reperfusion. The experiment was repeated independently for at least three times. (a) Digital photomicrograph under fluorescent illumination showing the morphology of mitochondria was detected using Tom20 staining. The confocal images and the enlarged section of the confocal images are displayed. Most of N2a cells displayed typical tubular and long mitochondria in normal condition or after 4 h OGD. Fragmented mitochondria were evident in N2a cells exposed to 4 h OGD plus 12 h reperfusion. Transfection with siRNA against Drp1 significantly attenuated OGDR induced fragmentation of mitochondria. (b) Representative immunoblots of Drp1 showed knockdown of Drp1 by the specific Drp1 siRNA. (c) Quantitation (mean ± SEM) of A from three independent experiments. (d, e) Knockdown of Drp1 in vitro reduced OGDR induced apoptosis. (f, g) Cytochrome c oxidase and mitochondrial ATP synthase activity in N2a cells with or without Drp1 depletion after OGDR insult. Drp1 protein knockdown significantly improved cytochrome c oxidase activity and mitochondrial ATP synthase activity after OGDR insult. Values are expressed as mean ± SEM. ∗ p < 0.05 and ∗∗ p < 0.01 compared to control. Figure 4 Inhibition of Drp1 by parkin overexpression confers protection against OGDR induced apoptosis and mitochondrial dysfunction. (a) Mouse N2a cells were untransfected or transfected with parkin-myc or vector, and western blot was performed to examine the expression of Drp1 and parkin proteins. Overexpression of parkin contributed to decreased expression of Drp1. (b, c) Mouse N2a cells untransfected or transfected with parkin-myc and Drp1-HA were subjected to 4 h OGD plus 12 h reperfusion. Enhanced expression of parkin was able to suppress OGDR induced apoptosis, which was abrogated by increased expression of Drp1. (d, e) Parkin overexpression improved cytochrome c oxidase and mitochondrial ATP synthase activities after OGDR injury, which was also attenuated by increased expression of Drp1. All data are representative of at least three independent experiments. Values are expressed as mean ± SEM. ∗ p < 0.05 compared to control. ==== Refs 1 Baek S.-H. Noh A. R. Kim K.-A. Modulation of mitochondrial function and autophagy mediates carnosine neuroprotection against ischemic brain damage Stroke 2014 45 8 2438 2443 10.1161/strokeaha.114.005183 2-s2.0-84905398294 24938837 2 Silachev D. N. Plotnikov E. Y. Zorova L. D. Neuroprotective effects of mitochondria-targeted plastoquinone and thymoquinone in a rat model of brain ischemia/reperfusion injury Molecules 2015 20 8 14487 14503 10.3390/molecules200814487 2-s2.0-84941585580 26270657 3 Buhlman L. Damiano M. Bertolin G. Functional interplay between Parkin and Drp1 in mitochondrial fission and clearance Biochimica et Biophysica Acta (BBA)—Molecular Cell Research 2014 1843 9 2012 2026 10.1016/j.bbamcr.2014.05.012 2-s2.0-84902477543 24878071 4 Kageyama Y. Hoshijima M. Seo K. Parkin-independent mitophagy requires Drp1 and maintains the integrity of mammalian heart and brain EMBO Journal 2014 33 23 2798 2813 10.15252/embj.201488658 2-s2.0-84913565821 25349190 5 Kim H. J. Shaker M. R. Cho B. Dynamin-related protein 1 controls the migration and neuronal differentiation of subventricular zone-derived neural progenitor cells Scientific Reports 2015 5 15962 10.1038/srep15962 2-s2.0-84946105759 6 Li Y. Wang P. Wei J. Inhibition of Drp1 by Mdivi-1 attenuates cerebral ischemic injury via inhibition of the mitochondria-dependent apoptotic pathway after cardiac arrest Neuroscience 2015 311 67 74 10.1016/j.neuroscience.2015.10.020 26477985 7 Kamp F. Exner N. Lutz A. K. Inhibition of mitochondrial fusion by α -synuclein is rescued by PINK1, Parkin and DJ-1 The EMBO Journal 2010 29 20 3571 3589 10.1038/emboj.2010.223 2-s2.0-77958450202 20842103 8 Wang H. Song P. Du L. Parkin ubiquitinates Drp1 for proteasome-dependent degradation: implication of dysregulated mitochondrial dynamics in Parkinson disease The Journal of Biological Chemistry 2011 286 13 11649 11658 10.1074/jbc.m110.144238 2-s2.0-79953231682 21292769 9 Takamura H. Koyama Y. Matsuzaki S. TRAP1 controls mitochondrial fusion/fission balance through Drp1 and Mff expression PLoS ONE 2012 7 12 e51912 10.1371/journal.pone.0051912 2-s2.0-84871458524 10 Di Y. He Y.-L. Zhao T. Methylene blue reduces acute cerebral ischemic injury via the induction of mitophagy Molecular Medicine 2015 21 420 429 10.2119/molmed.2015.00038 2-s2.0-84940395578 25998511 11 Kim S. Y. Shim M. S. Kim K.-Y. Weinreb R. N. Wheeler L. A. Ju W.-K. Inhibition of cyclophilin D by cyclosporin A promotes retinal ganglion cell survival by preventing mitochondrial alteration in ischemic injury Cell Death and Disease 2014 5 3 e1105 10.1038/cddis.2014.80 2-s2.0-84897415969 12 Zeng L. Tan J. Lu W. Hu Z. Blockade of Ser16-Hsp20 phosphorylation attenuates neuroprotection dependent upon Bcl-2 and bax Current Neurovascular Research 2013 10 3 208 215 10.2174/15672026113109990006 2-s2.0-84880264929 23713735 13 Zhang X.-M. Zhang L. Wang G. Suppression of mitochondrial fission in experimental cerebral ischemia: the potential neuroprotective target of p38 MAPK inhibition Neurochemistry International 2015 90 1 8 10.1016/j.neuint.2015.06.010 2-s2.0-84947033848 26116440 14 Tamura Y. Itoh K. Sesaki H. SnapShot: mitochondrial dynamics Cell 2011 145 7 1158 1158.e1 10.1016/j.cell.2011.06.018 2-s2.0-79959632321 21703455 15 Grohm J. Kim S.-W. Mamrak U. Inhibition of Drp1 provides neuroprotection in vitro and in vivo Cell Death and Differentiation 2012 19 9 1446 1458 10.1038/cdd.2012.18 2-s2.0-84865187064 22388349 16 Shimura H. Hattori N. Kubo S.-I. Familial Parkinson disease gene product, parkin, is a ubiquitin-protein ligase Nature Genetics 2000 25 3 302 305 10.1038/77060 2-s2.0-0033933048 10888878 17 Cha S.-H. Choi Y. R. Heo C.-H. Loss of parkin promotes lipid rafts-dependent endocytosis through accumulating caveolin-1: implications for Parkinson's disease Molecular Neurodegeneration 2015 10, article 63 10.1186/s13024-015-0060-5 2-s2.0-84957431131 18 Puschmann A. Monogenic Parkinson's disease and parkinsonism: clinical phenotypes and frequencies of known mutations Parkinsonism and Related Disorders 2013 19 4 407 415 10.1016/j.parkreldis.2013.01.020 2-s2.0-84875277277 23462481 19 Kitada T. Asakawa S. Hattori N. Mutations in the parkin gene cause autosomal recessive juvenile parkinsonism Nature 1998 392 6676 605 608 10.1038/33416 2-s2.0-0032499264 9560156 20 Mengesdorf T. Jensen P. H. Mies G. Aufenberg C. Paschen W. Down-regulation of parkin protein in transient focal cerebral ischemia: a link between stroke and degenerative disease? Proceedings of the National Academy of Sciences of the United States of America 2002 99 23 15042 15047 10.1073/pnas.232588799 2-s2.0-0037069337 12415119 21 Cook C. Stetler C. Petrucelli L. Disruption of protein quality control in Parkinson's disease Cold Spring Harbor Perspectives in Medicine 2012 2 5 a009423 10.1101/cshperspect.a009423 2-s2.0-84869014351 22 Zhao W. Bonem M. McWhite C. Silberg J. J. Segatori L. Sensitive detection of proteasomal activation using the Deg-On mammalian synthetic gene circuit Nature Communications 2014 5, article 3612 10.1038/ncomms4612 2-s2.0-84908180743 23 Chan N. C. Chan D. C. Parkin uses the UPS to ship off dysfunctional mitochondria Autophagy 2011 7 7 771 772 10.4161/auto.7.7.15453 2-s2.0-79959951425 21460629 24 Reddy P. H. Reddy T. P. Manczak M. Calkins M. J. Shirendeb U. Mao P. Dynamin-related protein 1 and mitochondrial fragmentation in neurodegenerative diseases Brain Research Reviews 2011 67 1-2 103 118 10.1016/j.brainresrev.2010.11.004 2-s2.0-79957937433 21145355 25 Brooks C. Wei Q. Cho S.-G. Dong Z. Regulation of mitochondrial dynamics in acute kidney injury in cell culture and rodent models The Journal of Clinical Investigation 2009 119 5 1275 1285 10.1172/jci37829 2-s2.0-66449121454 19349686 26 Ong S.-B. Subrayan S. Lim S. Y. Yellon D. M. Davidson S. M. Hausenloy D. J. Inhibiting mitochondrial fission protects the heart against ischemia/reperfusion injury Circulation 2010 121 18 2012 2022 10.1161/CIRCULATIONAHA.109.906610 2-s2.0-77952236126 20421521 27 Park S. W. Kim K.-Y. Lindsey J. D. A selective inhibitor of Drp1, Mdivi-1, increases retinal ganglion cell survival in acute ischemic mouse retina Investigative Ophthalmology and Visual Science 2011 52 5 2837 2843 10.1167/iovs.09-5010 2-s2.0-79956048283 21372007 28 Zhao Y.-X. Cui M. Chen S.-F. Dong Q. Liu X.-Y. Amelioration of ischemic mitochondrial injury and bax-dependent outer membrane permeabilization by Mdivi-1 CNS Neuroscience and Therapeutics 2014 20 6 528 538 10.1111/cns.12266 2-s2.0-84900553387 24712408 29 Yamamori T. Ike S. Bo T. Inhibition of the mitochondrial fission protein dynamin-related protein 1 (Drp1) impairs mitochondrial fission and mitotic catastrophe after x-irradiation Molecular Biology of the Cell 2015 26 25 4607 4617 10.1091/mbc.E15-03-0181 2-s2.0-84951130958 26466676 30 Park J. H. Ko J. Hwang J. Koh H. C. Dynamin-related protein 1 mediates mitochondria-dependent apoptosis in chlorpyrifos-treated SH-SY5Y cells NeuroToxicology 2015 51 145 157 10.1016/j.neuro.2015.10.008 2-s2.0-84945567484 26598294 31 Yung C. Sha D. Li L. Chin L.-S. Parkin protects against misfolded SOD1 toxicity by promoting its aggresome formation and autophagic clearance Molecular Neurobiology 2015 10.1007/s12035-015-9537-z 2-s2.0-84946949475 32 Petrucelli L. O'Farrell C. Lockhart P. J. Parkin protects against the toxicity associated with mutant α -synuclein: proteasome dysfunction selectively affects catecholaminergic neurons Neuron 2002 36 6 1007 1019 10.1016/s0896-6273(02)01125-x 2-s2.0-0037137702 12495618 33 Rosen K. M. Moussa C. E.-H. Lee H.-K. Parkin reverses intracellular β -amyloid accumulation and its negative effects on proteasome function Journal of Neuroscience Research 2010 88 1 167 178 10.1002/jnr.22178 2-s2.0-73949083524 19610108

PMC005xxxxxx/PMC5002298.txt

==== Front Case Rep PediatrCase Rep PediatrCRIPECase Reports in Pediatrics2090-68032090-6811Hindawi Publishing Corporation 10.1155/2016/5208753Case ReportA Case of Congenital Malignant Spinal Cord Glioma as a Cause of Congenital Ascites in a Neonate http://orcid.org/0000-0003-0912-1645Karber Bianca * Omesi Lenore Chang Sunny Handel Andrew Hegedus Monica Maduekwe Echezona Stony Brook Children's Hospital, Stony Brook, NY 11794, USA*Bianca Karber: biancakarber@gmail.comAcademic Editor: Ozgur Cogulu 2016 14 8 2016 2016 52087532 2 2016 11 7 2016 Copyright © 2016 Bianca Karber et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Congenital ascites is rare, but when it occurs, urinary ascites secondary to posterior urethral valve obstruction is the most common, and tumors are the least. Among the tumors in the pediatric population, the central nervous system tumors are common, but spinal cord tumors are rare. We describe a very rare case of congenital malignant spinal cord glioma presenting as isolated congenital ascites secondary to neurogenic bladder. A female infant was diagnosed sonographically with isolated congenital ascites at 40 weeks' gestational age, with uneventful development prior to 40 weeks' gestational age. Magnetic resonance imaging of the spine done within the first week of life identified a lobulated spinal mass with heterogeneous enhancement within the conus medullaris. Spinal fluid analysis showed evidence of small round blue cells and the pathology from the excision biopsy of the mass confirmed a WHO grade III or IV malignant glioma. The postoperative course was uneventful with resolution of the ascites and spontaneous micturition. The patient was discharged home without an indwelling urinary catheter. We report the first documented case of a newborn infant with isolated congenital ascites from neurogenic bladder secondary to a spinal cord glioma. ==== Body 1. Introduction Tumors of the central nervous system are common in the pediatric population, constituting the second most prevalent tumor type in children. Spinal cord tumors account for 1–10% of all pediatric central nervous system tumors [1–3]. However, only 2% of childhood tumors have been reported to appear in the neonatal period [4]. Neonatal tumors are more often benign and rarely present with urinary ascites. We report a first documented case of a congenital malignant spinal cord glioma. 2. Case Report A full-term 3.8 kg female of 40 weeks' gestation was born via C-section for isolated fetal ascites seen on sonogram. The prenatal course history was significant for conception through in vitro fertilization using a donor egg, anhydramnios, and isolated congenital ascites seen on sonogram at 40 weeks' gestational age. This repeat sonogram was done due to failed induction three days prior to delivery, after having had a normal sonogram four weeks earlier. At delivery, the baby was tachypneic and had a tense shiny distended abdomen (see Figure 1). She required minimal resuscitation and was given nasal CPAP prior to transfer to the neonatal intensive care unit (NICU). Apgar scores were 7 and 8 at 1 and 5 minutes, respectively. Abdominal X-ray and ultrasonography in the NICU confirmed massive ascites with centrally floating bowel (see Figure 2) with ultrasound reporting no hydronephrosis and a decompressed bladder. The kidneys were normal. Chylous ascites was ruled out on analysis of the peritoneal fluid which showed no lymphocytes. However, the creatinine, in the peritoneal fluid, was higher than the serum creatinine, suggestive of urinary ascites [4]. The patient's admission labs and peritoneal fluid analysis are available in the Appendix. Adequate urine output was only noted on the aspiration of an indwelling urinary catheter with a syringe, which was suggestive of neurogenic bladder. Cystography and cystoscopy did not show any evidence of vesicoureteric reflux or urinary leakage from the urinary system, and hydronephrosis was evident on a repeat renal sonogram. Further diagnostic investigation with an abdominal MRI demonstrated a spinal mass. A subsequent focused MRI evaluation of the brain and spine confirmed a lobulated mass with heterogeneous enhancement within the conus medullaris measuring 0.9 × 1.4 × 0.8 cm (see Figure 3). Spinal fluid analysis after MRI showed multiple small round blue cells. The mass was identified as WHO grade III/IV malignant glioma on excision biopsy (see Figure 4). The hydronephrosis resolved with continuous draining of urine through the indwelling urinary catheter which was eventually discontinued during her NICU course as she started voiding spontaneously after the subtotal excision biopsy. She was discharged home and is currently undergoing experimental chemotherapy for the glioma. 3. Discussion Isolated neonatal ascites defined as ascites without hydrops is rare. When it occurs, the most common cause is urinary ascites with urinary obstruction accounting for approximately 60 to 70% of cases, isolated fetal chylous ascites accounts for 4 to 20% of the cases, and hepatobiliary and intestinal causes account for 15 to 30% [5]. The most common cause of the urinary obstruction is posterior urethral valves and less commonly urethral atresia. All documented cases of urinary ascites in the literature are from prenatally diagnosed bladder rupture and all were in male infants except one, a female diagnosed in utero with spontaneous bladder rupture of unknown etiology [6]. There were two documented cases of urinary ascites due to spinal cord pathology: one was a case of bladder rupture and dysfunction secondary to neurogenic bladder from a sacrococcygeal teratoma [7] and the other as a result of complication from surgical repair of myelomeningocele. Mann et al. also reported a case of urinary ascites from bladder rupture secondary to a congenital neuroblastoma with complete block of the spinal canal in the lumbar region [8]. Our patient had urinary ascites with no evidence of bladder rupture. The origin of the urine might be from breaks in the bladder or at the level of calyceal fornices not seen on ultrasound or it may have been from transudation of urine from an intact urinary tract [7]. The most acceptable theory for urinary ascites caused by a neurogenic bladder is that extrinsic compression by a mass can disrupt nerve connections to the bladder resulting in bladder atonia and eventual rupture. When congenital urinary ascites occurs in neonates, it can be complicated by electrolyte disturbances including hyponatremia, hyperkalemia, and elevated creatinine as a result of “autodialysis” [8] when urine is in contact with peritoneal membrane. These complications were not noted in our patient, perhaps due to acute recent onset of anhydramnios with little time for serum changes to occur. Spinal cord tumors are rare in the neonatal period, with conus medullaris tumors being even rarer among this category. The most common neonatal tumors are teratomas which are most likely extragonadal and mainly present in the sacrococcygeal or mediastinal area. Neurologic features of spinal cord tumors in the lumbosacral-coccygeal area include lower extremity weakness and disturbance of sphincter function [9]. All previously reported 24 cases of congenital malignant glioma involved the brain (19 supratentorial cases, 1 infratentorial case, and 2 cases with locations not given) [10]. Among intraspinal neoplasms in the pediatric age group, the most frequently noted were astrocytoma (47%) and the ependymal neoplasm (24%) [11]. The median age at diagnosis was 10 years with a male-to-female ratio of 1 : 1. Our case, therefore, provides documentation of the youngest infant with malignant spinal cord glioma. Although it is a rare condition, a diagnosis of spinal cord tumor should be considered in neonates with congenital ascites and neurogenic bladder. Despite advances in imaging technology to diagnose spinal cord tumors in neonates, the treatment modalities have not had the same significant gains. Current treatment is still experimental chemotherapy, as is the case with our index patient. Appendix See Table 1. Disclosure The authors have no financial relationships relevant to this paper to disclose. Competing Interests All authors have no conflict of interests to disclose. Figure 1 Neonate soon after birth: physical exam suggestive of massive ascites. Figure 2 Abdominal X-ray, day of life 1: distended abdomen with centralization of bowel loops showing massive ascites. Figure 3 MRI spine, with and without contrast: expansile, somewhat lobulated mass in conus medullaris (see arrow). Figure 4 Surgical pathology: malignant neoplasm most consistent with malignant glioma. Table 1 Admission labs and peritoneal fluid analysis. (a) Basic metabolic panel Values Normal range Sodium (mmol/L) 138 135−148 Potassium (mmol/L) 5.2 3.5−5.1 Chloride (mmol/L) 103 98−108 Bicarbonate (mmol/L) 23 21−31 Glucose (mg/dL) 83 50−80 BUN (mg/dL) 11 5−20 Creatinine (mg/dL) 0.23 0.5−1.2 Calcium (mg/dL) 9.8 8.6−10.2 Phosphorus (mg/dL) 6.4 4.5−6.7 (b) Complete blood cell count/differentials Values Normal range WBC count (×103/μL) 17.77 7.0−17 RBC count (×106/μL) 3.03 3.46−6.26 Hemoglobin (g/dL) 9.7 14.6−20.1 Hematocrit (%) 30.1 50−69 MCV (fL) 99.3 96−120 MCH (pg) 32.0 27−31 MCHC (g/dL) 32.2 33−37 RDW (%) 16.3 11.2−14.8 PLT count (×103/μL) 243 150−350 MPV (fL) 10.4 8.0−12 Neutrophil (%) 51   Band (%) 2 0−15 Eos (%) 3 0−3 Lymphocyte (%) 30   (c) Peritoneal fluid analysis Values Color Yellow WBC count (per μL) 11 RBC count (per μL) 18 Neutrophil (%) 16 Mononuclear (%) 68 Eosinophils (%) 16 Albumin (g/dL) 1.6 BUN (mg/dL) 13 Cholesterol (mg/dL) 35 Creatinine (mg/dL) 0.4 LDH (IU/L) 141 pH 8.0 ==== Refs 1 Wilson P. E. Oleszek J. L. Clayton G. H. Pediatric spinal cord tumors and masses Journal of Spinal Cord Medicine 2007 30 Supplement 1 S15 S20 2-s2.0-34548240650 17874681 2 Hardison H. H. Packer R. J. Rorke L. B. Schut L. Sutton L. N. Bruce D. A. Outcome of children with primary intramedullary spinal cord tumors Child's Nervous System 1987 3 2 89 92 10.1007/BF00271131 2-s2.0-0023277136 3 Nadkarni T. D. Rekate H. L. Pediatric intramedullary spinal cord tumors: critical review of the literature Child's Nervous System 1999 15 1 17 28 10.1007/s003810050321 2-s2.0-0032910887 4 Chilakala S. K. Boulden T. F. Pourcyrous M. Ruptured remnant of urachal diverticulum: an unusual cause of congenital urinary ascites Journal of Perinatology 2012 32 12 978 980 10.1038/jp.2012.30 2-s2.0-84870515228 23190938 5 Herman T. E. Siegel M. J. Congenital chylous ascites Journal of Perinatology 2009 29 2 178 180 10.1038/jp.2008.125 2-s2.0-59549093129 19177050 6 Bracero L. A. Seybold D. J. Koike J. Prenatal diagnosis of spontaneous bladder rupture in a female fetus Urology 2011 77 6 1474 1476 10.1016/j.urology.2010.07.500 2-s2.0-79957926141 21256559 7 Zaninovic A. C. Westra S. J. Hall T. R. Sherman M. P. Wong L. Boechat M. I. Congenital bladder rupture and urine ascites secondary to a sacrococcygeal teratoma Pediatric Radiology 1992 22 7 509 511 10.1007/BF02012995 2-s2.0-0027049619 1491908 8 Mann C. M. Leape L. L. Holder T. M. Neonatal urinary ascites: a report of 2 cases of unusual etiology and a review of the literature Journal of Urology 1974 111 1 124 128 2-s2.0-0015950352 4855936 9 Volpe J. J. Clinical Key (Online Service). Neurology of the Newborn 2008 5th Philadelphia, Pa, USA Saunders/Elsevier 10 Winters J. L. Wilson D. Davis D. G. Congenital glioblastoma multiforme: a report of three cases and a review of the literature Journal of the Neurological Sciences 2001 188 1-2 13 19 10.1016/s0022-510x(01)00538-x 2-s2.0-0035878592 11489279 11 Farwell J. R. Dohrmann G. J. Flannery J. T. Central nervous system tumors in children Cancer 1977 40 6 3123 3132 2-s2.0-0017683186 201364

PMC005xxxxxx/PMC5002299.txt

==== Front Mediators InflammMediators InflammMIMediators of Inflammation0962-93511466-1861Hindawi Publishing Corporation 10.1155/2016/3481371Review ArticleDeubiquitinases: Novel Therapeutic Targets in Immune Surveillance? http://orcid.org/0000-0002-8585-3381Lopez-Castejon Gloria 1 * http://orcid.org/0000-0003-0459-7716Edelmann Mariola J. 2 1Manchester Collaborative Centre for Inflammation Research, University of Manchester, 46 Grafton Street, Manchester M13 9NT, UK2Department of Microbiology and Cell Science, College of Agricultural and Life Sciences, University of Florida, 1355 Museum Drive, Gainesville, FL 32611-0700, USA*Gloria Lopez-Castejon: gloria.lopez-castejon@manchester.ac.ukAcademic Editor: Luca Cantarini 2016 14 8 2016 2016 348137124 3 2016 1 6 2016 4 7 2016 Copyright © 2016 G. Lopez-Castejon and M. J. Edelmann.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Inflammation is a protective response of the organism to tissue injury or infection. It occurs when the immune system recognizes Pathogen-Associated Molecular Patterns (PAMPs) or Damage-Associated Molecular Pattern (DAMPs) through the activation of Pattern Recognition Receptors. This initiates a variety of signalling events that conclude in the upregulation of proinflammatory molecules, which initiate an appropriate immune response. This response is tightly regulated since any aberrant activation of immune responses would have severe pathological consequences such as sepsis or chronic inflammatory and autoimmune diseases. Accumulative evidence shows that the ubiquitin system, and in particular ubiquitin-specific isopeptidases also known as deubiquitinases (DUBs), plays crucial roles in the control of these immune pathways. In this review we will give an up-to-date overview on the role of DUBs in the NF-κB pathway and inflammasome activation, two intrinsically related events triggered by activation of the membrane TLRs as well as the cytosolic NOD and NLR receptors. Modulation of DUB activity by small molecules has been proposed as a way to control dysregulation or overactivation of these key players of the inflammatory response. We will also discuss the advances and challenges of a potential use of DUBs as therapeutic targets in inflammatory pathologies. National Institute of Food and Agriculture109705571271158 ==== Body 1. Introduction Ubiquitination is a posttranslational modification (PTM) that involves the attachment of a ubiquitin molecule (~9 kDa) to a target protein. It is now well accepted that most of the cellular processes required for the maintenance of the cell homeostasis are regulated by the ubiquitin-proteasome system (UPS), including the regulation of innate immune signalling. Ubiquitination is mediated by a set of three enzymes, a ubiquitin-activating enzyme (E1), a ubiquitin-conjugating enzyme (E2), and a ubiquitin ligase (E3). Ubiquitin (Ub) is attached as a monomer or as polyubiquitin (poly-Ub) chains. This attachment occurs between a lysine group in the target protein and the carboxy-terminal glycine of Ub. The formation of Ub chains, however, occurs by formation of a bond between the carboxy-terminal glycine of Ub and one of the seven lysines (K6, K11, K27, K29, K33, K48, and K63) or the methionine (M1) present in the acceptor Ub molecule [1] allowing the generation of a wide variety of poly-Ub chains. Each poly-Ub chain type will influence the fate of the target protein differently. For instance, K48-conjugated Ub chains are considered a signal for protein degradation at the proteasome while K63 and M1 chains play important roles in signalling pathways [1]. Ubiquitination is a reversible process, and its reversibility is mediated by a family of proteases called deubiquitinases (deubiquitinating enzymes, DUBs). Keeping the balance between the addition and removal of ubiquitin moieties is crucial in maintaining cellular homeostasis and any disturbances in this balance can have adverse consequences for the cell. 2. Mechanisms of Regulation of DUBs The human genome encodes ~100 DUBs that fall into five different families. There are four thiol protease families, the ubiquitin-specific proteases (USP), ubiquitin C-terminal hydrolases (UCH), ovarian tumour domain containing proteases (OTU), and Machado Joseph disease (MJD)/Josephin domain DUBs, and one zinc-metalloprotease group, the JAB1/MPN/Mov34 metalloenzyme family [2]. The main functions of DUBs are (i) generation/release of free ubiquitin from Ub precursors (de novo Ub synthesis), (ii) subtle editing of poly-Ub chains, and (iii) removal of the poly-Ub chains from substrates prior to degradation by proteasome-bound DUBs. DUBs, similarly to other proteases, are tightly regulated to avoid aberrant function that could be therefore detrimental to the cell. This is achieved by a combination of different layers of regulation at transcriptional and nontranscriptional levels. As many other proteins DUBs are regulated at the transcriptional level. One of the best examples of transcriptional regulation is A20 (TNFAIP3), which is a member of the OTU family of DUBs. A20 expression levels are highly upregulated in a proinflammatory environment (i.e., in response to TLR4 activation) [3], reflecting its important role as a negative regulator in the inflammatory response, as we will discuss below. There are other DUBs, which are regulated in response to cytokines, including DUB1, DUB2, USP17 (DUB3), and OTUD-6B. DUB1 is specifically induced by IL-3, IL-5, and GM-CSF, while DUB2 is stimulated by IL-2. USP17 (DUB3) is involved in the regulation of cell growth and survival and it is regulated by the cytokines IL-4 and IL-6 [4]. Ovarian tumour domain containing 6B (OTUD-6B) is a DUB, whose expression in B lymphocytes is induced by secretion of IL-3, IL-4, IL-13, or GM-CSF. With prolonged stimulation, these cytokines have an opposite effect and instead lead to a decrease in OTUD-6B expression. A higher expression of OTUD-6B was associated with inhibition of cell growth, an increase in apoptosis, and arrest of cells in G1 phase [5]. DUBs are also heavily regulated at the activity level by different mechanisms. DUBs can acquire specificity due to recruitment factors that guide them towards a specific substrate. One example is USP10 that requires the protein MCPIP-1 (monocyte chemotactic protein induced protein 1) to interact with and deubiquitinate its substrate NEMO inhibiting the NF-κB signalling cascade [6]. In other cases binding of the substrate actively contributes to DUB catalysis. For instance, USP7, whose catalytic triad exists in an inactive configuration, changes towards an active one upon ubiquitin binding suggesting that USP7 catalytic domain is only fully active when a ubiquitin molecule is correctly bound [7]. The presence of DUBs in molecular complexes is a common way to modulate their activity. This mechanism is essential for USP1, an inefficient enzyme alone, but its activity highly increases when bound to the WD40-repeat protein UAF-1 due to conformational changes that increase its catalytic activity [8]. USP1 is involved in DNA damage response, mainly in the Fanconi anemia (FA) pathway where it mediates the deubiquitination of FANCD2 and FANCI, a crucial step for the correct function of the FA pathway [9, 10]. Additionally DUB activity can be further adjusted by posttranslational modifications such as phosphorylation or ubiquitination [11]. For instance, phosphorylation of CYLD at Ser418 or USP7 at Ser18 led to an increase in the activities of these two DUBs. In the case of CYLD, this modification can be induced by LPS (lipopolysaccharide) or TNF-α (tumour necrosis factor) treatment and it can suppress its deubiquitinating activity on TNF receptor-associated factor 2 (TRAF2) [12]. Furthermore, this phosphorylation also occurs in dendritic cells (DCs) treated with LPS/Lex, which leads to a diminished activity of CYLD but not to its complete loss. This effect can be reversed by an inhibition of DC-SIGN signalling and also by depletion of IKKε [13]. Changes in the cellular microenvironment can also have an effect on DUB activity. One example is the production of reactive oxygen species (ROS) generated during mitochondrial oxidative metabolism as well as in cellular responses to cytokines or bacterial invasion, which can inhibit cellular DUB activity by oxidation of the catalytic cysteine residue [14, 15]. To summarize, more than one regulatory mechanism can apply to certain DUBs, which highlights the importance of a fine and multifaceted control of DUB expression and activity. 3. Deubiquitination in TLR- and NLR-Mediated Immune Signalling Innate immunity is triggered in response to either PAMPs, which are derived from microbial pathogens, or DAMPs such as ATP, cholesterol, or monosodium urate crystals. These danger signals are recognized by Pattern Recognition Receptors either at the cell membranes by Toll-Like Receptors (TLRs) or at the cytosol by receptors such as the NOD-like receptors (NLRs) [16]. Activation of these PRR receptors results in a variety of immune signalling cascades which lead to the induction of immune mediators and proinflammatory cytokines, such as TNFα or IL-1β, capable of triggering appropriate immune responses. These cytokines lead to the recruitment of immune cells to the site of infection or tissue damage, which initiates an inflammatory response. TLR- and NLR-mediated signalling is heavily controlled by the ubiquitin system, which plays an essential role in maintaining the appropriate regulation of these cellular pathways [1]. Although DUBs can be involved in many other inflammatory aspects, here we will discuss how DUBs contribute to TLR and NLR-induced pathways, focusing on the activation of two very important and related processes, NF-κB pathway and inflammasome activation. 3.1. TLR Signalling TLRs are transmembrane glycoproteins, which play a key role in the immune response against microbes. Ten human TLRs have been identified to date and they either localize to the cell surface (TLRs 1, 2, 4, 5, 6, and 10) or have endosomal localization (TLRs 3, 7, 8, and 9 [17]). There are two distinguishable pathways of TLR signalling, one via the MyD88 (myeloid differentiation primary-response protein 88) and the second one via TRIF (TIR domain containing adaptor protein inducing interferon α/β) and apart from TLR3, most other TLRs are associated with the MyD88 pathway [18]. Ubiquitination is critically involved in optimal TLR-triggered MyD88 and TRIF signalling (Figure 1). TLR3 engagement induces the recruitment of TRIF and modification of TRAF3 with K63 poly-Ub, which consequently recruits the TBK1 (TRAF family member-associated NF-κB activator-binding kinase)/IKKε kinase complex. Finally, this cascade of events causes IRF3 activation and INFγ production. In contrast, TLR4 or TLR2 activation leads to the assembly of the MyD88 signalling complex, recruiting TRAF6, cIAP1, and cIAP2. These ubiquitin ligases mediate K48-linked poly-Ub of TRAF3, and TRAF3 is consequently degraded by the proteasome [19]. TRAF6 ubiquitin ligase activity is essential for the synthesis of K63-linked poly-Ub chains, which act as a scaffold to recruit other proteins required for signalling. TRAF6 K63-linked poly-Ub chains recruit both the TAK1 and IKK complexes through their respective ubiquitin-binding subunits, TAB2/3 and NEMO. This occurs with the help of the LUBAC ubiquitin ligase complex, which leads to the linear ubiquitination of NEMO required for the recruitment of the IKK complex (IKKα and IKKβ). As a result, TAK1 phosphorylates IKKβ, which in turn phosphorylates IκB and subsequently undergoes ubiquitination and proteasomal degradation [20]. This allows NF-κB to translocate to the nucleus from cytosol and regulate the transcription of a variety of target genes (Figure 1). Deubiquitination also plays a key role in TLR signalling pathways by reversing the effect of ubiquitination and controlling the intensity of the immune response (Figure 1). Several DUBs have been identified to participate in the TLR signalling, the most studied and best characterized being A20 (TNFAIP3) and CYLD. A20 plays an essential role in restricting TLR signalling and maintaining immune homeostasis. A20 contains an OTU domain, which has DUB activity specific towards several NF-κB signalling factors, such as TRAF6, RIPK1, or NEMO, which consequently leads to suppressed NF-κB activation [21]. A20 is an unusual DUB because it encodes seven zinc-finger (ZnF) motifs, which confer E3 ubiquitin ligase activity on A20. This allows A20 to perform an editing function: in addition to removing K63-linked polyubiquitin chains from substrates such as RIPK1, A20 can introduce K48-polyubiquitin chains in the same substrate tagging it for a proteasomal degradation [21]. In addition to this, A20 can also regulate NF-κB independently of its enzymatic activity. A20 can bind polyubiquitin chains through its ZnF domain allowing the interaction of ubiquitinated NEMO with A20. This ubiquitin-induced recruitment of A20 to NEMO is sufficient to block IKK phosphorylation by its upstream kinase TAK1 preventing NF-κB activation [22]. In contrast CYLD is a tumour suppressor, whose loss leads to familial cylindromatosis, a skin tumour hereditary disorder, but that also controls NF-κB activation. CYLD achieves this by specifically cleaving K63-linked poly-Ub chains and linear poly-Ub chains from RIPK1, TRAF2, and NEMO and similarly to A20 negatively regulates NF-κB signalling [23]. USP7 was first identified as a herpesvirus associated protein, hence its alternative name HAUSP (herpesvirus associated USP). USP7 presents dual roles in the regulation of NF-κB. It can regulate NF-κB transcriptional activity in the nucleus, by deubiquitinating NF-κB and preventing its degradation, hence increasing its transcriptional activity [24]. But USP7 can also act as a negative cytosolic regulator by deubiquitinating NEMO and consequently decreasing proteasomal degradation of IκBα. This in turn retains NF-κB in the cytoplasm and further suppresses NF-κB activity [25]. These two reported and opposing roles suggest that USP7 can perform different functions roles, depending on substrate recognition or cellular localization, highlighting the tight activity control of this protease. As previously mentioned, USP10 is required for mediated inhibition of NF-κB activation. By mediating USP10-dependent deubiquitination of NEMO, MCPIP1 serves in a negative feedback mechanism for attenuation of NF-κB activation [6]. TRAF family member-associated NF-κB activator (TANK) interacts with both MCPIP1 and USP10, which leads to decrease in TRAF6 ubiquitination and the termination of the NF-κB activation in response to TLR activation [26]. In accordance with this, depletion of USP10 is associated with TLR-triggered increase in NF-κB activation [26]. USP18 is responsible for counteracting ISG15 conjugation and it is an important negative regulator of the IFN responses, thereby playing important roles in viral responses [27]. However, we now know that USP18 also mediates and regulates TLR-induced NF-κB activation by cleavage of K63-polyubiquitin chains, but not K48 chains, of TAK1 and NEMO [28]. In addition to the DUBs described here there are several others implicated in the downregulation of the NF-κB pathway upon TLR activation, although these are not well characterized. These include the USP family members USP2a, USP4, USP15, USP21, and USP31 and the member of the JAMM family MYSM1 and their substrates have been summarized in Table 1. 3.2. NLR Signalling The NLR family presents a characteristic tripartite domain architecture with a variable C-terminus, a middle NACHT domain, and a Leucine Rich Repeat (LRR) N-terminus. The C-terminal LRR domain is involved in the ligand binding or activator sensing while the N-terminal domain performs effector functions by interacting with other proteins. NLRs are classified into four subfamilies according to their N-terminal domains: the acidic transactivation domain (NLRA), the baculoviral inhibitory repeat-like domain (NLRB) that includes NOD1 and NOD2, the caspase activation and recruitment domain (CARD; NLRC), and the pyrin domain (NLRP). NLRs can recognize a wide variety of ligands including pathogens, endogenous molecules, or environmental factors [29]. Their functions can vary and they are divided into four steps: inflammasome formation, signalling transduction, transcription activation, and autophagy [29]. Similarly to TLRs, NLR activation is also tightly regulated and PTMs play an important role here. Although ubiquitination in NLR signalling is well accepted, the role of DUBs in these pathways is just emerging. 3.2.1. NOD1 and NOD2 NOD1 and NOD2 receptors are important bacterial sensors, which recognize peptidoglycan (PGN). NOD1 senses the iE-DAP dipeptide, which is found in PGN of all Gram-negative and certain Gram-positive bacteria, while NOD2 recognizes MDP (muramyl dipeptide), the minimal bioactive peptidoglycan motif common to all bacteria (Figure 2(a)). Upon encountering with these ligands, NOD1 and NOD2 form oligomeric complexes, leading to the activation of NF-κB and MAPK. IAPs (cIAP1, cIAP2, and XIAP) are central regulators of NOD1 and NOD2 signalling. Upon oligomerization RIPK2 is recruited to this complex. cIAP1, cIAP2, and XIAP contribute to K63-linked ubiquitination of RIPK2. This allows the recruitment of TAK1/TAB2/TAB3 complex and LUBAC, which can also mediate the linear ubiquitination of RIPK2, and further contributes to the NF-κB and MAPK pathway activation by ubiquitination of NEMO [30, 31]. Ubiquitin can directly bind to the CARD domain of NOD1 or NOD2 and compete with RIPK2 for its association with these receptors, suggesting that ubiquitin might play a negative regulatory role [32, 33] (Figure 2(b)). A20 also plays a regulatory role in NOD2 signalling by deubiquitinating RIPK2 to control the extent of the inflammatory signals. A20-deficient cells present an amplified response to MDP, including increased RIPK2 ubiquitination and NF-κB signalling [34]. One of the DUBs, which is relatively poorly characterized but which has been shown to play key functions in NOD2 signalling, is OTULIN. This protein specifically deconjugates linear (M1) poly-Ub chains assembled by LUBAC and in this way it modulates linear ubiquitination of LUBAC's substrates and provides fine-tuning of the initial activation of NF-κB. By deubiquitinating RIPK2, OTULIN prevents NEMO binding and hence decreases its downstream signalling. Because LUBAC continuously ubiquitinates itself and other substrates, OTULIN plays an important role to avoid accumulation of Met1-Ub chains and overactivation of this pathway [35] (Table 1, Figure 2(b)). 3.2.2. The Inflammasome Another crucial function of NLR receptors is their contribution to the inflammasome. The inflammasome is a molecular complex, which consists of a sensor molecule (NLR, e.g., NLRP1, NLRP3, NLRC4, or NLRP6), an adaptor protein (ASC, apoptosis-associated speck-like protein containing a CARD domain), and an effector molecule (caspase-1) [36]. The main function of the effector molecule is to induce the cleavage and activation of the proinflammatory cytokines, IL-1β and IL-18. These proinflammatory proteins are synthesized as precursor molecules and require caspase-1 activation within the inflammasome in order to be released and cleaved and perform their biological activity. Activation of inflammasomes occurs in two steps. First, an NF-κB mediated initial step leads to increased expression of NLRP3 and pro-IL-1β. Then an activating signal triggers rapid activation of caspase-1. Caspase-1 activation can be achieved by several K+-releasing molecules, including nigericin, crystals, or extracellular ATP through the activation of the ATP-gated P2X7 receptor (P2X7R) [37]. After the inflammasome is fully activated, it can lead to pyroptotic cell death, which can be distinguished from other cell death types by pore formation in the plasma membrane followed by osmotic cell lysis and finally the release of IL-1β and IL-18 [36]. Given the important role of ubiquitin in signalling cascades derived from TLR and NLR activation, it is not surprising to find that assembly and activation of an inflammasome is also regulated by the ubiquitin system. Ubiquitination can regulate canonical inflammasome activation by modulation of three major components: NLR, ASC, and caspase-1. Ubiquitin ligases can also directly influence NLRP3 inflammasome activation. This can be exemplified by MARCH7, which promotes ubiquitination of NLRP3, and this causes its degradation upon dopamine stimulation as a mean to control inflammasome activation [38]. Another example is SCFFBXL2, whose activity is impaired upon LPS priming preventing NLRP3 ubiquitination and its consequent degradation [39] (Figure 3). Other ubiquitin ligases have also been involved in control of NLRP3 ubiquitination. For instance, TRIM30 can negatively regulate NLRP3 inflammasome by modulating the levels of ROS species in the cell. TRIM30−/− macrophages produce higher levels of ROS and potentiate NLRP3 inflammasome activation; however the mechanisms by which TRIM30 controls this remain unknown [40]. However, TRIM33 is essential for cytosolic RNA-induced NLRP3 inflammasome activation. TRIM33 ubiquitinates DHX33, a cytosolic dsDNA sensor for NLRP3, allowing DHX33-NLRP3 interactions and consequent inflammasome activation [41]. Similarly to the NOD2 receptor activation, cIAP E3s are also involved in the inflammasome activation. Attenuation of cIAP activities, either by their deletion or by inhibition, triggers NLRP3 and caspase-1 activation as well as RIP3 kinase-dependent IL-1β processing and secretion [42]. On the other hand, cIAP1 and cIAP2 can attach K63-linked poly-Ub chains to caspase-1, thereby facilitating caspase-1 activation and IL-1β release [43]. Caspase-1 ubiquitination also occurs in response to the NLRP1 activator anthrax lethal toxin [43, 44] although the type of ubiquitin chains and whether this is a requirement for caspase-1 activation still remain unclear. In addition to NLR and caspase-1, ubiquitin-mediated inflammasome activation can be also promoted by modification of the adaptor protein ASC. Activation of the inflammasome can induce autophagy as a mean to control inflammasome activation. In this situation, K63 poly-Ub modification of ASC allows for its interaction with the autophagic adaptor p62 and delivery of ASC to the autophagosome [45]. TRAF3 ubiquitin ligase ubiquitinates ASC, and abolishment of the target lysine (K174) prevents inflammasome activation and IL-1β release in response to viral infection [46]. Also, TRAF6-mediated ASC ubiquitination has been recently reported in response to far-infrared and proposed to constitute a mechanism, which dampens inflammasome activation in repair processes [47]. Interestingly ASC has been identified as a substrate of HOIL-1L, a member of linear ubiquitination complex LUBAC, and HOIL deficient macrophages present an impaired inflammasome response [48]. In line with this, macrophages deficient in SHARPIN, which is a different member of the LUBAC complex, are not able to mount an optimal inflammasome response [49]. All this evidence reveals that ubiquitination is an essential modification for the control of the inflammasome activation. It is then logical to assume that DUBs are important players of these regulatory mechanisms. This was first suggested by Juliana et al., who showed that NLRP3 is ubiquitinated in resting macrophages and that, upon cell activation with priming (LPS) and activating signals (ATP, nigericin, and MSU crystals), these ubiquitin chains are removed by DUBs, allowing activation of the complex [50]. This report was quickly followed by two other studies supporting these results [51, 52], and it was Py et al. who identified BRCC3 as the first DUB to be directly involved in inflammasome activation. These reports showed that inhibition of DUB activity with the DUB inhibitors bAP-15, WP1130, PR-619, and G5 blocks NLRP3 but not NLRC4 or AIM2 mediated IL-1β release and pyroptosis (Figure 3; Table 1). Moreover, a recent report has demonstrated that histone deacetylase 6 (HDAC6) negatively regulates NLRP3 inflammasome activation. HDAC6 interacts with NLRP3's ubiquitin-binding domain and treatment with the DUB inhibitor PR-619 results in an increased interaction of NLRP3 with HDAC6. The authors suggest this is due to an increased ubiquitination of NLRP3 and the consequent inhibition of NLRP3-dependent caspase-1 activation [53]. The ability of these DUB inhibitors to block inflammasome activation could explain the inhibitory effect of the compound Bay 11-7082 on NLRP3 inflammasome independently of its NF-κB inhibitory activity [54] since this compound can inhibit components of the ubiquitin system, including DUBs [55, 56]. The other DUB, which has been directly implicated in the inflammasome activation, is A20. In contrast to BRCC3, A20 acts as a negative regulator of NLRP3 and suppresses inflammasome activation by restricting ubiquitination of IL-1β and NLRP3 activation [57, 58]. Given the fine-tuning and the layers of regulation required for both the inflammasome and DUB activation, it is quite likely to think that different DUBs might perform opposing functions pertaining to the inflammasome activation. Whether DUBs regulate the ubiquitination state of ASC or caspase-1 involved in the inflammasome assembly still remains unknown. 4. Pathogen Manipulation of DUBs to Control PRR Signalling During pathogenesis, deubiquitinating enzymes are regulated both by microorganisms and by a host cell. Pathogens can exploit the host ubiquitin system by expressing their own ubiquitin-specific enzymes, and the host cell can up- or downregulate expression and/or activity of host DUBs [59]. First, an example of a pathogen-encoded deubiquitinase disturbing the host innate immune pathways is Salmonella's AvrA, which is a DUB that facilitates inhibition of the NF-κB pathway. AvrA leads to stabilization of IκBα and prevents nuclear translocation of NF-κB p65. Also, depletion of AvrA in Salmonella leads to significantly increased secretion of cytokine IL-6 in the host cell, which is dependent on NF-κB pathway [60–63]. As a second example, Chlamydia trachomatis encodes two DUBs, ChlaDub1 and ChlaDub2, which are specific for ubiquitin but they also harbour deneddylating activity [64]. ChlaDub1 binds and stabilizes IκBα, most likely via its deubiquitination, and finally this can lead to an inhibition of NF-κB activation [65]. Since several known bacterial DUBs directly target important functions in the host immune system, development of selective inhibitors for pathogenic DUBs could be exploited as a therapeutic approach in the treatment of infections. Bacterial infection can induce inflammasome activation in the host cell [36] and deubiquitination has been implicated in this process. Salmonella Typhimurium infection leads to changes in the activity of several host DUBs, such as USP4, USP5, UCHL3, and UCHL5, and increased activity of UCHL5 was found to contribute to the inflammasome activation during this infection [66]. Additionally, enteropathogenic Escherichia coli protein NleA associates with and interrupts deubiquitination of NLRP3, thereby repressing inflammasome activation [67]. 5. Deubiquitinases and Inflammatory Disease Accumulating evidence indicates that somatic mutations in DUBs are correlated with human disease. DUBs are genetically altered in many human cancers (i.e., CYLD, A20, or USP6) or contribute to the stability of oncogenes or tumour suppressors (i.e., USP7, USP8, or BRCC3) [68]. Here we will highlight DUBs with potential implications in immune disease although the scope for other DUBs contributing to disease is very high. Although many of the studies mentioned in this review have been performed in vitro in cell culture models, the involvement of DUBs in inflammatory responses has been also studied by using animal models, highlighting the relevance of these proteases in a relevant tissue and immune context (Table 1). Mutations in the CYLD gene lead to a subtype of the benign cancer predisposition syndrome of skin appendages also known as Brooke-Spiegler syndrome, although inactivation or downregulation of CYLD is also observed in a variety of other cancers, including melanoma, and breast, colon, lung, breast, cervical, and, recently, prostate cancer. As previously mentioned CYLD can bind to NEMO and NF-κB that have been identified as its substrates. It is possible that the negative regulation of NF-κB mediated by CYLD contributes to its tumour suppression function given the increasingly recognized role for NF-κB in cancer advancement. CYLD deactivation could provide specific advantage to tumour cells by enhanced NF-κB signalling [69–71]. CYLD-deficient mice present abnormalities in their immune system. They show increased basal and induced NF-κB activation and can develop autoimmune symptoms and colonic inflammation with features of human inflammatory bowel disease [72], and their inflammatory responses in response to pathogenic infection are potentiated [73]. A20 is an important negative regulator of immune response as we have mentioned before. Multiple mutations in the A20 gene have been identified; however no inheritable syndrome has so far been linked with A20 abnormalities. This could be explained if these mutations were developmentally critical. A20 mutations are strongly linked to autoimmunity, lymphomas, and asthma [74, 75], highlighting important differences to CYLD despite both targeting NF-κB. This might be explained by different chain preference, K48 and K11 for A20 compared to the K63 and M1 chain preference showed by CYLD [68]. A20−/− mice fail to regulate NF-κB responses, develop severe inflammation and are hypersensitive to LPS or TNFα leading to premature death [76]. Cell specific ablation of A20 has revealed important knowledge about the contribution of A20 to disease pathogenesis and generated very useful mouse models for several conditions like rheumatoid arthritis, lupus erythematosus, or inflammatory bowel disease [75]. USP18 has been thoroughly studied in the context of viral responses, since it regulates protein ISGylation in response to viral infection. However Liu et al. also demonstrated that USP18 deficient mice are resistant to experimental autoimmune encephalomyelitis (EAE) [77]. This study proposes that USP18 regulates TAK1-TAB interaction and is hence necessary for Th17 differentiation and autoimmune response. DUBs can contribute to disease not only by mutations, but also by an altered expression or activity. An example of this is USP7, whose increased activity mediates the deubiquitination and destabilization of a number of critical tumour suppressors, including p53 or PTEN, and is by inference an oncogenic prosurvival protein. The interrelationship between p53, USP7, and MDM2 ubiquitin ligase is quite unique and complex. USP7 can deubiquitinate and stabilize p53, but interestingly it can also deubiquitinate and stabilize MDM2 indirectly leading to p53 destabilization and its degradation by the proteasome [78]. USP7 also interacts and stabilizes the ICP0 ubiquitin E3 ligase of herpes simplex virus (HSV), which is required for the effective initiation of the lytic cycle, facilitating lytic viral growth [79]. USP7 can also interact with other viral proteins, such as the EBNA1 protein of the Epstein-Barr virus (EBV) [80] and the Viral Interferon Regulatory Factor 1 (vIRF1) of a Kaposi sarcoma herpesvirus protein [81]. In addition, and as mentioned before, USP7 plays a role by regulating NF-κB signalling [24, 25]. Unfortunately USP7−/− mice are embryonically lethal explaining the lack of in vivo studies to further characterize the role of USP7 in immune responses and associated pathologies [82]. 6. Modulating DUB Activity as a Novel Inflammatory Therapeutic Approach Given the importance of DUBs in inflammatory and other pathological responses, it is certainly easy to think of DUBs as potential therapeutic targets, whose modulation could be beneficial for inflammatory conditions. However, up to date there are no DUB targeting compounds that have been approved for clinical use, either in the inflammatory or in cancer context. The identification and success of inhibitors that target other elements of the ubiquitin system suggest that altering inflammation by targeting the ubiquitin system, including DUBs, could be a viable approach to develop novel anti-inflammatory treatments. An example of successful development of UPS inhibitors has been achieved with the proteasomal inhibitors Bortezomib or Carfilzomib, which have been effected in multiple myeloma treatment [83]. Another compound, MLN4924 (Nedd8-E1 enzyme inhibitor), has reached phase I clinical trials [84] and SMAC mimetics, which promote proteasomal degradation of cIAPs, have recently proved to work in cancer patients through phase I clinical trials [85]. DUB targeting drugs present a great potential as novel therapeutic agents. DUBs present the advantage of being druggable targets since they have a catalytic domain, and unlike other UPS members, such as the E3 ubiquitin ligase family with approx. 600 members, targeting the DUB family seems an achievable target. Given the clear evidence of the contribution of DUBs to disease there is a considerable effort put into the development of compounds that modulate DUB activity. Intensive research is being channelled to develop selective DUB inhibitors, which could be applied to such diseases like cancer, neurological and inflammatory disorders, or infectious disease. Despite these intensive efforts and great advances in the DUB field, selective compounds have not reached clinical trials yet. Although no DUB-selective compound has yet reached clinical trials, the field is moving fast and in the right direction. MISSION Therapeutics is developing new DUB inhibitors that present good oral bioavailability and low EC50s in cell viability assays. Proteostasis Therapeutics in collaboration with Biogen is developing very promising USP14 inhibitor series, while Genentech and Almac might be developing a new therapeutic generation of USP7 inhibitors [86, 87]. This is due to two main challenges: first not all DUBs work in the same manner hence different strategies need to be followed to develop these compounds and second we do not completely understand how these enzymes function and/or are regulated. In addition, many of the studies, which address DUB functions, have been developed in in vitro systems using either isolated proteins or cell lines that are not relevant to function or disease. This might not reflect the reality of DUB behaviour in a tissue-specific context and more work has to be developed using in vivo mice models and primary human cells. To achieve this, new and more powerful tools are required, including in-cell based assays to discriminate selective DUB function and cytotoxicity and the development of inducible mouse models, which would allow for the study of tissue-specific DUB functions. It is fundamental that basic research and drug development teams work in close collaboration to allow the success of these compounds [86, 87]. Based on our actual knowledge on DUBs thinking that not all DUBs will be good therapeutic targets is likely, since some of them might share more than one substrate, which play opposing roles in different tissues or be essential to maintain homeostasis and health. For instance, targeting USP7 in the oncology context would be a good therapeutic strategy [88]; however we need to very carefully consider the possible effects of inhibiting USP7 on the inflammatory response to the tumour. Whether this would be detrimental or beneficial still remains unknown. Similarly, we could argue that potentiating A20 function in an inflammatory context would be a plausible treatment; however more detailed studies in the consequences of this approach are required. The presence of DUBs in pathogens causative of disease, such as virus, bacteria, or parasites, has also highlighted the possibility of developing DUB inhibitors, which specifically target the pathogen and not the host. In the following years new knowledge emerging from ongoing research will allow scientists to discern those that constitute good targets and offer promising new alternatives to existing therapeutics. 7. Concluding Remarks Immune responses are strongly regulated by the addition and removal of ubiquitin molecules, and although the roles of E3 ubiquitin ligases in these signalling pathways are well established, it is still unclear how DUBs contribute to PRR signalling. The advances in this field due to novel tools and approaches including advanced mass spectroscopic techniques, ubiquitin linkage-specific antibodies, and structural and biochemical studies will provide new insights into the regulatory mechanism of immune signalling molecules by DUBs and vice versa. Since the involvement of DUBs in several inflammatory conditions is clear, development of potent and selective DUB-specific inhibitors or agonists could provide new therapeutics to treat these conditions. For instance, given the high regulation of NOD1/2 by ubiquitin and the contribution of NOD mutations to inflammatory diseases such as inflammatory bowel disease (IBD) or Crohn's disease, it is possible that DUBs could be used as a target in NOD-associated inflammatory conditions. Similarly to the kinase research area 20 years ago the DUB field is in its infancy. There are many challenges that remain to be solved to further advance our understanding of DUB function, specificity, and activity and to develop compounds that inhibit this activity. However, the field is advancing quickly, and hopefully new highly selective DUB inhibitors will be developed very soon. Acknowledgments Gloria Lopez-Castejon is supported by a Sir Henry Dale Fellowship from the Wellcome Trust/Royal Society (UK); Mariola J. Edelmann's work was supported by USDA National Institute of Food and Agriculture, Hatch Multistate Project FLA-MCS-005482. Competing Interests The authors declare that they have no competing interests. Figure 1 Regulation of TLR4 signalling by the ubiquitin-proteasome system. In MyD88-dependent signalling, TRAF6 and cIAP1/2s mediate K48 polyubiquitination and consequent degradation of TRAF3 by the proteasome. TRAF6 synthesizes K63 poly-Ub chains, which act as a scaffold for TAK1 and IKK complexes, TAB2/3 and NEMO. This occurs with the help of LUBAC, which leads to the linear ubiquitination of NEMO required for the recruitment of the IKK complex (IKKα and IKKβ). As a result, TAK1 phosphorylates IKKβ, which in turn phosphorylates IκB and subsequently undergoes ubiquitination and proteasomal degradation. This event frees NF-κB (p50/p65) to translocate to the nucleus and initiate transcription. Several DUBs (in blue) remove ubiquitin chains from TRAF6, NEMO, or NF-κB, negatively regulating this signalling pathway. USP7 can also prevent NF-κB degradation hence positively regulating transcription. MyD88-independent signalling occurs through TRAM/TRIF. In this case K63 poly-Ub chains are added to TRAF3, which consequently recruits the TBK1/IKKε kinase complex. This phosphorylates IRF3 allowing nuclear translocation and initiation of transcription. The DUB MYSM1 can deubiquitinate TRAF3, controlling the extent of this signalling. Figure 2 Regulation of NOD signalling by the ubiquitin-proteasome system. (a) NOD1 receptors recognize iE-DAP while NOD2 main ligand is muramyl dipeptide (MDP). (b) Similarly to NOD1, NOD2 receptors oligomerize upon ligand binding. This triggers the recruitment of RIPK2 to this complex and cIAP- and XIAP-mediated K63-ubiquitination of RIPK2. This allows the recruitment of TAK1/TAB2/TAB3 complex and LUBAC, which can also mediate the linear ubiquitination of RIPK2. TAK1 then phosphorylates IKKβ, which in turn phosphorylates IκB and subsequently undergoes ubiquitination and proteasomal degradation. This frees NF-κB (p50/p65) to translocate to the nucleus and initiate transcription. Deubiquitinases A20 and OTULIN are negative regulators of these events by deubiquitinating K63 and M1 poly-Ub chains, respectively. Figure 3 Regulation of the NLRP3 inflammasome activation by the ubiquitin-proteasome system. Assembly of the NLRP3 inflammasome complex occurs in response to a wide variety of danger signals including ATP, bacterial toxins, or particulate matter such as monosodium urate crystals. The ubiquitin ligases MARCH7 and SCFBXL2 add K48-linked poly-Ub chains to NLRP3 as a mean to control its levels by proteasomal degradation. cIAPs on the contrary add K63 poly-Ub chains to NLRP3 and caspase-1, contributing to the assembly of the complex. A20 also acts as a negative regulator of this complex. However, BRCC3 can deubiquitinate NLRP3, allowing it to form the complex and acting as a positive regulator of this pathway. TRAF3, TRAF6, and LUBAC also ubiquitinate ASC by K63 or M1 poly-Ub chains and this contributes to complex assembly. How other DUBs contribute to the assembly of the NLRP3 complex still remains unknown. Table 1 DUBs involved in TLR, NOD1/2, or inflammasome activation. Knock-out mouse available for these DUBs has been indicated. Mouse model validation of target in which these mice have been used to demonstrate their function on that substrate. This table does not include studies where these mice have been used in other models of inflammation. DUB PRR Target KO mouse available Mouse model validation of target Ref. USP2a TLR TRAF6 Yes No [89] USP4 TLR TAK1 Yes No [90] USP7 TLR NF-κB, NEMO No, lethal No [24, 25, 82] USP10 TLR NEMO, TRAF6 No, lethal No [6, 26, 91] USP15 TLR IκBα Yes No [92] USP18 TLR TAK1, NEMO Yes Yes [28, 77] USP20 TLR TRAF6 No No [93] USP21 TLR RIPK1 Yes No [94, 95] USP25 TLR TRAF3 Yes Yes [96, 97] USP31 TLR   Yes No [98] A20 TLR TRAF6, RIPK1, NEMO Yes Yes [3, 76, 99, 100] NOD1/2 RIPK2 Yes [34] NLRP3 inflammasome   Yes [57, 58] Cezanne TLR TRAF6 Yes Yes [101, 102] OTULIN TLR NEMO No, lethal No [103, 104] NOD2 RIPK2     [103, 105] CYLD TLR RIPK1, TRAF2, NEMO Yes Yes [12, 72] MYSM1 TLR TRAF3, TRAF6 Yes Yes [105, 106] BRCC3 NLRP3 inflammasome NLRP3 No No [51] ==== Refs 1 Komander D. Rape M. The ubiquitin code Annual Review of Biochemistry 2012 81 203 229 10.1146/annurev-biochem-060310-170328 2-s2.0-84861877407 2 Eletr Z. M. Wilkinson K. D. Regulation of proteolysis by human deubiquitinating enzymes Biochimica et Biophysica Acta (BBA)—Molecular Cell Research 2014 1843 1 114 128 10.1016/j.bbamcr.2013.06.027 2-s2.0-84890190983 23845989 3 Yuk J.-M. Kim T. S. Kim S. Y. Orphan nuclear receptor ERRα controls macrophage metabolic signaling and A20 expression to negatively regulate TLR-induced inflammation Immunity 2015 43 1 80 92 10.1016/j.immuni.2015.07.003 2-s2.0-84937714668 26200012 4 Lim K.-H. Ramakrishna S. Baek K.-H. Molecular mechanisms and functions of cytokine-inducible deubiquitinating enzymes Cytokine and Growth Factor Reviews 2013 24 5 427 431 10.1016/j.cytogfr.2013.05.007 2-s2.0-84884706110 23773437 5 Xu Z. Zheng Y. Zhu Y. Kong X. Hu L. Evidence for OTUD-6B participation in B lymphocytes cell cycle after cytokine stimulation PLoS ONE 2011 6 1 e14514 10.1371/journal.pone.0014514 2-s2.0-79251537162 6 Niu J. Shi Y. Xue J. USP10 inhibits genotoxic NF-κ B activation by MCPIP1-facilitated deubiquitination of NEMO EMBO Journal 2013 32 24 3206 3219 10.1038/emboj.2013.247 2-s2.0-84890568965 24270572 7 Faesen A. C. Dirac A. M. G. Shanmugham A. Ovaa H. Perrakis A. Sixma T. K. Mechanism of USP7/HAUSP activation by its C-Terminal ubiquitin-like domain and allosteric regulation by GMP-synthetase Molecular Cell 2011 44 1 147 159 10.1016/j.molcel.2011.06.034 2-s2.0-80053594090 21981925 8 Villamil M. A. Chen J. Liang Q. Zhuang Z. A noncanonical cysteine protease USP1 is activated through active site modulation by USP1-associated factor 1 Biochemistry 2012 51 13 2829 2839 10.1021/bi3000512 2-s2.0-84859410891 22439892 9 Nijman S. M. B. Huang T. T. Dirac A. M. G. The deubiquitinating enzyme USP1 regulates the Fanconi anemia pathway Molecular Cell 2005 17 3 331 339 10.1016/j.molcel.2005.01.008 2-s2.0-13244291457 15694335 10 Sims A. E. Spiteri E. Sims R. J. III FANCI is a second monoubiquitinated member of the Fanconi anemia pathway Nature Structural and Molecular Biology 2007 14 6 564 567 10.1038/nsmb1252 2-s2.0-34249281152 11 Kessler B. M. Edelmann M. J. PTMs in conversation: activity and function of deubiquitinating enzymes regulated via post-translational modifications Cell Biochemistry and Biophysics 2011 60 1-2 21 38 10.1007/s12013-011-9176-6 2-s2.0-79955941973 21480003 12 Reiley W. Zhang M. Wu X. Granger E. Sun S.-C. Regulation of the deubiquitinating enzyme CYLD by Iκ B kinase gamma-dependent phosphorylation Molecular and Cellular Biology 2005 25 10 3886 3895 10.1128/mcb.25.10.3886-3895.2005 2-s2.0-18144368948 15870263 13 Gringhuis S. I. Kaptein T. M. Wevers B. A. Mesman A. W. Geijtenbeek T. B. H. Fucose-specific DC-SIGN signalling directs T helper cell type-2 responses via IKKε -and CYLD-dependent Bcl3 activation Nature Communications 2014 5, article 3898 10.1038/ncomms4898 2-s2.0-84901707773 14 Cotto-Rios X. M. Békés M. Chapman J. Ueberheide B. Huang T. T. Deubiquitinases as a signaling target of oxidative stress Cell Reports 2012 2 6 1475 1484 10.1016/j.celrep.2012.11.011 2-s2.0-84871699184 23219552 15 Lee J.-G. Baek K. Soetandyo N. Ye Y. Reversible inactivation of deubiquitinases by reactive oxygen species in vitro and in cells Nature Communications 2013 4, article 1568 10.1038/ncomms2532 2-s2.0-84875912087 16 Trinchieri G. Sher A. Cooperation of Toll-like receptor signals in innate immune defence Nature Reviews Immunology 2007 7 3 179 190 10.1038/nri2038 2-s2.0-33847183077 17 Jiménez-Dalmaroni M. J. Gerswhin M. E. Adamopoulos I. E. The critical role of toll-like receptors—from microbial recognition to autoimmunity: a comprehensive review Autoimmunity Reviews 2016 15 1 1 8 10.1016/j.autrev.2015.08.009 2-s2.0-84952908271 26299984 18 Hennessy E. J. Parker A. E. O'Neill L. A. J. Targeting Toll-like receptors: emerging therapeutics? Nature Reviews Drug Discovery 2010 9 4 293 307 10.1038/nrd3203 2-s2.0-77950542752 20380038 19 Tseng P.-H. Matsuzawa A. Zhang W. Mino T. Vignali D. A. A. Karin M. Different modes of ubiquitination of the adaptor TRAF3 selectively activate the expression of type I interferons and proinflammatory cytokines Nature Immunology 2010 11 1 70 75 10.1038/ni.1819 2-s2.0-74049136127 19898473 20 Wertz I. E. Dixit V. M. Signaling to NF-kappaB: regulation by ubiquitination Cold Spring Harbor Perspectives in Biology 2010 2 3 a003350 10.1101/cshperspect.a003350 2-s2.0-77957284071 21 Wartz I. E. O'Rourke K. M. Zhou H. De-ubiquitination and ubiquitin ligase domains of A20 downregulate NF-κ B signalling Nature 2004 430 7000 694 699 10.1038/nature02794 2-s2.0-3943054838 15258597 22 Skaug B. Chen J. Du F. He J. Ma A. Chen Z. J. Direct, noncatalytic mechanism of IKK inhibition by A20 Molecular Cell 2011 44 4 559 571 10.1016/j.molcel.2011.09.015 2-s2.0-81355150892 22099304 23 Fulda S. Rajalingam K. Dikic I. Ubiquitylation in immune disorders and cancer: from molecular mechanisms to therapeutic implications EMBO Molecular Medicine 2012 4 7 545 556 10.1002/emmm.201100707 2-s2.0-84863454687 22730341 24 Colleran A. Collins P. E. O'Carroll C. Deubiquitination of NF-kappaB by Ubiquitin-Specific Protease-7 promotes transcription Proceedings of the National Academy of Sciences of the United States of America 2013 110 2 618 623 23267096 25 Li T. Guan J. Li S. Zhang X. Zheng X. HSCARG downregulates NF-κ B signaling by interacting with USP7 and inhibiting NEMO ubiquitination Cell Death and Disease 2014 5 5 e1230 10.1038/cddis.2014.197 2-s2.0-84901048218 26 Wang W. Huang X. Xin H.-B. Fu M. Xue A. Wu Z.-H. TRAF family member-associated NF-κ B activator (TANK) inhibits genotoxic nuclear factor κ B activation by facilitating deubiquitinase USP10-dependent deubiquitination of TRAF6 ligase The Journal of Biological Chemistry 2015 290 21 13372 13385 10.1074/jbc.m115.643767 2-s2.0-84930010074 25861989 27 Ritchie K. J. Hahn C. S. Kim K. I. Role of ISG15 protease UBP43 (USP18) in innate immunity to viral infection Nature Medicine 2004 10 12 1374 1378 10.1038/nm1133 2-s2.0-11144294823 28 Yang Z. Xian H. Hu J. USP18 negatively regulates NF-κ B signaling by targeting TAK1 and NEMO for deubiquitination through distinct mechanisms Scientific Reports 2015 5 12738 10.1038/srep12738 2-s2.0-84938809839 29 Chen G. Shaw M. H. Kim Y.-G. Nuñez G. NOD-like receptors: role in innate immunity and inflammatory disease Annual Review of Pathology: Mechanisms of Disease 2009 4 365 398 10.1146/annurev.pathol.4.110807.092239 2-s2.0-66149086936 30 Damgaard R. B. Nachbur U. Yabal M. The ubiquitin ligase XIAP recruits LUBAC for NOD2 signaling in inflammation and innate immunity Molecular Cell 2012 46 6 746 758 10.1016/j.molcel.2012.04.014 2-s2.0-84863000898 22607974 31 Bertrand M. J. M. Doiron K. Labbé K. Korneluk R. G. Barker P. A. Saleh M. Cellular inhibitors of apoptosis cIAP1 and cIAP2 are required for innate immunity signaling by the pattern recognition receptors NOD1 and NOD2 Immunity 2009 30 6 789 801 10.1016/j.immuni.2009.04.011 2-s2.0-66949138341 19464198 32 Ver Heul A. M. Fowler C. A. Ramaswamy S. Piper R. C. Ubiquitin regulates caspase recruitment domain-mediated signaling by nucleotide-binding oligomerization domain-containing proteins NOD1 and NOD2 The Journal of Biological Chemistry 2013 288 10 6890 6902 10.1074/jbc.m112.413781 2-s2.0-84874857071 23300079 33 Zurek B. Schoultz I. Neerincx A. TRIM27 negatively regulates NOD2 by ubiquitination and proteasomal degradation PLoS ONE 2012 7 7 e41255 10.1371/journal.pone.0041255 2-s2.0-84864053162 34 Hitotsumatsu O. Ahmad R.-C. Tavares R. The ubiquitin-editing enzyme A20 restricts nucleotide-binding oligomerization domain containing 2-triggered signals Immunity 2008 28 3 381 390 10.1016/j.immuni.2008.02.002 2-s2.0-40449136493 18342009 35 Fiil B. K. Damgaard R. B. Wagner S. A. OTULIN restricts Met1-linked ubiquitination to control innate immune signaling Molecular Cell 2013 50 6 818 830 10.1016/j.molcel.2013.06.004 2-s2.0-84880851215 23806334 36 Latz E. Xiao T. S. Stutz A. Activation and regulation of the inflammasomes Nature Reviews Immunology 2013 13 6 397 411 10.1038/nri3452 2-s2.0-84878237993 37 Muñoz-Planillo R. Kuffa P. Martínez-Colón G. Smith B. Rajendiran T. Núñez G. K+ efflux is the common trigger of NLRP3 inflammasome activation by bacterial toxins and particulate matter Immunity 2013 38 6 1142 1153 10.1016/j.immuni.2013.05.016 2-s2.0-84879596906 23809161 38 Yan Y. Jiang W. Liu L. Dopamine controls systemic inflammation through inhibition of NLRP3 inflammasome Cell 2015 160 1-2 62 73 10.1016/j.cell.2014.11.047 2-s2.0-84920990887 25594175 39 Han S. Lear T. B. Jerome J. A. Lipopolysaccharide primes the NALP3 inflammasome by inhibiting its ubiquitination and degradation mediated by the SCFFBXL2 E3 ligase The Journal of Biological Chemistry 2015 290 29 18124 18133 10.1074/jbc.M115.645549 26037928 40 Hu Y. Mao K. Zeng Y. Tripartite-motif protein 30 negatively regulates NLRP3 inflammasome activation by modulating reactive oxygen species production The Journal of Immunology 2010 185 12 7699 7705 10.4049/jimmunol.1001099 2-s2.0-78650636576 21048113 41 Weng L. Mitoma H. Tricot C. The E3 ubiquitin ligase tripartite motif 33 is essential for cytosolic RNA-Induced NLRP3 inflammasome activation The Journal of Immunology 2014 193 7 3676 3682 10.4049/jimmunol.1401448 2-s2.0-84907210913 25172487 42 Vince J. E. Wong W. W.-L. Gentle I. Inhibitor of apoptosis proteins limit RIP3 kinase-dependent interleukin-1 activation Immunity 2012 36 2 215 227 10.1016/j.immuni.2012.01.012 2-s2.0-84857404572 22365665 43 Labbé K. McIntire C. R. Doiron K. Leblanc P. M. Saleh M. Cellular inhibitors of apoptosis proteins cIAP1 and cIAP2 are required for efficient caspase-1 activation by the inflammasome Immunity 2011 35 6 897 907 10.1016/j.immuni.2011.10.016 2-s2.0-84255178199 22195745 44 Van Opdenbosch N. Gurung P. Vande Walle L. Fossoul A. Kanneganti T.-D. Lamkanfi M. Activation of the NLRP1b inflammasome independently of ASC-mediated caspase-1 autoproteolysis and speck formation Nature Communications 2014 5, article 3209 10.1038/ncomms4209 2-s2.0-84893835896 45 Shi C.-S. Shenderov K. Huang N.-N. Activation of autophagy by inflammatory signals limits IL-1β production by targeting ubiquitinated inflammasomes for destruction Nature Immunology 2012 13 3 255 263 10.1038/ni.2215 2-s2.0-84857195479 22286270 46 Guan K. Wei C. Zheng Z. MAVS promotes inflammasome activation by targeting ASC for K63-linked ubiquitination via the E3 Ligase TRAF3 The Journal of Immunology 2015 194 10 4880 4890 10.4049/jimmunol.1402851 2-s2.0-84929119017 25847972 47 Chiu H.-W. Chen C.-H. Chang J.-N. Chen C.-H. Hsu Y.-H. Far-infrared promotes burn wound healing by suppressing NLRP3 inflammasome caused by enhanced autophagy Journal of Molecular Medicine 2016 94 7 809 819 10.1007/s00109-016-1389-0 2-s2.0-84957634376 26864306 48 Rodgers M. A. Bowman J. W. Fujita H. The linear ubiquitin assembly complex (LUBAC) is essential for NLRP3 inflammasome activation Journal of Experimental Medicine 2014 211 7 1333 1347 10.1084/jem.20132486 2-s2.0-84903761551 24958845 49 Gurung P. Lamkanfi M. Kanneganti T.-D. Cutting edge: SHARPIN is required for optimal NLRP3 inflammasome activation The Journal of Immunology 2015 194 5 2064 2067 10.4049/jimmunol.1402951 2-s2.0-84924358739 25637014 50 Juliana C. Fernandes-Alnemri T. Kang S. Farias A. Qin F. Alnemri E. S. Non-transcriptional priming and deubiquitination regulate NLRP3 inflammasome activation The Journal of Biological Chemistry 2012 287 43 36617 36622 10.1074/jbc.m112.407130 2-s2.0-84867770402 22948162 51 Py B. F. Kim M.-S. Vakifahmetoglu-Norberg H. Yuan J. Deubiquitination of NLRP3 by BRCC3 critically regulates inflammasome activity Molecular Cell 2013 49 2 331 338 10.1016/j.molcel.2012.11.009 2-s2.0-84872782298 23246432 52 Lopez-Castejon G. Luheshi N. M. Compan V. Deubiquitinases regulate the activity of caspase-1 and interleukin-1β secretion via assembly of the inflammasome The Journal of Biological Chemistry 2013 288 4 2721 2733 10.1074/jbc.m112.422238 2-s2.0-84873834109 23209292 53 Hwang I. Lee E. Jeon S.-A. Yu J.-W. Histone deacetylase 6 negatively regulates NLRP3 inflammasome activation Biochemical and Biophysical Research Communications 2015 467 4 973 978 10.1016/j.bbrc.2015.10.033 2-s2.0-84946564306 26471297 54 Juliana C. Fernandes-Alnemri T. Wu J. Anti-inflammatory compounds parthenolide and bay 11-7082 are direct inhibitors of the inflammasome The Journal of Biological Chemistry 2010 285 13 9792 9802 10.1074/jbc.m109.082305 2-s2.0-77951240589 20093358 55 Strickson S. Campbell D. G. Emmerich C. H. The anti-inflammatory drug BAY 11-7082 suppresses the MyD88-dependent signalling network by targeting the ubiquitin system Biochemical Journal 2013 451 3 427 437 10.1042/BJ20121651 2-s2.0-84876240899 23441730 56 Ritorto M. S. Ewan R. Perez-Oliva A. B. Screening of DUB activity and specificity by MALDI-TOF mass spectrometry Nature Communications 2014 5, article 4763 10.1038/ncomms5763 2-s2.0-84907358982 57 Duong B. H. Onizawa M. Oses-Prieto J. A. A20 restricts ubiquitination of pro-interleukin-1β protein complexes and suppresses NLRP3 inflammasome activity Immunity 2015 42 1 55 67 10.1016/j.immuni.2014.12.031 2-s2.0-84921305910 25607459 58 Vande Walle L. Van Opdenbosch N. Jacques P. Negative regulation of the NLRP3 inflammasome by A20 protects against arthritis Nature 2014 512 1 69 73 10.1038/nature13322 2-s2.0-84905562455 25043000 59 Edelmann M. J. Kessler B. M. Ubiquitin and ubiquitin-like specific proteases targeted by infectious pathogens: emerging patterns and molecular principles Biochimica et Biophysica Acta (BBA)—Molecular Basis of Disease 2008 1782 12 809 816 10.1016/j.bbadis.2008.08.010 2-s2.0-56749161705 18817868 60 Neish A. S. Gewirtz A. T. Zeng H. Prokaryotic regulation of epithelial responses by inhibition of Iκ B-α ubiquitination Science 2000 289 5484 1560 1563 10.1126/science.289.5484.1560 2-s2.0-0034284803 10968793 61 Collier-Hyams L. S. Zeng H. Sun J. Cutting edge: Salmonella AvrA effector inhibits the key proinflammatory, anti-apoptotic NF-κ B pathway The Journal of Immunology 2002 169 6 2846 2850 10.4049/jimmunol.169.6.2846 2-s2.0-0037105562 12218096 62 Ye Z. Petrof E. O. Boone D. Claud E. C. Sun J. Salmonella effector AvrA regulation of colonic epithelial cell inflammation by deubiquitination The American Journal of Pathology 2007 171 3 882 892 10.2353/ajpath.2007.070220 17690189 63 Lu R. Wu S. Zhang Y.-G. Enteric bacterial protein AvrA promotes colonic tumorigenesis and activates colonic beta-catenin signaling pathway Oncogenesis 2014 3, article e105 10.1038/oncsis.2014.20 2-s2.0-84902243899 64 Misaghi S. Balsara Z. R. Catic A. Spooner E. Ploegh H. L. Starnbach M. N. Chlamydia trachomatis-derived deubiquitinating enzymes in mammalian cells during infection Molecular Microbiology 2006 61 1 142 150 10.1111/j.1365-2958.2006.05199.x 2-s2.0-33745186215 16824101 65 Le Negrate G. Krieg A. Faustin B. ChlaDub1 of Chlamydia trachomatis suppresses NF-κ B activation and inhibits Iκ Bα ubiquitination and degradation Cellular Microbiology 2008 10 9 1879 1892 10.1111/j.1462-5822.2008.01178.x 2-s2.0-48749090431 18503636 66 Kummari E. Alugubelly N. Hsu C. Y. Dong B. Nanduri B. Edelmann M. J. Activity-based proteomic profiling of deubiquitinating enzymes in salmonella-infected macrophages leads to identification of putative function of UCH-L5 in inflammasome regulation PLoS ONE 2015 10 8 e0135531 10.1371/journal.pone.0135531 67 Yen H. Sugimoto N. Tobe T. Enteropathogenic Escherichia coli uses NleA to inhibit NLRP3 inflammasome activation PLoS Pathogens 2015 11 9 23 e1005121 10.1371/journal.ppat.1005121 2-s2.0-84943523865 68 Heideker J. Wertz I. E. DUBs, the regulation of cell identity and disease Biochemical Journal 2015 465 1 1 26 10.1042/bj20140496 2-s2.0-84918538847 25631680 69 Brummelkamp T. R. Nijman S. M. B. Dirac A. M. G. Bernards R. Loss of the cylindromatosis tumour suppressor inhibits apoptosis by activating NF-κ B Nature 2003 424 6950 797 801 10.1038/nature01811 2-s2.0-0042467554 12917690 70 Kovalenko A. Chable-Bessia C. Cantarella G. Israël A. Wallach D. Courtois G. The tumour suppressor CYLD negatively regulates NF-κ B signalling by deubiquitination Nature 2003 424 6950 801 805 10.1038/nature01802 2-s2.0-0041967054 12917691 71 Trompouki E. Hatzivassillou E. Tsichritzis T. Farmer H. Ashworth A. Mosialos G. CYLD is a deubiquitinating enzyme that negatively regulates NF-κ B activation by TNFR family members Nature 2003 424 6950 793 796 10.1038/nature01803 2-s2.0-0042467558 12917689 72 Zhang J. Stirling B. Temmerman S. T. Impaired regulation of NF-κ B and increased susceptibility to colitis-associated tumorigenesis in CYLD-deficient mice The Journal of Clinical Investigation 2006 116 11 3042 3049 10.1172/jci28746 2-s2.0-33750598021 17053834 73 Courtois G. Tumor suppressor CYLD: negative regulation of NF-κ B signaling and more Cellular and Molecular Life Sciences 2008 65 7-8 1123 1132 10.1007/s00018-007-7465-4 2-s2.0-42449146639 18193168 74 Schuijs M. J. Willart M. A. Vergote K. Farm dust and endotoxin protect against allergy through A20 induction in lung epithelial cells Science 2015 349 6252 1106 1110 10.1126/science.aac6623 2-s2.0-84941102754 26339029 75 Ma A. Malynn B. A. A20: linking a complex regulator of ubiquitylation to immunity and human disease Nature Reviews Immunology 2012 12 11 774 785 10.1038/nri3313 2-s2.0-84867898782 76 Boone D. L. Turer E. E. Lee E. G. The ubiquitin-modifying enzyme A20 is required for termination of Toll-like receptor responses Nature Immunology 2004 5 10 1052 1060 10.1038/ni1110 2-s2.0-5444223519 15334086 77 Liu X. Li H. Zhong B. Usp18 inhibits NF-κ B and NFAT activation during Th17 differentiation by deubiquitinating the TAK1-TAB1 complex The Journal of Experimental Medicine 2013 210 8 1575 1590 10.1084/jem.20122327 2-s2.0-84884234601 23825189 78 Brooks C. L. Li M. Hu M. Shi Y. Gu W. The p53—Mdm2—HAUSP complex is involved in p53 stabilization by HAUSP Oncogene 2007 26 51 7262 7266 10.1038/sj.onc.1210531 2-s2.0-36049010619 17525743 79 Boutell C. Canning M. Orr A. Everett R. D. Reciprocal activities between herpes simplex virus type 1 regulatory protein ICP0, a ubiquitin E3 ligase, and ubiquitin-specific protease USP7 Journal of Virology 2005 79 19 12342 12354 10.1128/JVI.79.19.12342-12354.2005 2-s2.0-25144467949 16160161 80 Holowaty M. N. Frappier L. HAUSP/USP7 as an Epstein-Barr virus target Biochemical Society Transactions 2004 32 5 731 732 10.1042/BST0320731 2-s2.0-8744284703 15494000 81 Chavoshi S. Egorova O. Lacdao I. K. Farhadi S. Sheng Y. Saridakis V. Identification of kaposi sarcoma herpesvirus (KSHV) vIRF1 protein as a novel interaction partner of human deubiquitinase USP7 Journal of Biological Chemistry 2016 291 12 6281 6291 10.1074/jbc.m115.710632 26786098 82 Kon N. Kobayashi Y. Li M. Brooks C. L. Ludwig T. Gu W. Inactivation of HAUSP in vivo modulates p53 function Oncogene 2010 29 9 1270 1279 10.1038/onc.2009.427 2-s2.0-77949267635 19946331 83 Adams D. H. Ruzehaji N. Strudwick X. L. Attenuation of Flightless I, an actin-remodelling protein, improves burn injury repair via modulation of transforming growth factor (TGF)-β 1 and TGF-β 3 British Journal of Dermatology 2009 161 2 326 336 10.1111/j.1365-2133.2009.09296.x 2-s2.0-73349109483 19519830 84 Nawrocki S. T. Griffin P. Kelly K. R. Carew J. S. MLN4924: a novel first-in-class inhibitor of NEDD8-activating enzyme for cancer therapy Expert Opinion on Investigational Drugs 2012 21 10 1563 1573 10.1517/13543784.2012.707192 2-s2.0-84866532722 22799561 85 Beug S. T. Tang V. A. LaCasse E. C. Smac mimetics and innate immune stimuli synergize to promote tumor death Nature Biotechnology 2014 32 2 182 190 10.1038/nbt.2806 2-s2.0-84893825107 86 Farshi P. Deshmukh R. R. Nwankwo J. O. Deubiquitinases (DUBs) and DUB inhibitors: a patent review Expert Opinion on Therapeutic Patents 2015 25 10 1191 1208 10.1517/13543776.2015.1056737 2-s2.0-84942107490 26077642 87 Kemp M. Recent advances in the discovery of deubiquitinating enzyme inhibitors Progress in Medicinal Chemistry 2016 55 149 192 10.1016/bs.pmch.2015.10.002 26852935 88 Fan Y.-H. Cheng J. Vasudevan S. A. USP7 inhibitor P22077 inhibits neuroblastoma growth via inducing p53-mediated apoptosis Cell Death and Disease 2013 4 10, article e867 10.1038/cddis.2013.400 2-s2.0-84887471999 89 He X. Li Y. Li C. USP2a negatively regulates IL-1β - and virus-induced NF-κ B activation by deubiquitinating TRAF6 Journal of Molecular Cell Biology 2013 5 1 39 47 10.1093/jmcb/mjs024 2-s2.0-84874332545 22611252 90 Fan Y.-H. Yu Y. Mao R.-F. USP4 targets TAK1 to downregulate TNFα -induced NF-κ B activation Cell Death and Differentiation 2011 18 10 1547 1560 10.1038/cdd.2011.11 2-s2.0-80052691620 21331078 91 (IMPC) I.M.P.C., http://www.mousephenotype.org/data/genes/MGI:894652 92 Schweitzer K. Bozko P. M. Dubiel W. Naumann M. CSN controls NF-κ B by deubiquitinylation of Iκ Bα The EMBO Journal 2007 26 6 1532 1541 10.1038/sj.emboj.7601600 2-s2.0-33947606938 17318178 93 Jean-Charles P. Y. Zhang L. Wu J. Ubiquitin-specific protease 20 regulates the reciprocal functions of β -arrestin2 in toll-like receptor 4-promoted nuclear factor κ B (NFκ B) activation Journal of Biological Chemistry 2016 291 14 7450 7464 10.1074/jbc.m115.687129 26839314 94 Xu G. Tan X. Wang H. Ubiquitin-specific peptidase 21 inhibits tumor necrosis factor α -induced nuclear factor κ B activation via binding to and deubiquitinating receptor-interacting protein 1 Journal of Biological Chemistry 2010 285 2 969 978 10.1074/jbc.M109.042689 2-s2.0-74049114641 19910467 95 Fan Y. Mao R. Yu Y. USP21 negatively regulates antiviral response by acting as a RIG-I deubiquitinase Journal of Experimental Medicine 2014 211 2 313 328 10.1084/jem.20122844 2-s2.0-84893721948 24493797 96 Zhong B. Liu X. Wang X. Ubiquitin-specific protease 25 regulates TLR4-dependent innate immune responses through deubiquitination of the adaptor protein TRAF3 Science Signaling 2013 6 275, article ra35 10.1126/scisignal.2003708 2-s2.0-84877973388 97 Lin D. Zhang M. Zhang M.-X. Induction of USP25 by viral infection promotes innate antiviral responses by mediating the stabilization of TRAF3 and TRAF6 Proceedings of the National Academy of Sciences of the United States of America 2015 112 36 11324 11329 10.1073/pnas.1509968112 2-s2.0-84941248283 26305951 98 Tzimas C. Michailidou G. Arsenakis M. Kieff E. Mosialos G. Hatzivassiliou E. G. Human ubiquitin specific protease 31 is a deubiquitinating enzyme implicated in activation of nuclear factor-κ B Cellular Signalling 2006 18 1 83 92 10.1016/j.cellsig.2005.03.017 2-s2.0-26244434134 16214042 99 Draber P. Kupka S. Reichert M. LUBAC-recruited CYLD and A20 regulate gene activation and cell death by exerting opposing effects on linear ubiquitin in signaling complexes Cell Reports 2015 13 10 2258 2272 10.1016/j.celrep.2015.11.009 26670046 100 Verhelst K. Carpentier I. Kreike M. A20 inhibits LUBAC-mediated NF-κ B activation by binding linear polyubiquitin chains via its zinc finger 7 The EMBO Journal 2012 31 19 3845 3855 10.1038/emboj.2012.240 2-s2.0-84867062977 23032186 101 Enesa K. Zakkar M. Chaudhury H. NF-κ B suppression by the deubiquitinating enzyme cezanne: a novel negative feedback loop in pro-inflammatory signaling Journal of Biological Chemistry 2008 283 11 7036 7045 10.1074/jbc.m708690200 2-s2.0-43749122598 18178551 102 Hu H. Brittain G. C. Chang J.-H. OTUD7B controls non-canonical NF-κ B activation through deubiquitination of TRAF3 Nature 2013 494 7437 371 374 10.1038/nature11831 2-s2.0-84874193578 23334419 103 Schaeffer V. Akutsu M. Olma M. H. Gomes L. C. Kawasaki M. Dikic I. Binding of OTULIN to the PUB domain of HOIP controls NF-κ B signaling Molecular Cell 2014 54 3 349 361 10.1016/j.molcel.2014.03.016 2-s2.0-84899982792 24726327 104 Rivkin E. Almeida S. M. Ceccarelli D. F. The linear ubiquitin-specific deubiquitinase gumby regulates angiogenesis Nature 2013 498 7454 318 324 10.1038/nature12296 2-s2.0-84879390723 23708998 105 Elliott P. R. Nielsen S. V. Marco-Casanova P. Molecular basis and regulation of OTULIN-LUBAC interaction Molecular Cell 2014 54 3 335 348 10.1016/j.molcel.2014.03.018 2-s2.0-84899911195 24726323 106 Panda S. Nilsson J. A. Gekara N. O. Deubiquitinase MYSM1 regulates innate immunity through inactivation of TRAF3 and TRAF6 complexes Immunity 2015 43 4 647 659 10.1016/j.immuni.2015.09.010 2-s2.0-84944685637 26474655

PMC005xxxxxx/PMC5002300.txt

==== Front Stem Cells IntStem Cells IntSCIStem Cells International1687-966X1687-9678Hindawi Publishing Corporation 10.1155/2016/8432314EditorialStem Cells in Musculoskeletal Regeneration: From Benchtop to Bedside http://orcid.org/0000-0002-7485-8275Fan Jiabing 1 * Wang Dong-An 2 Liu Haifeng 3 http://orcid.org/0000-0002-7656-9530Fan Hongbin 4 http://orcid.org/0000-0002-4022-7643Yang Fang 5 1Division of Advanced Prosthodontics, School of Dentistry, University of California, Los Angeles, CA 90095, USA2Division of Bioengineering, Nanyang Technological University, Singapore 6374573Key Laboratory for Biomechanics and Mechanobiology of Ministry of Education, School of Biological Science and Medical Engineering, Beihang University, Beijing 100191, China4Institute of Orthopedic Surgery, Xijing Hospital, The Fourth Military Medical University, Xi'an 710032, China5Department of Biomaterials, Radboud University Medical Centre, 6500 HB Nijmegen, Netherlands*Jiabing Fan: jiabing2011@ucla.edu2016 14 8 2016 2016 84323143 7 2016 3 7 2016 Copyright © 2016 Jiabing Fan et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. ==== Body The continuing existent challenges of conventional surgical approaches for musculoskeletal reconstruction have led to the favorite of stem cell-mediated treatments over the last decade [1]. However, the in-depth understanding of underlying therapeutic mechanisms, regulatory mechanisms of stem cell differentiation, interaction of stem cells and inductive factors/delivery vehicles, and evaluation criteria in vitro and in vivo are indispensable to the wide practice of stem cells-based therapeutics for functionally restoring the musculoskeletal tissues from bench to bedside [2]. We are pleased to present a special issue encompassing both basic and translational aspects of stem cells-related research and review work in musculoskeletal regeneration including cartilage, bone, ligament, spinal cord, and meniscus, aiming to blossom in the effective use of stem cells for clinical application. The proper cell source is one of the determinant elements to the success of stem cells-based tissue regeneration [3]. Mesenchymal stem cells (MSCs) hold the high accessibility as well as capacity of self-renewal and multidifferentiation, representing an attractive cell source for musculoskeletal tissue engineering [3]. In this special issue, S. Morikawa et al. presented a comprehensive review work recapitulating fundamental biology of MSCs and highlighting that neural crest may be a new cell pool attributable to the harvest of MSCs in favor of craniofacial bone repair. And also S. Liu et al. reviewed a large number of literatures to discuss recent advances in meniscus tissue engineering using cell-based strategy, suggesting that scaffold-free cell self-assembly method is a potential approach to manufacture a functional meniscus graft with robust mechanical properties. Another review work conducted by Z. Deng et al. concluded that currently matrix-assisted autologous chondrocyte transplantation is an optimal approach to the repair of cartilage defects after comparing the efficacy and safety of various tissue engineering approaches through a systemic review and meta-analyses. This review also indicated that there are no enough studies collected to verify the efficacy of MSC-based treatment in cartilage repair. Natural, synthetic, or nanoscale scaffolds, as delivery vehicles, have been widely employed to carrier stem cells or inductive factors (gene, protein, and DNA) for tissue repair [4]. However, it necessitates the further enhancement of scaffold in both conductivity and inductivity [4]. Q. Li et al. conducted a comparative assessment of how two calcium phosphate/collagen composite materials affect osteogenic differentiation of adipose-derived stem cells (ASCs) which are thought to be a promising cell source in bone tissue engineering, demonstrating the hydroxyapatite/β-tricalcium phosphate composite scaffold is a better stimulator for ASCs proliferation and osteogenesis. To overcome the worrisome complications likely caused by exogenous FDA-approved bone morphogenetic protein 2 (BMP-2) in bone repair, Q. Xie et al. developed a core-shell PEI (polyethylenimine)/pBMP2- (plasmid BMP2-) PLGA (poly(lactic-co-glycolic acid)) electrospun scaffold using both gene modification and coaxial electrospinning techniques, which was revealed to be capable of sustaining expression of BMP-2 as well as enhancing the osteogenic differentiation of periodontal ligament stem cells. The ultimate goal of fundamental studies involving stem cells and musculoskeletal regeneration is to translate stem cells to clinical application [5]. In a research article, K. Yaghoobi et al. discovered that herbal drug, Lavandula angustifolia, significantly promoted human umbilical mesenchymal Wharton's jelly stem cells in the treatment of spinal cord injury created in Wistar rats. Additionally, Z. Kakabadze et al. who are working in several well-known research institutes around world carried out a pioneering phase 1 clinical trial using autologous human bone marrow stem cell transplantation to treat patients with spinal cord injury. Through evaluation of 18 patients after treatment, the therapeutic effects are encouraging in the recovery of spinal cord injury, but further improvements such as establishment of a standard of in vivo evaluation are still needed in future investigation. Together, these findings may provide an alternative approach to the use of stem cell-mediated therapy for the restoration of spinal cord injury that is still a challenge in the current clinical treatment. Furthermore, L. Sun et al. performed a profound work in exploring the effect of mechanical stretch on proliferation and matrix formation of BMSCs and anterior cruciate ligament fibroblasts, two types of cell identified as major seed cells serving ligament reconstruction. The outcomes of this study are also reminiscent of mechanical property that may be emphasized in the following study of stem cells-mediated tissue engineering. Lastly, R. J. F. C. Amaral et al. reported that human blood collected with an anticoagulant of sodium citrate may yield higher amount of human platelet-rich plasma (PRP) and exert higher proliferation of MSCs. The interesting findings suggest PRP as a potential supplement to promote MSC proliferation and differentiation. In summary, the cutting-edge review and research articles presented by experts in the field of orthopedic surgery, stem cells, and tissue engineering were collected to be published in this special issue, prospectively being a cornerstone to spur stem cell therapy to be applied to musculoskeletal regeneration in clinic. Jiabing Fan Jiabing Fan Dong-An Wang Dong-An Wang Haifeng Liu Haifeng Liu Hongbin Fan Hongbin Fan Fang Yang Fang Yang ==== Refs 1 Tuan R. S. Regenerative medicine in 2012: the coming of age of musculoskeletal tissue engineering Nature Reviews Rheumatology 2013 9 2 74 76 10.1038/nrrheum.2012.235 23321611 2 Dawson J. I. Kanczler J. Tare R. Kassem M. Oreffo R. O. C. Concise review: bridging the gap: bone regeneration using skeletal stem cell-based strategies—where are we now? STEM CELLS 2014 32 1 35 44 10.1002/stem.1559 2-s2.0-84891750619 24115290 3 Fan J. Varshney R. R. Ren L. Cai D. Wang D.-A. Synovium-derived mesenchymal stem cells: a new cell source for musculoskeletal regeneration Tissue Engineering Part B: Reviews 2009 15 1 75 86 10.1089/ten.teb.2008.0586 2-s2.0-65549095406 19196118 4 Fan H. Liu H. Toh S. L. Goh J. C. H. Anterior cruciate ligament regeneration using mesenchymal stem cells and silk scaffold in large animal model Biomaterials 2009 30 28 4967 4977 10.1016/j.biomaterials.2009.05.048 2-s2.0-67849094106 19539988 5 Hunsberger J. Harrysson O. Shirwaiker R. Manufacturing road map for tissue engineering and regenerative medicine technologies Stem Cells Translational Medicine 2015 4 2 130 135 10.5966/sctm.2014-0254 2-s2.0-84921807245 25575525

PMC005xxxxxx/PMC5002301.txt

==== Front Biomed Res IntBiomed Res IntBMRIBioMed Research International2314-61332314-6141Hindawi Publishing Corporation 10.1155/2016/8495630Research ArticlePhytochemical Contents and Antioxidant and Antiproliferative Activities of Selected Black and White Sesame Seeds http://orcid.org/0000-0003-4318-5132Zhou Lin 1 2 * http://orcid.org/0000-0003-3848-7048Lin Xiaohui 2 Abbasi Arshad Mehmood 2 3 Zheng Bisheng 2 1Guangdong Province Key Laboratory for Biotechnology Drug Candidates, School of Biosciences and Biopharmaceutics, Guangdong Pharmaceutical University, Guangzhou 510006, China2School of Light Industry and Food Sciences, South China University of Technology, Guangzhou 510641, China3Department of Environmental Sciences, COMSATS Institute of Information Technology, Abbottabad 22060, Pakistan*Lin Zhou: zhoulin@gdpu.edu.cnAcademic Editor: Michael D. Coleman 2016 14 8 2016 2016 849563015 4 2016 7 6 2016 30 6 2016 Copyright © 2016 Lin Zhou et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Sesame (Sesamum indicum L.) seeds are popular nutritional food but with limited knowledge about their antioxidant and antiproliferative activities of various varieties. Phytochemical profiles and antioxidant and antiproliferative activities of six varieties of sesame (Sesamum indicum L.) seeds were studied. Fenheizhi3 (black) cultivar exhibited the maximum contents of total phenolics and lignans and values of total oxygen radical absorbance capacity (ORAC) and antiproliferative activity (EC50) against HepG2 cells. Bound ORAC values showed strong associations with bound phenolics contents (r = 0.976, p < 0.01); in bound phenolic extracts, EC50 values showed strong negative associations with phenolic contents (r = −0.869, p < 0.05) and ORAC values (r = −0.918, p < 0.01). Moreover, the contents of free phenolics were higher than that of the bound phenolics, and the three black sesame seeds generally depicted higher total phenolics compared to the three white varieties. The antioxidant (ORAC values) and antiproliferation activities of six sesame seeds were both associated with contents of bound phenolics (r > 0.8, p < 0.05). Interestingly, nonlignan components in bound phenolics contributed to the antioxidant and antiproliferative activities. This study suggested that Fenheizhi3 variety is superior to the other five varieties as antioxidant supplements. Fundamental Research Funds for the Central Universities2015 D2155140Natural Science Foundation of Guangdong Province44555019 ==== Body 1. Introduction Sesame “Sesamum indicum L.” is a commonly growing plant species mainly in tropical and subtropical regions of the world, particularly in Burma, India, China, and Sudan [1]. Sesame seeds are preferably used along with bread, biscuits, crackers, and so forth and as seasoning in food around the world [2]. Sesame seeds are important sources of oil, protein, carbohydrates, and minerals for human nutrition [3]. Sesame seeds color varied from cream-white to charcoal-black, whereas white and black are the typical skin color. Black sesame seed is superior to white one as food for health in traditional beliefs of Asian countries and is included in Pharmacopoeia of the People's Republic of China (PPRC, 2015) as liver and kidney benefiting traditional Chinese medicine (TCM). It has been reported that the seed colors of sesame affect the phytochemical contents and their biological activities [4, 5]. Phytochemical compounds in sesame seed such as sesamin, sesamol, and anthrasesamone F have been proved to have in vitro/in vivo antioxidant and antiaging activity [6–9]. Moreover, sesamin and sesamolin showed anti-inflammatory, antihypertensive, and anticarcinogenic effects in numerous studies [10–12]. The common nutritional evaluations of sesame seeds are based on the contents of proteins, oils, and lignans. However, little work was conducted on the correlations between multicomponents and antioxidant activity and multicomponents and antiproliferative activity in various sesame seeds varieties. Therefore, the phytochemical profiles including the contents of phenolics, flavonoids, and lignans, in vitro antioxidant activities, and antiproliferative activities against HepG2 cells of six varieties of sesame seeds in China were studied in the present work. Particularly, the six sesame seeds (three of black and three of white skin color, nonscaled planting) are newly bred varieties in midwest of China for improved nutrition values. The knowledge of the antioxidant and antiproliferative activities of various sesame seeds will benefit the sesame seed planters, relevant manufacturers, and ordinary consumers. 2. Materials and Methods 2.1. Chemicals and Materials Methanol (MeOH), ethanol (EtOH), n-hexane, ethyl acetate, hydrochloric acid (HCl), and acetic acid (HAC) were purchased from Guanghua Sci-Tech Co., Ltd. (Guangdong, China). Potassium chloride (KCl), sodium acetate (NaAC), sodium carbonate (NaCO3), sodium hydroxide (NaOH), potassium phosphate monobasic (KH2PO4), and potassium phosphate dibasic (K2HPO4) were of analytical grade and were purchased from Sangon Biotech Co., Ltd. (Shanghai, China). 2,2′-Azobis(2-amidinopropane) dihydrochloride (ABAP), 2′,7′-dichlorofluorescin diacetate (DCFH-DA), fluorescein disodium salt, catechin hydrate (HPLC, ≥98%), Folin-Ciocalteu reagent, ascorbic acid, 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox), formic acid (chromatographic grade), and methanol (chromatographic grade) were purchased from Sigma-Aldrich, Inc. (St. Louis, MO, USA). Trifluoroacetic acid (TFA, analytical grade) and aluminum chloride (AlCl3·6H2O, analytical grade) were purchased from Fisher Scientific (Fair Lawn, NJ, USA). Sodium borohydride (NaBH4, reagent grade), chloranil (analytical grade), vanillin (analytical grade, 99%), and gallic acid (analytical grade, 99%) were purchased from Aladdin, Inc. (Shanghai, China). Sesamol, sesamolin, and sesamin (HPLC, ≥98%) were purchased from Chengdu Pufei De Biotech Co., Ltd. (Sichuan, China). HepG2 human liver cancer cells were purchased from American Type Culture Collection (Rockville, MD, USA). WME medium, foetal bovine serum (FBS), insulin, and other cell culture reagents were purchased from Gibco Life Technologies Co. (Grand Island, NY, USA). Six sesame seed varieties (three black and three white) were kindly donated by Shanxi Academy of Agricultural Sciences (Taiyuan city, China), which were harvested in 2014 and stored in desiccator at room temperature until analysis. 2.2. Moisture Content The moisture content was analyzed by oven-dry method [13]. Briefly, 1 g of sample was dried in electric oven at 105°C to a constant weight and moisture content was expressed as percent of dry weight (DW) in triplicate. 2.3. Phenolics Extraction and Determination 2.3.1. Free Phenolics Extraction Free phenolic compounds were extracted using the modified method reported previously [14]. Briefly, 1 g of sesame seed was blended with 5 mL of hexane for 1 min. After centrifugation at 2700 ×g for 3 min, residue was defatted two more times. Add 10 mL of 80% chilled acetone to the residue and blend 1 min. The supernatant was collected after being centrifuged at 2700 ×g for 3 min, repeated twice. The supernatant was pooled and evaporated to dryness at 45°C under vacuum. The solution was reconstituted in 10 mL of 70% methanol and free phenolic extracts (100 mg mL−1) were stored at −40°C until analysis. 2.3.2. Bound Phenolics Extraction Bound phenolics were extracted using the method described by Chen et al. [14] with modifications. Briefly, the residues obtained after extraction of free phenolics were flushed with nitrogen gas for 2 min, sealed, and digested with 20 mL of 4 M NaOH at room temperature for 1 h. The mixture was neutralized with 10 M concentrated HCl. Then 20 mL hexane was added to extract residual lipids in the mixture for 10 min (not necessary if no residual lipids). After centrifuging at 2700 ×g for 5 min, remaining residues were extracted five times with ethyl acetate. The ethyl acetate fractions were pooled and evaporated to dryness at 45°C under vacuum. The bound phenolics were reconstituted to 10 mL of 70% methanol and bound phenolic extracts (100 mg mL−1) were stored at −40°C until analysis. 2.3.3. Determination of Total Phenolic Contents Total phenolic contents of each sample were determined using the modified Folin-Ciocalteu colorimetric method [14, 15]. Briefly, free phenolic and bound phenolic extracts were diluted 10–20 times with distilled water and were reacted with Folin-Ciocalteu reagent and then neutralized with Na2CO3. After 90 min incubation, the absorbance of the resulting solution was recorded at 760 nm using MRX II Dynex plate reader (Dynex Technologies Inc., Chantilly, VA, USA). Total phenolic contents were expressed as grams of gallic acid equivalents per kg of sample on DW (g GAE kg−1). 2.4. Determination of Total Flavonoid Contents Total flavonoid contents were determined by sodium borohydride-chloranil (SBC) protocol [16]. Briefly, 1 mL of free and 1 mL of bound phenolic extracts were added to test tubes (15 × 150 mm) and were dried under nitrogen gas. The residues and catechin hydrate standards (0.1–10.0 mM) were prepared in 1 mL of THF/EtOH (1 : 1, v/v). Extracts or standard solution was mixed with 0.5 mL of 50 mM NaBH4 solution and 0.5 mL of 74.6 mM AlCl3 solution. After 30 min shaking in an orbital shaker at room temperature, 0.1 mL of 50.0 mM NaBH4 solution was added with continued shaking for 30 min at room temperature. 0.4 mL of cold 0.8 M acetic acid solution was added and the mixtures were kept in the dark for 15 min after thorough mixing. 0.2 mL of 20.0 mM chloranil was added and the mixture was heated at 95°C with shaking for 60 min. Then, the reaction solutions were cooled using tap water and were brought to 1 mL using methanol. Afterward, 0.2 mL of 16% (w/v) vanillin was added and mixed, followed by addition of 0.4 mL of 12 M HCl and the reaction solutions were kept in the dark for 15 min after thorough mixing. At last, the absorbance was recorded at 490 nm using MRX microplate reader with Revelation workstation (Dynex Technologies, Inc.). Total flavonoid contents were expressed as grams of catechin equivalents per kg of sample on DW (g CE kg−1). 2.5. Estimation of Lignan Contents Both the free and bound phenolic extracts (100 mg mL−1) were used for determination of lignan contents. The chromatographic analysis was performed to estimate lignan contents in sesame seeds as described by Reshma et al. [17] with some modifications. Briefly, three representative lignans (sesamol, sesamin, and sesamolin) were analyzed for the contents using Waters HPLC system (Waters Corp, Milford, MA) with a Waters C18 column (5 μm, 250 mm × 4.6 mm). The mobile phase consisted of methanol/water (75 : 25 v/v) with 0.1% formic acid in water at a flow rate of 1 mL min−1. The UV detector was set at 290 nm with column temperature of 30°C and injection volume of 10 μL. Peak was identified by calibration standards of sesamol, sesamin, and sesamolin with a retention time of 3.4, 8.3, and 10.2 min, respectively. The concentration range of standards of sesamol, sesamin, and sesamolin was from 2 to 200 μg mL−1 with recovery at 99.50 ± 1.00%, 99.47 ± 1.02%, and 99.62 ± 1.13%, respectively. The results were expressed as milligrams per kg on DW (mg kg−1). 2.6. Determination of Total Antioxidant Activity 2.6.1. Oxygen Radical Absorbance Capacity (ORAC) ORAC values were determined using a Fluoroskan Ascent fluorescent spectrophotometer (Molecular Devices, Sunnyvale, CA) [18, 19]. Briefly, 20 μL extracts (100 mg mL−1) or Trolox was mixed with 200 μL of fluorescein and incubated at 37°C for 20 min, followed by adding 20 μL of freshly prepared 119.4 mM ABAP into black walled 96-well plates. Fluorescence intensity was recorded automatically at 37°C, excitation of 485 nm, and emission of 535 nm for 35 cycles every 5 min. Final values of ORAC were expressed as micromoles of Trolox equivalents (TE) per gram of sesame seeds on DW basis (μmol TE g−1 DW). 2.6.2. Peroxyl Radical Scavenging Capacity (PSC) PSC values were determined by the method described by Adom and Liu [20]. Briefly, 100 μL extracts (100 mg mL−1) or standards solutions (freshly prepared ascorbic acid solutions) were added in each well of the 96-well plate. Then, 100 μL of DCFH dye (13.26 μM), hydrolyzing with 1 mM KOH just before being used in the reaction, was also added. After adding 50 μL of peroxyl radicals producer (ABAP, 40 mM), the mixtures were kept at 37°C for 40 min. Fluorescent was dynamically recorded at excitation of 485 nm and emission of 538 nm using a multimode microplate reader (FilterMax F5, Molecular Devices, USA). The results were calculated as micromoles of vitamin C equivalents (VCE) per gram of sesame seeds on DW basis (μmol VCE g−1 DW). 2.7. Determination of Antiproliferative Activity The antiproliferative effects of sesame seeds extracts were assessed by modified methylene blue assay [21]. Briefly, HepG2 human liver cancer cell was grown in WME medium supplemented with 5% FBS, 10 mM N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (HEPES), 2 mM L−1 glutamine, 5 g mL−1 insulin, 0.05 g mL−1 hydrocortisone, 50 units mL−1 penicillin, 50 g mL−1 streptomycin, and 100 g mL−1 gentamycin at 37°C and 5% CO2 as described by Liu et al. [22]. HepG2 cells were cultured at a density of 2.5 × 104cells/well in a 96-well microplate with 100 μL of growth medium. After 4 h of incubation at 37°C, growth medium was displaced by 100 μL of fresh medium containing 10–150 mg mL−1 of sesame seeds extracts. The wells containing fresh medium without extracts were set as the control. After 72 h of incubation, the number of the viable cells was counted by the methylene blue assay. The antiproliferative effects were assessed by the EC50 values and expressed as milligrams of sesame seeds extracts per milliliter. 2.8. Statistical Analysis Statistical analysis was performed using SPSS software version 19.0 (LEAD Technologies, Inc., Chicago, IL). Variation of means was analyzed by ANOVA and Duncan's test. Significance of correlations was calculated by the Pearson coefficient. Dose-Effect analysis was performed using CalcuSyn software, version 2.0 (Biosoft, Cambridge, UK). Statistical significance was set at p < 0.05. All data were expressed as the mean ± SD for three replications. 3. Results and Discussion 3.1. Moisture Content Moisture content and description of six varieties of sesame seeds are given in Table 1. Except B1 variety, the average moisture content of the other five sesame seed varieties ranged from 4.36 to 5.23%, which was consistent with the published data [3, 13]. The possible reason for the lower moisture content in B1 variety was the moisture loss during storage before transportation. 3.2. Phytochemical Composition of Selected Sesame Seeds 3.2.1. Contents of Phenolics and Flavonoids Measured levels of total phenolic and total flavonoid contents of six sesame seed varieties were presented in Figures 1(a) and 1(b). From Figure 1(a), free phenolic contents were higher than that of the bound phenolic. Black sesame seed variety B2 exhibited maximum free and bound phenolics (4.99 ± 0.47 and 2.33 ± 0.36 g GAE kg−1, resp.), while minimum amount was determined in free phenolics of W1 (2.58 ± 0.07 g GAE kg−1) and bound phenolics of W2 (0.80 ± 0.11 g GAE kg−1). In B2 cultivar, free phenolics were 68.19% and bound phenolics were 31.81% of total phenolic contents. Total phenolic contents in black sesame varieties (B1, B2, and B3) ranged from 4.54 to 7.32 g GAE kg−1 with maximum value in B2, whereas among white sesame seed W3 exhibited the highest total phenolic contents of 4.04 ± 0.13 g GAE kg−1, followed by W2 (4.00 ± 0.18 g GAE kg−1) and W1 (3.56 ± 0.08 g GAE kg−1) variety. On the whole, the black sesame varieties contain more phenolics compared to the white varieties. However, no significant differences between free, bound, and total phenolics were observed between black and white sesame varieties. Nadeem et al. [23] reported the total phenolics extracted from sesame cake were 1.72 g GAE kg−1. Shahidi et al. [4] reported that the total phenolic contents of two sesame seed cultivars (black and white) were 29.9 ± 0.6 and 10.6 ± 1.6 g catechin equivalents kg−1 crude ethanolic extract, respectively. These studies were not comparable with our data for different samples (sesame cake) or reference control (catechin) used in the determination. From Figure 1(b), the contents of free flavonoid ranged from 2.88 (B3) to 4.61 (W3) g CE kg−1, and the bound flavonoids were between 2.24 (W2) and 3.52 (B2) g CE kg−1, while total flavonoid contents ranged from 5.80 (B3) to 8.04 (W3) g CE kg−1. W3 variety exhibited the highest level of total flavonoids (8.04 ± 0.26) and free flavonoids (4.61 ± 0.06) g CE kg−1, while the highest levels of bound flavonoids were present in B2 (3.52 ± 0.08 g CE kg−1). On the whole, no significant differences of the flavonoid contents (free, bound, and total flavonoids) were observed between black and white sesame varieties. 3.2.2. Contents of Lignans Lignans here including sesamol, sesamin, and sesamolin in free and bound extracts of sesame seeds were analyzed by HPLC (Figure 2, Table 2). Overall, free phenolic extracts contain more levels of lignans (>89%) compared to the bound extracts. Sesamol and sesamolin contents were detected in the free extracts of all varieties. The lignans in free phenolic extracts of white and black sesame varieties ranged from 29.28 to 53.76 and 82.83 to 251.91 mg kg−1, respectively. The average lignan content of the three black cultivars (167.34 mg kg−1) was three times higher than that of the white cultivars (47.54 mg kg−1). B2 variety showed the highest sesamol (187.25 ± 10.56 mg kg−1) in free phenolic extracts, followed by B3 (121.48 ± 3.28 mg kg−1) and B1 (79.21 ± 3.62 mg kg−1). In white varieties, sesamol in free phenolic extracts ranged from 7.15 ± 0.75 (W2) to 34.58 ± 1.99 mg kg−1 (W1). Likewise, B2 variety had the highest sesamolin content (25.12 ± 0.95 mg kg−1) in free phenolic extracts, followed by B3 (10.30 ± 0.20 mg kg−1) and B1 (not detected). Sesamin was detected in the free extracts of four varieties with the highest value in W2 (39.57 ± 1.26 mg kg−1) and B2 (39.55 ± 0.38 mg kg−1), followed by W1 (14.25 ± 1.74 mg kg−1) and B3 (4.95 ± 0.20 mg kg−1). However, sesamin was not detected in bound phenolic extracts. It has been reported [24–26] that sesamin and sesamolin contents in sesame seeds ranged from 1550 to 4200 mg kg−1 and 620 to 3590 mg kg−1, respectively. In the present study, the possible reason for lower sesamin and sesamolin contents (<40 mg kg−1) was the extraction solvents. Although 70~80% methanol is common solvent for the extraction of total phenolics and flavonoids [14, 20], lipid-soluble lignans (sesamin and sesamolin) are partially soluble in 70% methanol based on visual observation. Likewise, the relatively higher content of sesamol compared to sesamin and sesamolin was observed in the free and bound extracts, which is possibly related to the relatively higher dissolving ability of sesamol in 70% methanol. Mostly previous studies were designed to determine the maximal content of lignans in sesame seeds, so the extraction organic solvents commonly used were weak polar ethanol, methanol, and acetonitrile. Here, our aim was to evaluate the lignan levels and their contribution to in vitro antioxidant activity in extracted fractions of sesame seeds, which leads to the relatively lower contents of lignans determined. Sesamin and sesamolin were successfully separated in setting chromatographic parameters (Figure 2), but the baseline separation of sesamol was not attained in the present work. 3.3. In Vitro Antioxidant Activity Due to the complex components in sesame seeds, two in vitro antioxidant activity assays (ORAC and PSC) were used in the present study. The common principle of determination was to measure radical chain breaking ability of antioxidants by monitoring the inhibition of peroxyl radical, while different oxidant-fluorescein probes were used: ABAP-fluorescein in ORAC assay and ABAP-DCFH in PSC assay [18, 20]. Measured values of ORAC and PSC based on DW are given in Figures 3(a) and 3(b). ORAC values ranged from 34.85 (B1) to 95.48 (W2) μmol TE g−1 in free fractions (Figure 3(a)), while in bound fractions it was between 14.97 (W2) and 40.74 (B3) μmol TE g−1. Black sesame variety B2 exhibited the highest total ORAC value (132.33 μmol TE g−1), whereas B1 showed the lowest total ORAC value (59.51 μmol TE g−1). White sesame variety W1 had the highest total PSC value among all varieties, which was 4 times higher than the lowest one (W2), followed by B3, B1, B2, W3, and W2 (Figure 3(b)). PSC value of free fractions ranged from 2.55 (W2) to 12.27 (B3) μmol VCE g−1, while in bound fractions it ranged from 1.33 (W3) to 6.45 (W1) μmol VCE g−1. The total PSC values ranged from 4.50 (W2) to 18.55 (W1) μmol VCE g−1. Ishiyama et al. [27] reported that ORAC values of two sesame seed varieties (Japan Kanto No. 1 and Gomazou) were 658.3 and 8.27 mg TE g−1, respectively. These values correspond to 26.30 and 33.04 μmol TE g−1 of sesame seed, which was comparable to the free fractions of B1 and B3 but was nearly one-fourth to one-third of the corresponding values of varieties B2, W1, W2, and W3. Othman et al. [28] recently reported that the average ORAC values of polar-soluble crude extracts for white and gold sesame seed were 347.20 and 217.00 μmol TE g−1, respectively. These reported data were nearly 2.6- and 1.6-fold higher than total ORAC value in B2 (the highest in this study), which verified the significant antioxidant activity of polar-soluble extracts. Correlation analysis of the in vitro antioxidant activity with the multicomponents of selected sesame seeds was shown in Table 3. Bound ORAC values showed strong associations with contents of bound phenolics (r = 0.976, p < 0.01). Moreover, total ORAC values showed strong association with contents of total phenolics (r = 0.701, p < 0.01). Moderate associations were observed between total ORAC values and contents of total flavonoids (r = 0.412, p < 0.05). However, associations between total lignans and antioxidant activities (ORAC and PSC values) were not observed. From ORAC values of B2, W1, W2, and W3, PSC values of B1, B3, W1, and W3, and their corresponding contents of lignans (Figure 3, Table 3), it can be inferred that the relatively higher antioxidant activities of free phenolic extracts compared to bound phenolic extracts were related to the relatively higher contents of lignans in free phenolic extracts. Comparable antioxidant activity was observed in PSC values of B2 and W2 variety and ORAC values of B2 and B3 variety in free and bound phenolic extracts (Figure 3, Table 3), even none or less lignans determined in bound phenolic extracts, which indicated that nonlignan components contribute to their corresponding antioxidant activities. There were no significant associations between PSC values (free, bound, and total) with corresponding contents of phenolic and flavonoid. Though specific reason is not clear, synergistic effects of various phytochemicals may involve complex antioxidant behavior of these constituents. Many similar reports were observed in selected components of sesame seed or sesame seed oil, such as synergistic antioxidant activity between lignans (sesamin, sesamol, sesamolin) and tocopherols (α, β, and γ) in various test system [29, 30]. Except the weak polar lignan components, the water soluble extracts of sesame seed raised concern. Moazzami et al. [31] quantified lignan glucosides in 65 different sesame seed cultivars, but no significant difference between black and white seeds was observed. Othman et al. [28] reported that extracts from white sesame seed had relatively higher antioxidant capacity compared to extracts from gold sesame seeds and ascribed the possible reason to the differences in the contents of antioxidants. Ide et al. [32] compared the physiological activities of sesame seeds with different concentration of lignans in rats and found that sesame seeds rich in lignans, irrespective of composition of lignans, greatly affect hepatic fatty acid oxidation and serum triacylglycerol levels. However, the researchers stated that there is a possibility that compounds other than lignans are involved in the physiological activities of tested sesame seeds. These reports indicate the complex antioxidant activity of multicomponents of sesame seeds. 3.4. Antiproliferation in HepG2 Cancer Cell Cell culture models provide efficient approach that imitates the uptake, distribution, and metabolism of antioxidant compounds in vivo [33], and human liver cancer HepG2 cell model is widely used in the antioxidant, antiproliferative activity of various compounds [14, 34, 35]. The antiproliferative effect of free and bound phenolic extracts against HepG2 cells was present in Table 4. The EC50 of free phenolic extracts ranged from 21.04 ± 0.60 to 124.91 ± 2.79 mg mL−1 for HepG2 cells, while corresponding values for bound phenolic extracts ranged from 23.57 ± 0.88 to 109.53 ± 1.23 mg mL−1. Lower values of EC50 indicate a higher antiproliferative activity. Both free and bound phenolic extracts of B2 variety showed antiproliferative activity. This result is consistent with the relatively higher contents of phenolics, flavonoids, and lignans and ORAC values of B2 variety. Interestingly, the black sesame seed varieties showed higher antiproliferative activity compared to the white varieties in both free phenolic and bound phenolic extracts. Moreover, in bound phenolic extracts, highly negative significant correlations were observed between EC50 and bound phenolics (r = −0.869, p < 0.05) and EC50 and bound ORAC values (r = −0.918, p < 0.01); in free phenolic extracts, highly negative significant correlations were observed between EC50 and sesamol contents (r = −0.858, p < 0.05) (data not shown). Sesamin and sesamol have been reported to have anticancer activities in many cancer cell lines including HepG2. Sesamin of 25–125 μM could induce HepG2 cell death involving apoptosis [36]. Content of sesamin in free phenolic extract of B2 variety was nearly 39.55 mg kg−1 DW, which is roughly equivalent to 11 μM sesamin. As for sesamol, at dosage of 50 μM, apoptosis against HepG2 cell could be observed by Liu et al. [37], while the content of sesamol in free phenolic extracts of B2 variety was nearly 187.25 mg kg−1 DW, which is roughly equivalent to 135 μM of sesamol. These results were consistent with the significant correlation observed between antiproliferation activity and content of sesamol in free phenolic extracts. Additionally, in bound phenolic extracts of B1, B2, W1, and W2 variety, contents of lignans (sesamol, sesamin, and sesamolin) were far less compared to their free phenolic extracts (Table 2). However, the corresponding antiproliferative activities in free phenolic extracts and bound phenolic extracts were comparable (Table 4). Thus, it can be inferred that the antiproliferative activities of sesame seed were not only related to the contents of phenolics and lignans (sesamol) of free phenolic extracts but also related to nonlignan components in bound phenolic extracts, which need further investigation. 4. Conclusions Sesame seed variety of Fenheizhi3 (black, B2) could be more valuable than the others (Zhenzhou1, 05H27, Jizhi157, Fenzhi2, and Jizhi1) based on the ex vivo antioxidant and antiproliferative activities. In selected sesame seeds, free phenolic contents were higher than the bound phenolic contents; black sesame seeds generally depicted higher contents of total phenolic compared to the white varieties. Moreover, the antioxidant activities (total ORAC values) were associated with the contents of total phenolics and flavonoids (r > 0.4, p < 0.05); the antiproliferation activities (bound EC50) were associated with contents of bound phenolics and sesamol (r > 0.8, p < 0.05). Additionally, nonlignan components in bound phenolic extracts contributed to the corresponding antioxidant and antiproliferative activities. Supplementary Material The inhibition of HepG2 cells proliferation by the free and bound fractions were observed in a dose-dependent manner in Figure S1. The EC50 of free phenolic extracts ranged from 21.04 ± 0.60 to 124.91 ± 2.79 mg mL− 1 for HepG2 cells, while corresponding values for bound phenolic extracts ranged from 23.57 ± 0.88 to 109.53 ± 1.23 mg mL− 1. Lower values of EC50 indicate a higher antiproliferative activity. Both free and bound phenolic extracts of B2 variety showed highest antiproliferative activity. Acknowledgments The authors are grateful for the financial support from the Fundamental Research Funds for the Central Universities (2015 D2155140) and Natural Science Foundation of Guangdong Province (44555019). Professor Rui hai Liu from Department of Food Science, Cornel University, provided technical support on the antioxidant determination. Competing Interests All authors declared that they have no conflict of interests. Authors' Contributions Lin Zhou designed and performed the experiments; Xiaohui Lin participated in the experiments; Arshad Mehmood Abbasi and Bisheng Zhen analyzed data and provided reagents, materials, and analytical tools. Lin Zhou provided financial support. Figure 1 Free, bound, and total phenolic and flavonoid contents in six varieties of sesame seeds (B1, B2, B3, W1, W2, and W3). Bars with no letters in common are significant difference at p < 0.05. (a) Phenolic contents. (b) Flavonoid contents. Figure 2 Elution curve of lignans (sesamol, sesamin, and sesamolin) of free phenolic extracts of B2 variety (mobile phase: methanol/water 75 : 25; flow rate: 1 mL min−1, wavenumber: 290 nm; and Waters C18 column). Figure 3 In vitro antioxidant activities of six sesame seed varieties (B1, B2, B3, W1, W2, and W3) (mean ± SD; n = 3). Bars with no letters in common are significant difference at p < 0.05. (a) ORAC values. (b) PSC value. Table 1 Description and moisture contents of six sesame seeds varieties. Varieties Abbreviation Harvest year Color Origin Moisture content (%) Zhenzhou1 B1 2014 Black Zhenzhou, Henan 1.63 ± 0.13 Fenheizhi3 B2 2014 Black Agriculture academy of Shanxi 5.14 ± 0.09 05H27 B3 2014 Black Shanxi Academy of Agricultural Sciences 5.22 ± 0.01 Jizhi157 W1 2014 White Hebei 4.50 ± 0.14 Fenzhi2 W2 2014 White Shanxi 4.63 ± 0.13 Jizhi1 W3 2014 White Shanxi Academy of Agricultural Sciences 4.63 ± 0.03 Data were reported as mean ± SD. Table 2 Lignans (sesamol, sesamin, and sesamolin) contents of free and bound phenolics extracts in six sesame seeds varieties. Varieties Sesamol (mg kg−1) Sesamin (mg kg−1) Sesamolin (mg kg−1) Total (mg kg−1) Free extracts Bound extracts Free extracts Bound extracts Free extracts Bound extracts B1 79.21 ± 3.62 7.14 ± 0.23 ND ND ND ND 86.35 B2 187.25 ± 10.56 30.54 ± 2.98 39.55 ± 0.38 ND 25.12 ± 0.95 ND 282.46 B3 121.48 ± 3.28 ND 4.95 ± 0.20 ND 10.30 ± 0.20 ND 136.72 W1 34.58 ± 1.99 ND 14.25 ± 1.74 ND 4.94 ± 0.21 ND 53.76 W2 7.15 ± 0.75 ND 39.57 ± 1.26 ND 12.86 ± 0.76 ND 59.57 W3 27.32 ± 3.46 ND ND ND 1.97 ± 0.22 ND 29.28 ND means not detected. Data were reported as mean ± SD. Limit of detection was 0.06 mg kg−1 for sesamol, 0.83 mg kg−1 for sesamin, and 1.01 mg kg−1 for sesamolin. Table 3 Correlation of multicomponents (phenolics, flavonoids, and lignans) with antioxidant (ORAC, PSC) and antiproliferation (EC50) activities of six sesame seed varieties.   Free phenolics Bound phenolics Total phenolics Free flavonoids Bound flavonoids Total flavonoids Lignans Free PSC values −0.406 NC NC −0.265 NC NC NC Bound PSC values NC 0.604 NC NC 0.294 NC NC Total PSC values NC NC 0.020 NC NC 0.130 0.542 Free ORAC values 0.286 NC NC 0.634 NC NC NC Bound ORAC value NC 0.976∗∗ NC NC 0.417 NC NC Total ORAC values NC NC 0.701∗∗ NC NC 0.412∗ 0.533 Free EC50 −0.750 NC NC 0.465 NC NC NC Bound EC50 NC −0.869∗ NC NC −0.666 NC NC ∗∗ means significant difference at p < 0.01. ∗ means significant difference at p < 0.05. NC means not computed. Table 4 Antiproliferative activities of the free and bound phenolic extracts from six sesame seeds varieties against HepG2 cell. Varieties Antiproliferative capacity to HepG2, EC50 (mg mL−1) Free phenolic extract Bound phenolic extract B1 66.08 ± 0.35b 55.49 ± 9.01b B2 21.04 ± 0.60a 23.57 ± 0.88a B3 63.45 ± 0.85b 30.06 ± 0.82a W1 87.29 ± 3.57c 77.74 ± 1.29c W2 82.12 ± 3.10c 109.53 ± 1.23d W3 124.91 ± 2.79d 67.18 ± 1.92c Data were reported as mean ± SD. Values in the same rows with different letters differ significantly at p < 0.05. ==== Refs 1 Abou-Gharbia H. A. Shehata A. A. Y. Shahidi F. Effect of processing on oxidative stability and lipid classes of sesame oil Food Research International 2000 33 5 331 340 10.1016/S0963-9969(00)00052-1 2-s2.0-0034130296 2 Namiki M. Nutraceutical functions of sesame: a review Critical Reviews in Food Science and Nutrition 2007 47 7 651 673 10.1080/10408390600919114 2-s2.0-35348929361 17943496 3 Elleuch M. Besbes S. Roiseux O. Blecker C. Attia H. Quality characteristics of sesame seeds and by-products Food Chemistry 2007 103 2 641 650 10.1016/j.foodchem.2006.09.008 2-s2.0-33846911232 4 Shahidi F. Liyana-Pathirana C. M. Wall D. S. Antioxidant activity of white and black sesame seeds and their hull fractions Food Chemistry 2006 99 3 478 483 10.1016/j.foodchem.2005.08.009 2-s2.0-33646518383 5 Zhang H. Miao H. Wei L. Li C. Zhao R. Wang C. Genetic analysis and QTL mapping of seed coat color in sesame (Sesamum indicum L.) PLoS ONE 2013 8 5 e63898 10.1371/journal.pone.0063898 2-s2.0-84877948550 6 Zuo Y. Peng C. Liang Y. Sesamin extends the mean lifespan of fruit flies Biogerontology 2013 14 2 107 119 10.1007/s10522-012-9413-4 2-s2.0-84878614389 23291977 7 Tada M. Ono Y. Nakai M. Evaluation of antioxidative effects of sesamin on the in vivo hepatic reducing abilities by a radiofrequency ESR method Analytical Sciences 2013 29 1 89 94 10.2116/analsci.29.89 2-s2.0-84873831580 23303091 8 Srisayam M. Weerapreeyakul N. Barusrux S. Kanokmedhakul K. Antioxidant, antimelanogenic, and skin-protective effect of sesamol Journal of Cosmetic Science 2014 65 2 69 79 2-s2.0-84902545294 24797023 9 Furumoto T. Nishimoto K. Identification of a characteristic antioxidant, anthrasesamone F, in black sesame seeds and its accumulation at different seed developmental stages Bioscience, Biotechnology and Biochemistry 2016 80 2 350 355 10.1080/09168451.2015.1091717 2-s2.0-84959036771 10 Hsu D.-Z. Liu C.-T. Chu P.-Y. Li Y.-H. Periasamy S. Liu M.-Y. Sesame oil attenuates ovalbumin-induced pulmonary edema and bronchial neutrophilic inflammation in mice BioMed Research International 2013 2013 7 905670 10.1155/2013/905670 2-s2.0-84877276892 11 Xu P. Cai F. Liu X. Guo L. Sesamin inhibits lipopolysaccharide-induced proliferation and invasion through the p38-MAPK and NF-κ B signaling pathways in prostate cancer cells Oncology Reports 2015 33 6 3117 3123 10.3892/or.2015.3888 2-s2.0-84928898588 25845399 12 Dar A. A. Arumugam N. Lignans of sesame: purification methods, biological activities and biosynthesis—a review Bioorganic Chemistry 2013 50 1 10 10.1016/j.bioorg.2013.06.009 2-s2.0-84881172802 23933354 13 Kanu P. J. Biochemical analysis of black and white sesame seeds from China American Journal of Biochemistry and Molecular Biology 2011 1 2 145 157 10.3923/ajbmb.2011.145.157 2-s2.0-79952022278 14 Chen Y. Chen G. Fu X. Liu R.-H. Phytochemical profiles and antioxidant activity of different varieties of Adinandra tea (Adinandra Jack) Journal of Agricultural and Food Chemistry 2015 63 1 169 176 10.1021/jf503700v 2-s2.0-84921019434 25494237 15 Singleton V. L. Orthofer R. Lamuela-Raventós R. M. Analysis of total phenols and other oxidation substrates and antioxidants by means of Folin-Ciocalteu reagent Methods in Enzymology 1998 299 152 178 10.1016/s0076-6879(99)99017-1 2-s2.0-0031790242 16 He X. Liu D. Liu R. H. Sodium borohydride/chloranil-based assay for quantifying total flavonoids Journal of Agricultural and Food Chemistry 2008 56 20 9337 9344 10.1021/jf070954 2-s2.0-55549117860 18798633 17 Reshma M. V. Balachandran C. Arumughan C. Extraction, separation and characterisation of sesame oil lignan for nutraceutical applications Food Chemistry 2010 120 4 1041 1046 10.1016/j.foodchem.2009.11.047 2-s2.0-74849121361 18 Huang D. Ou B. Hampsch-Woodill M. Flanagan J. A. Prior R. L. High-throughput assay of oxygen radical absorbance capacity (ORAC) using a multichannel liquid handling system coupled with a microplate fluorescence reader in 96-well format Journal of Agricultural and Food Chemistry 2002 50 16 4437 4444 10.1021/jf0201529 2-s2.0-0037205604 12137457 19 Zhang M. W. Zhang R. F. Zhang F. X. Liu R. H. Phenolic profiles and antioxidant activity of black rice bran of different commercially available varieties Journal of Agricultural and Food Chemistry 2010 58 13 7580 7587 10.1021/jf1007665 2-s2.0-77954557997 20521821 20 Adom K. K. Liu R. H. Rapid peroxyl radical scavenging capacity (PSC) assay for assessing both hydrophilic and lipophilic antioxidants Journal of Agricultural and Food Chemistry 2005 53 17 6572 6580 10.1021/jf048318o 2-s2.0-24344469424 16104768 21 Felice D. L. Sun J. Liu R. H. A modified methylene blue assay for accurate cell counting Journal of Functional Foods 2009 1 1 109 118 10.1016/j.jff.2008.09.014 2-s2.0-67651171248 22 Liu R. H. Jacob J. R. Hotchkiss J. H. Cote P. J. Gerin J. L. Tennant B. C. Woodchuck hepatitis virus surface antigen induces nitric oxide synthesis in hepatocytes: possible role in hepatocarcinogenesis Carcinogenesis 1994 15 12 2875 2877 10.1093/carcin/15.12.2875 2-s2.0-0028587328 8001249 23 Nadeem M. Situ C. Mahmud A. Antioxidant activity of sesame (Sesamum indicum L.) cake extract for the stabilization of olein based butter Journal of the American Oil Chemists' Society 2014 91 6 967 977 10.1007/s11746-014-2432-3 2-s2.0-84903904245 24 Kim J. H. Seo W. D. Lee S. K. Comparative assessment of compositional components, antioxidant effects, and lignan extractions from Korean white and black sesame (Sesamum indicum L.) seeds for different crop years Journal of Functional Foods 2014 7 1 495 505 10.1016/j.jff.2014.01.006 2-s2.0-84898042896 25 Kwan S. K. Park S. H. Choung M. G. Nondestructive determination of lignans and lignan glycosides in sesame seeds by near infrared reflectance spectroscopy Journal of Agricultural and Food Chemistry 2006 54 13 4544 4550 10.1021/jf0605603 2-s2.0-33746501574 16786996 26 Rangkadilok N. Pholphana N. Mahidol C. Variation of sesamin, sesamolin and tocopherols in sesame (Sesamum indicum L.) seeds and oil products in Thailand Food Chemistry 2010 122 3 724 730 10.1016/j.foodchem.2010.03.044 2-s2.0-77951665971 27 Ishiyama K. Nagashima M. Yasumoto T. Fukuda Y. Changes in amounts of lignan compounds and radical scavenging activities in improved high-lignan breed of sesame, ‘Gomazou’, during germination Journal of the Japanese Society for Food Science and Technology 2006 53 1 8 16 10.3136/nskkk.53.8 2-s2.0-33644820002 28 Othman S. B. Katsuno N. Kanamaru Y. Yabe T. Water-soluble extracts from defatted sesame seed flour show antioxidant activity in vitro Food Chemistry 2015 175 306 314 10.1016/j.foodchem.2014.11.155 2-s2.0-84916890258 25577085 29 Dachtler M. Van De Put F. H. M. Stijn F. V. Beindorff C. M. Fritsche J. On-line LC-NMR-MS characterization of sesame oil extracts and assessment of their antioxidant activity European Journal of Lipid Science and Technology 2003 105 9 488 496 10.1002/ejlt.200300835 2-s2.0-0141460615 30 Ghafoorunissa Hemalatha S. Rao M. V. V. Sesame lignans enhance antioxidant activity of vitamin E in lipid peroxidation systems Molecular and Cellular Biochemistry 2004 262 1-2 195 202 10.1023/b:mcbi.0000038235.01389.a9 2-s2.0-4344560306 15532724 31 Moazzami A. A. Haese S. L. Kamal-Eldin A. Lignan contents in sesame seeds and products European Journal of Lipid Science and Technology 2007 109 10 1022 1027 10.1002/ejlt.200700057 2-s2.0-35548960909 32 Ide T. Azechi A. Kitade S. Comparative effects of sesame seeds differing in lignan contents and composition on fatty acid oxidation in rat liver Journal of Oleo Science 2015 64 2 211 222 10.5650/jos.ess14182 25748381 33 Wolfe K. L. Rui H. L. Cellular antioxidant activity (CAA) assay for assessing antioxidants, foods, and dietary supplements Journal of Agricultural and Food Chemistry 2007 55 22 8896 8907 10.1021/jf0715166 2-s2.0-36148987289 17902627 34 Meyers K. J. Watkins C. B. Pritts M. P. Liu R. H. Antioxidant and antiproliferative activities of strawberries Journal of Agricultural and Food Chemistry 2003 51 23 6887 6892 10.1021/jf034506n 2-s2.0-0242267934 14582991 35 Zhang L. Z. Liu R. H. Phenolic and carotenoid profiles and antiproliferative activity of foxtail millet Food Chemistry 2015 174 495 501 10.1016/j.foodchem.2014.09.089 2-s2.0-84912553676 25529711 36 Deng P. Wang C. Chen L. Sesamin induces cell cycle arrest and apoptosis through the inhibition of signal transducer and activator of transcription 3 signalling in human hepatocellular carcinoma cell line HepG2 Biological and Pharmaceutical Bulletin 2013 36 10 1540 1548 10.1248/bpb.b13-00235 2-s2.0-84887056416 24088253 37 Liu Z. Xiang Q. Du L. Song G. Wang Y. Liu X. The interaction of sesamol with DNA and cytotoxicity, apoptosis, and localization in HepG2 cells Food Chemistry 2013 141 1 289 296 10.1016/j.foodchem.2013.02.105 2-s2.0-84877620903 23768360

PMC005xxxxxx/PMC5002302.txt

==== Front Scientifica (Cairo)Scientifica (Cairo)SCIENTIFICAScientifica2090-908XHindawi Publishing Corporation 10.1155/2016/9324279Research ArticlePreliminary Evaluation of Clinical and Angiographic Outcomes with Biodegradable Polymer Coated Sirolimus-Eluting Stent in De Novo Coronary Artery Disease: Results of the MANIPAL-FLEX Study http://orcid.org/0000-0002-1672-2652Shetty Ranjan 1 * Prajapati Jayesh 2 Pai Umesh 1 Shetty Kiran 1 1Kasturba Medical College and Hospital, Manipal, Karnataka 576104, India2Apollo Hospitals International Limited, Gandhinagar, Gujarat 382428, India*Ranjan Shetty: ranjanshettyk@yahoo.comAcademic Editor: Klaus Kettering 2016 14 8 2016 2016 932427924 12 2015 6 5 2016 2 6 2016 Copyright © 2016 Ranjan Shetty et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Objective. The objective of the MANIPAL-FLEX study was to evaluate the feasibility, preliminary safety, and efficacy of the Supraflex sirolimus-eluting stent (SES) implantation, in de novo coronary artery disease, using clinical and quantitative coronary angiography (QCA) follow-ups. Methods. This was a prospective, nonrandomized, multicenter, single-arm study that enrolled 189 patients with de novo coronary artery disease who were treated with the Supraflex SES. Of 189 patients enrolled, the first 61 consecutive patients who consented to a 9-month follow-up evaluation by QCA, irrespective of presence of symptoms, were to be followed up with angiography at 9 months. The primary endpoint of the study was target lesion failure (TLF), including cardiac death, myocardial infarction, and target lesion revascularization during 12-month follow-up after the index procedure. Results. The mean age of the study population was 58 ± 11 years, with 51.3% (97/189) of hypertensive patients. Total of 66 lesions, analyzed by offline QCA, showed good scaffolding of the target vessel with in-stent late lumen loss at 9 months of 0.18 ± 0.23 mm. The observed TLF at 30-day, 6-month, and 12-month follow-up were 2 (1.1%), 6 (3.2%), and 10 (5.3%), respectively. Conclusion. This study provides preliminary evidence for the feasibility, safety, and efficacy of the Supraflex sirolimus-eluting stent. ==== Body 1. Introduction In spite of the reduction of restenosis and the need for repeat revascularization with drug-eluting stents (DESs), the persistence of adverse events with first-generation durable polymer-based DES enlightens the opportunity for further improvement in DES technology [1–3]. The potential link between the durable polymer and late adverse events has prompted the development of biodegradable sources of drug delivery to a stented vessel [4–6]. The use of a biodegradable polymer has the potential to reduce the sustained inflammatory response of the arterial wall, facilitating reendothelialisation and minimising the risk of thrombus formation and late restenosis [7, 8]. Furthermore, first-generation DESs, with thick-strut design, are associated with more thrombotic events than thin-strut comparators in ex vivo and experimental models [9]. To resolve these issues, newer DES technologies, such as the use of stents with thinner struts and the use of biodegradable coatings to improve biocompatibility, have been developed. These properties might confer an additional benefit in terms of reendothelialisation and restenosis [10, 11]. The design and engineering of stent platforms is constantly evolving with major advancements made over the years [12–14]. Furthermore, stent design including strut thickness, metal composition, and radial strength can each influence short- and long-term clinical outcomes [15, 16]. The Supraflex [Sahajanand Medical Technologies Pvt. Ltd. (SMT), India] was developed using the same proven biodegradable polymer-based, sirolimus-eluting technology as on the Supralimus-Core (SMT, India) sirolimus-eluting stent (SES) but was applied to the newer Flexinnium (SMT, India) cobalt-chromium bare-metal stent. This new stent design has thinner struts (60 μm) with highly flexible “S-link.” It has been considered that such stent design could facilitate deliverability, especially in complex lesions, without compromising safety and efficacy. The MANIPAL-FLEX study was designed to evaluate the feasibility, preliminary safety, and efficacy of the Supraflex SES implantation, in de novo coronary artery disease (CAD), using clinical and quantitative coronary angiography (QCA) follow-ups. 2. Materials and Methods 2.1. Study Design and Patient Population This was a prospective, nonrandomized, multicenter, single-arm study which enrolled 189 patients with de novo CAD, who were treated with the Supraflex SES, between November 2012 and March 2014. Of 189 patients enrolled, the first 61 consecutive patients who consented to a 9-month follow-up evaluation by QCA, irrespective of presence of symptoms, were to be followed up with angiography at 9 months. Clinical follow-up was intended at 30 days, 6 months, and 12 months for all enrolled patients (Figure 1). Clinical inclusion criteria were age ≥18 years, symptoms of angina or ischemia, and acceptable candidacy for coronary artery bypass surgery. Angiographic criteria included de novo coronary lesions ≤44 mm in length, reference diameter of 2.0 to 4.5 mm, stenosis ≥50%, and agreement to undergo all protocol-required follow-up examinations including angiographic follow-up at 9 months. Exclusion criteria were cardiogenic shock, left ventricular ejection fraction <30%, ostial lesion location, major surgery within 6 weeks before planned stenting procedure, stenosis >50% of the left main coronary artery, angiographic evidence of thrombus or poor distal flow at the lesion site, lesion at a junction of a side branch with a diameter >2 mm, and staged procedure planned within 3 months of the index procedure. The study was reviewed and approved by institutional ethics committees [Kasturba Hospital, Manipal (IEC 471/2012), and Apollo Hospitals International Ltd. (ECR/30/Inst/GJ/2013)]. The study was conducted in accordance with the principles of good clinical practice and the Declaration of Helsinki. Patients' participation in the study was voluntary and eligible patients signed written informed consent prior to the interventional procedure. 2.2. Study Device The Supraflex SES is designed, using the Flexinnium bare-metal stent as a platform, with highly flexible “S-link” (Figure 2). The Supraflex SES has L605 cobalt-chromium (Co-Cr) alloy, having strut thickness of 60 μm with biodegradable polymers and drug load of 1.4 μg/mm2. The stent is mounted on a rapid-exchange catheter with a high-pressure, semicompliant balloon. The coating layer comprises drug sirolimus blended together with biodegradable polymeric matrix. This matrix includes different biodegradable polymers, Poly(L-lactide), 50/50 Poly(DL-lactide-co-glycolide), and Polyvinylpyrrolidone, to control the drug elution from stent coating. The average coating thickness of the Supraflex SES is between 4 and 5 μm. The coating of the polymer is conformal around the stent struts and not limited to abluminal surface of the stent. Figure 3 shows scanning electron microscopy (SEM) images of a normal, crimped, and expanded Supraflex stent. These images demonstrate that the surface of the stent was coated uniformly with a thin film that conformed faithfully to the stent surface, the contours of the stent struts, and the balloon assembly. About 70% of drug is released within 7 days and remaining drug is released over a period of 48 days (data on file at SMT, India) (Figure 4). The polymers retain their properties for a limited period and then gradually degrade into biologically acceptable molecules that are metabolized and removed from the body via normal metabolic pathways. This process takes about 9–12 months. The Supraflex SES is available in diameters of 2.0, 2.25, 2.5, 2.75, 3.0, 3.50, 4.0, and 4.5 mm and in length sizes of 8, 12, 16, 20, 24, 28, 32, 36, 40, 44, and 48 mm. 2.3. Coronary Intervention Procedures and Adjuvant Medications Coronary interventional procedures and adjuvant medications were performed according to standard guidelines [17]. All patients received dual antiplatelet therapy (DAPT) including a loading dose of aspirin (300 mg) and clopidogrel (600 mg) or prasugrel (60 mg) or ticagrelor (90 mg). The procedural anticoagulation was achieved either with heparin or with bivalirudin. However, the intraprocedural administration of glycoprotein IIb/IIIa inhibitor was at the investigator's discretion. All patients were advised to maintain DAPT (aspirin; 75–300 mg daily indefinitely and clopidogrel; 75 mg daily or prasugrel; 10 mg daily or ticagrelor; 90 mg twice daily for at least 12 months) after the procedure. 2.4. Patient Follow-Up Clinical data were collected before and after the index procedure and recorded in a dedicated database for all patients. Total 61 patients underwent angiographic evaluation at 9 months. Clinical/telephonic follow-up was scheduled for all enrolled patients at 30 days, 6 months, and 12 months, and collected details like assessment of angina status, monitoring of cardiovascular and antithrombotic drug use, interim hospitalisations, occurrence of major adverse cardiac events, and any invasive and noninvasive diagnostic test or interventional treatment that had occurred since the previous contact. An independent clinical events committee (CEC) reviewed and adjudicated all major adverse cardiac events. 2.5. Endpoints of the Study and Definitions The primary endpoint of the study was target lesion failure (TLF), including cardiac death, myocardial infarction (MI), and target lesion revascularization (TLR) during 12-month follow-up after the index procedure. Secondary endpoints included acute success (lesion success, device success, and procedural success) and in-stent late lumen loss at 9 months by QCA. Lesion success was defined as a final diameter stenosis <30% with any device with TIMI 3 flow; device success was defined as stenosis of <30% diameter with TIMI 3 flow in the absence of device malfunctions; and procedure success was defined as a final diameter stenosis <30% with TIMI 3 flow, without the occurrence of death, MI, or repeat revascularization of the target lesion during the hospital stay. Any death was considered cardiac, unless clear noncardiac causes could be determined. MI was defined by elevation of cardiac troponin (cTn) values [>5 × 99th percentile upper reference limit (URL)] in patients with normal baseline values (≤99th percentile URL) or a rise of cTn values >20% if the baseline values are elevated and are stable or falling. TLR was defined as repeat percutaneous intervention of the stented lesion including 5 mm proximal and distal from the edge of the stent or bypass surgery of the target vessel that was performed for a clinical indication and was due to restenosis or closure of the target lesion. Target vessel revascularization (TVR) was defined as repeat revascularization by either percutaneous coronary intervention (PCI) or CABG of the target vessel. We also evaluated the incidence of stent thrombosis (ST) as a safety endpoint during 12-month follow-up period, as defined by the Academic Research Consortium (ARC) [18]. 2.6. Angiographic Assessments Angiographic assessments were performed at baseline, after procedure, and at 9-month follow-up, in two orthogonal views after the intracoronary administration of nitroglycerin. A quantitative offline coronary angiographic analysis was performed [Coronary Angiography Analysis System (CAAS, version 5.9.2); Pie Medical Imaging, Maastricht, Netherlands] with automatic edge detection by senior interventional cardiologist who had experience in reading coronary angiograms with performing QCA and was blinded to the clinical data. For every patient, in-stent (bordered by the stent margins) and in-segment (in-stent region plus 5 mm margins proximal and distal to the stent) measurements were performed by QCA. The quantitative angiographic parameters including reference vessel diameter (RVD), percentage diameter stenosis (DS), minimal lumen diameter (MLD), late luminal loss, binary restenosis, and acute gain were evaluated. Percent diameter stenosis was defined as 1 − [MLD/RVD] × 100, late luminal loss was defined as the difference between the MLD at the completion of the stenting procedure and the MLD measured at angiographic follow-up, binary restenosis was defined as a diameter stenosis of ≥50% at angiographic follow-up, and acute gain was defined as the MLD immediately after the procedure minus the MLD before the procedure. 2.7. Statistical Analysis The MANIPAL-FLEX study was designed to provide adequate preliminary data about the safety and efficacy of the Supraflex SES and was therefore considered explorative. Resulting values will be the basis for sample size calculations for future studies. Based on similar studies with other drug-eluting stent systems, a sample size of 189 patients (61 patients for angiographic assessments) was expected to be sufficient for preliminary evaluation of the safety and efficacy of the device. Data are presented using descriptive statistical methods. Continuous variables were presented as mean ± standard deviation, whereas categorical variables were expressed as percentages. All data were processed using the Statistical Package for Social Sciences, version 15 (SPSS, Chicago, IL, USA). 3. Results 3.1. Baseline, Lesion, and Procedural Characteristics The baseline clinical characteristics are outlined in Table 1. Among 189 enrolled patients, the mean age was 58 ± 11 years and 30.7% (58/189) had diabetes mellitus. Lesion and procedural characteristics are listed in Table 2. For patients undergoing examination by quantitative coronary angiography (n = 61) at 9 months, the mean lesion length was 9.70 ± 6.94 mm, and the average RVD was 2.42 ± 0.49 mm (Table 3). Among overall population most lesions (48.4%; 105/217) were located in the left anterior descending artery and 181 (83.4%) of 217 lesions were type B2/C, according to the American College of Cardiology/American Heart Association classification scheme. Lesion, device, and procedural success were achieved in 100% of patients. 3.2. Quantitative Coronary Angiography Outcomes Pre- and postprocedural and 9-month QCA measurements were performed on the first 61 consecutive patients (66 lesions) who consented to a 9-month follow-up evaluation by QCA, irrespective of presence of symptoms, and are presented in Table 3. The mean in-stent late lumen loss at 9 months was 0.18 ± 0.23 mm, while the mean in-segment late lumen loss was 0.11 ± 0.33 mm. In-stent diameter stenosis after 9 months was 16.14 ± 7.52 percentage with binary restenosis rate of 1.5% (1/66). Figure 5 demonstrates the cumulative frequency distribution curves of in-stent MLD at preprocedure, postprocedure, and 9-month follow-up. 3.3. Clinical Outcomes The in-hospital incidence of cardiac death, MI, TLR, TVR-nontarget lesion, and/or stent thrombosis was 0%. Clinical follow-up up to 12 months was completed in all 189 patients (100%) and is summarized in Table 4. No TLR was observed during the first 30 days, though from 30 days to 6 months, 2 patients (1.1%) had TLR and from 6 months to 12 months, another two patients had TLR, resulting in a 12-month cumulative TLR of 2.1% (4/189). The observed TLF at 30-day, 6-month, and 12-month follow-up were 2 (1.1%), 6 (3.2%), and 10 (5.3%), respectively. According to ARC, there was only one case (0.5%) of late stent thrombosis and was recorded 11 months following the index procedure. One patient died from a new MI during 30-day follow-up and another two sudden deaths were noted at 6-month and 12-month clinical follow-up with cause of death adjudicated as cardiac death but without MI or stent thrombosis. A total of 98.9% and 73.7% of patients were on DAPT at 6 months and 12 months, respectively. 4. Discussion Due to concern related to impaired healing and inflammation associated with first-generation durable polymer DES, scientific interest has been created in developing newer devices, with improved polymer and metallic platform technologies [1, 19, 20]. Thus, the concept of thin-strut, biodegradable polymer coated DES emerged in the recent years. Biodegradation ensures that both the drug and coating are absorbed from the stent surface after their function is accomplished, leaving only a metal stent covered by neointima and endothelium without further continuous irritation of the arterial wall [21]. Moreover, the randomized LEADERS studies showed that biodegradable polymer stents reduced the risk of cardiac events associated with very late stent thrombosis, which might improve long-term clinical outcomes compared with durable polymer sirolimus-eluting stents [22, 23]. Furthermore, in clinical circumstances, thin-strut stents have been shown to produce less neointimal hyperplasia than thick-strut bare-metal stents [10, 11]. Also, several clinical studies using thin-strut bare-metal cobalt-chromium stents have shown the advantages of enhanced visibility, deliverability and radial strength, and reducing restenosis when compared with bare-metal stainless steel stents [24–27]. In this context, the Supraflex SES was developed using the same proven biodegradable polymer-based, sirolimus-eluting technology as on the Supralimus-Core SES but was applied to the thinner struts (60 μm) cobalt-chromium stent platform (Flexinnium bare-metal stent). The clinical safety and effectiveness of the Supralimus-Core was already established in various clinical studies [28–34]. In this MANIPAL-FLEX study, patient population had higher rates of hypertension (51.3%), hypercholesterolemia (32.3%), type B2/C (83.4%) lesions, and total occluded (31.8%) lesions compared to real-world Supralimus-Cores' S-CORE registry and CORE registry [32, 34]. The combination of these factors makes the patient population for this study unusually complex. Regardless of this individuality, the results of this experience confirm the excellent acute performance of the Supraflex SES. The ease of deliverability, flexibility, and crossing of the Supraflex SES across the lesion was confirmed with 100% lesion, device, and procedural success, which preliminarily indicates feasibility for this device. Additionally, in-stent late lumen loss at 9-month poststent implantation was one of the secondary endpoints of the MANIPAL-FLEX study. The main aim to select this endpoint was that late lumen loss as an angiographic factor is confirmed to be a dependable predictor of the long-term clinical efficacy of DES and also strong predictor of binary restenosis and TLR [35]. The 9-month in-stent late lumen loss of 0.18 ± 0.23 mm observed in the current study situates Supraflex SES among the most effective DES systems, which resembles to the late lumen loss reported for other DESs using limus drugs such as Promus Element (0.15 mm), Xience (0.12 mm), Biomatrix (0.26 mm), Nobori (0.10 mm), or Synergy (0.10 mm) [36–39]. The observed late lumen loss, in this study, with the Supraflex SES translated into an unpretentious level of binary in-stent restenosis (1.5%) and binary in-segment restenosis (1.5%) which is also comparable to sirolimus-eluting Ultimaster stent [40]. In this study, the primary outcome, 5.3% TLF rate at 12 months, compares well to the 5.1% TLF rate reported for the Orsiro DES in the BIOFLOW-III registry [41]. Furthermore, it is noted that, at 12 months, the TLR rate in the MANIPAL-FLEX study was 2.1% which is lower to the rate of TLR with the Ultimaster DES (3.8%) at one year [40]. Also, analysis of 50 patients from the MANIPAL-FLEX study, presented at the Euro-PCR meeting 2015, showed good clinical outcomes [42]. The promising clinical outcomes in this study might be attributed to the biodegradable polymer and the unique design of the Supraflex stent. In summary, this study provided preliminary evidence for the feasibility, safety, and efficacy of the Supraflex sirolimus-eluting stent in the treatment of de novo CAD. 4.1. Limitations of the Study This study is limited by the fact that it is a single-arm, nonrandomized study without comparator or adequate statistical power for conclusive definition of angiographic outcomes. In addition, limited clinical outcomes up to 12 months may not be enough to capture all the adverse events, especially very late stent thrombosis. Accordingly, these results must be measured in a comparative study with larger population, over longer follow-up duration. 5. Conclusion This MANIPAL-FLEX study provided preliminary evidence for the feasibility, safety, and efficacy of the Supraflex SES. In this study, the Supraflex, thin-strut biodegradable polymer coated sirolimus-eluting stent demonstrated high-level efficacy, by the relatively low late lumen loss, at 9-month angiographic follow-up. Also, implantation of the Supraflex SES was proven to be safe, with high acute lesion, procedure, and device success rates. A randomized study with larger study population and over long-term follow-up is needed for further information and validation. Competing Interests The authors declare that there is no conflict of interests regarding the publication of this paper. Figure 1 Flow chart of the study. Figure 2 Supraflex sirolimus-eluting stent: a schematic view of the stent structure (circle shows flexible “S-link”). Figure 3 Scanning electron microscopy (SEM) images of a (a) normal, (b) crimped, and (c) expanded Supraflex sirolimus-eluting stent. Figure 4 Drug release kinetics of the Supraflex sirolimus-eluting stent. Figure 5 Cumulative frequency distribution curves of in-stent minimal lumen diameter by QCA. Table 1 Baseline clinical characteristics for the entire population (n = 189) and for patients undergoing examination by quantitative coronary angiography (n = 61) at 9 months. Variables n = 189 n = 61 Age, (mean ± SD, years) 58 ± 11 56 ± 10 Male, n (%) 136 (72.0%) 49 (80.3%) Cardiovascular risk      Diabetes mellitus, n (%) 58 (30.7%) 13 (21.3%)  Hypertensive, n (%) 97 (51.3%) 26 (42.6%)  Hypercholesterolemia, n (%) 61 (32.3%) 18 (29.5%)  Current smoker, n (%) 45 (23.8%) 11 (18.0%)  Family history of CAD, n (%) 12 (6.3%) 4 (6.6%)  Previous MI, n (%) 13 (6.9%) 4 (6.6%)  Previous PCI, n (%) 17 (9.0%) 5 (8.2%) Clinical presentation      Stable angina, n (%) 8 (4.2%) 5 (8.2%)  Unstable angina, n (%) 18 (9.5%) 16 (26.2%)  ST-elevation myocardial infarction, n (%) 58 (30.7%) 18 (29.5%)  Non-ST-elevation myocardial infarction, n (%) 105 (55.6%) 22 (36.1%) CAD = coronary artery disease; MI = myocardial infarction; PCI = percutaneous coronary intervention. Table 2 Lesion and procedural characteristics for the entire population (n = 189) and for patients undergoing examination by quantitative coronary angiography (n = 61) at 9 months. Variables n = 189 n = 61 Number of lesions (n) 217 66 Treated coronary artery      Left anterior descending artery, n (%) 105 (48.4%) 33 (50.0%)  Right coronary artery, n (%) 73 (33.6%) 22 (33.3%)  Left circumflex artery, n (%) 38 (17.5%) 11 (16.7%)  Left main, n (%) 1 (0.5%) 0 (0) Lesion classification (ACC/AHA score)      Type A, n (%) 6 (2.8%) 4 (6.1%)  Type B1, n (%) 30 (13.8%) 17 (25.8%)  Type B2, n (%) 108 (49.8%) 30 (45.5%)  Type C, n (%) 73 (33.6%) 15 (22.7%) Total occlusion, n (%) 69 (31.8%) 14 (21.2%) Total number of stents, (n) 230 76 Number of stents per patient (mean ± SD, mm) 1.22 ± 0.47 1.34 ± 0.60 Number of stents per lesion (mean ± SD, mm) 1.05 ± 0.29 1.12 ± 0.45 Average stent length (mean ± SD, mm) 24.88 ± 7.72 15.94 ± 7.95 Average stent diameter (mean ± SD, mm) 3.05 ± 0.34 3.15 ± 0.45 Predilation, n (%) 187 (98.9%) 59 (96.7%) Postdilation, n (%) 127 (67.2%) 34 (55.7%) Lesion success (%) 100% 100% Device success (%) 100% 100% Procedure success (%) 100% 100% ACC/AHA = American College of Cardiology/American Heart Association. Table 3 Results of quantitative coronary angiography analysis at preprocedure, postprocedure, and 9-month follow-up (n = 61 patients).   In-segment In-stent Preprocedure      Reference vessel diameter (mm) 2.42 ± 0.49 —  Diameter stenosis (%) 69.54 ± 18.81 —  Minimal lumen diameter (mm) 0.75 ± 0.53 —  Lesion length (mm) 9.70 ± 6.94 — Postprocedure      Reference vessel diameter (mm) 2.56 ± 0.44 2.65 ± 0.43  Diameter stenosis (%) 21.90 ± 8.49 14.07 ± 7.40  Minimal lumen diameter (mm) 2.01 ± 0.46 2.27 ± 0.39  Acute gain (mm) 1.26 ± 0.52 1.52 ± 0.53 9-month follow-up      Reference vessel diameter (mm) 2.42 ± 0.43 2.49 ± 0.41  Diameter stenosis (%) 22.00 ± 8.65 16.14 ± 7.52  Minimal lumen diameter (mm) 1.90 ± 0.46 2.09 ± 0.44  Late lumen loss (mm) 0.11 ± 0.33 0.18 ± 0.23  Binary restenosis (%) 1.5 1.5 Table 4 Cumulative clinical outcomes up to 12 months (n = 189 patients). Variables 30 days 6 months 12 months Cardiac death, n (%) 1 (0.5%) 3 (1.6%) 3 (1.6%) Noncardiac death, n (%) 0 (0%) 2 (1.1%) 6 (3.2%) Myocardial infarction, n (%) 1 (0.5%) 1 (0.5%) 3 (1.6%) Target lesion revascularization, n (%) 0 (0%) 2 (1.1%) 4 (2.1%) Target vessel revascularization-non-TL, n (%) 0 (0%) 1 (0.5%) 3 (1.6%) Stent thrombosis, n (%) 0 (0%) 0 (0%) 1 (0.5%) Definite stent thrombosis, n (%) 0 (0%) 0 (0%) 1 (0.5%) Target lesion failure, n (%) 2 (1.1%) 6 (3.2%) 10 (5.3%) TL = target lesion. ==== Refs 1 Daemen J. Wenaweser P. Tsuchida K. Early and late coronary stent thrombosis of sirolimus-eluting and paclitaxel-eluting stents in routine clinical practice: data from a large two-institutional cohort study The Lancet 2007 369 9562 667 678 10.1016/s0140-6736(07)60314-6 2-s2.0-33847167065 2 Kimura T. Morimoto T. Nakagawa Y. Very late stent thrombosis and late target lesion revascularization after sirolimus-eluting stent implantation: five-year outcome of the j-cypher registry Circulation 2012 125 4 584 591 10.1161/circulationaha.111.046599 22203694 3 Byrne R. A. Kastrati A. Tiroch K. 2-year clinical and angiographic outcomes from a randomized trial of polymer-free dual drug-eluting stents versus polymer-based cypher and endeavor, drug-eluting stents Journal of the American College of Cardiology 2010 55 23 2536 2543 10.1016/j.jacc.2010.03.020 2-s2.0-77952744608 20417052 4 Nebeker J. R. Virmani R. Bennett C. L. Hypersensitivity cases associated with drug-eluting coronary stents: a review of available cases from the Research on Adverse Drug Events and Reports (RADAR) project Journal of the American College of Cardiology 2006 47 1 175 181 10.1016/j.jacc.2005.07.071 2-s2.0-29344432871 16386683 5 Park S.-J. Kang S.-J. Virmani R. Nakano M. Ueda Y. In-stent neoatherosclerosis: a final common pathway of late stent failure Journal of the American College of Cardiology 2012 59 23 2051 2057 10.1016/j.jacc.2011.10.909 2-s2.0-84861596931 22651862 6 Nordmann A. J. Briel M. Bucher H. C. Mortality in randomized controlled trials comparing drug-eluting vs. bare metal stents in coronary artery disease: a meta-analysis European Heart Journal 2006 27 23 2784 2814 10.1093/eurheartj/ehl282 2-s2.0-33751247472 17020889 7 Joner M. Finn A. V. Farb A. Pathology of drug-eluting stents in humans: delayed healing and late thrombotic risk Journal of the American College of Cardiology 2006 48 1 193 202 10.1016/j.jacc.2006.03.042 2-s2.0-33745233024 16814667 8 Hofma S. H. Van Der Giessen W. J. Van Dalen B. M. Indication of long-term endothelial dysfunction after sirolimus-eluting stent implantation European Heart Journal 2006 27 2 166 170 10.1093/eurheartj/ehi571 2-s2.0-30344439818 16249221 9 Kolandaivelu K. Swaminathan R. Gibson W. J. Stent thrombogenicity early in high-risk interventional settings is driven by stent design and deployment and protected by polymer-drug coatings Circulation 2011 123 13 1400 1409 10.1161/CIRCULATIONAHA.110.003210 2-s2.0-79954633893 21422389 10 Briguori C. Sarais C. Pagnotta P. In-stent restenosis in small coronary arteries: impact of strut thickness Journal of the American College of Cardiology 2002 40 3 403 409 10.1016/s0735-1097(02)01989-7 2-s2.0-0037036821 12142103 11 Jabara R. Geva S. Ribeiro H. B. A third generation ultra-thin strut cobalt chromium stent: histopathological evaluation in porcine coronary arteries EuroIntervention 2009 5 5 619 626 10.4244/eijv5i5a99 2-s2.0-77955499599 20142184 12 Kereiakes D. J. Cox D. A. Hermiller J. B. Usefulness of a cobalt chromium coronary stent alloy American Journal of Cardiology 2003 92 4 463 466 10.1016/S0002-9149(03)00669-6 2-s2.0-0041562602 12914881 13 Martin D. M. Boyle F. J. Drug-eluting stents for coronary artery disease: a review Medical Engineering and Physics 2011 33 2 148 163 10.1016/j.medengphy.2010.10.009 2-s2.0-79951773825 21075668 14 Sprague E. A. Pomeranz M. L. Odess I. Furnish S. M. Granada J. F. Surface material, surface treatment and nanotechnology in cardiovascular stent development EuroIntervention 2008 4 C60 C62 2-s2.0-63149193857 19202698 15 Rogers C. D. Optimal stent design for drug delivery Reviews in Cardiovascular Medicine 2004 5, supplement S9 S15 15184829 16 Kastrati A. Mehilli J. Dirschinger J. Intracoronary stenting and angiographic results: strut thickness effect on restenosis outcome (isar-stereo) trial Circulation 2001 103 23 2816 2821 10.1161/01.cir.103.23.2816 11401938 17 King S. B. III Smith S. C. Jr. Hirshfeld J. W. Jr. 2007 Focused update of the ACC/AHA/SCAI 2005 guideline update for percutaneous coronary intervention: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines: 2007 writing group to review new evidence and update the ACC/AHA/SCAI 2005 guideline update for percutaneous coronary intervention, writing on behalf of the 2005 Writing Committee Circulation 2008 117 2 261 295 18079354 18 Cutlip D. E. Windecker S. Mehran R. Clinical end points in coronary stent trials: a case for standardized definitions Circulation 2007 115 17 2344 2351 10.1161/circulationaha.106.685313 2-s2.0-34247558672 17470709 19 Virmani R. Guagliumi G. Farb A. Localized hypersensitivity and late coronary thrombosis secondary to a sirolimus-eluting stent: should we be cautious? Circulation 2004 109 6 701 705 10.1161/01.cir.0000116202.41966.d4 2-s2.0-10744229988 14744976 20 Kahn J. TCT Daily: ESTROFA: late thrombosis not increased in DES patients: Spanish registry of drug-eluting stents shows STEMI, LAD artery lesions, significant risk factors Journal of interventional cardiology 2007 20 1 p. 32 10.1111/j.1540-8183.2007.00231.x 2-s2.0-33947494324 21 Buszman P. Trznadel S. Milewski K. Novel paclitaxel-eluting, biodegradable polymer coated stent in the treatment of de novo coronary lesions: a prospective multicenter registry Catheterization and Cardiovascular Interventions 2008 71 1 51 57 10.1002/ccd.21392 2-s2.0-38549103648 18098182 22 Stefanini G. G. Kalesan B. Serruys P. W. Long-term clinical outcomes of biodegradable polymer biolimus-eluting stents versus durable polymer sirolimus-eluting stents in patients with coronary artery disease (LEADERS): 4 year follow-up of a randomised non-inferiority trial The Lancet 2011 378 9807 1940 1948 10.1016/s0140-6736(11)61672-3 2-s2.0-82755195016 23 Wykrzykowska J. Serruys P. Buszman P. The three year follow-up of the randomised ‘all-comers’ trial of a biodegradable polymer biolimus-eluting stent versus permanent polymer sirolimus-eluting stent (Leaders) EuroIntervention 2011 7 7 789 795 10.4244/eijv7i7a125 2-s2.0-82655171644 22027687 24 Beijk M. A. M. Piek J. J. XIENCE V everolimus-eluting coronary stent system: a novel second generation drug-eluting stent Expert Review of Medical Devices 2007 4 1 11 21 10.1586/17434440.4.1.11 2-s2.0-33846550022 17187467 25 Moliterno D. J. Restenosis following thin-strut bare-metal stents versus thick-strut drug-eluting stents Catheterization and Cardiovascular Interventions 2007 69 6 799 800 10.1002/ccd.21057 2-s2.0-34249870397 17377984 26 Sketch M. H. Jr. Ball M. Rutherford B. Popma J. J. Russell C. Kereiakes D. J. Evaluation of the medtronic (Driver) cobalt-chromium alloy coronary stent system American Journal of Cardiology 2005 95 1 8 12 10.1016/j.amjcard.2004.08.055 2-s2.0-11144248125 15619386 27 Legrand V. Kelbaek H. Hauptmann K. E. Clinical and angiographic analysis with a Cobalt Alloy Coronary Stent (Driver) in stable and unstable angina pectoris American Journal of Cardiology 2006 97 3 349 352 10.1016/j.amjcard.2005.08.051 2-s2.0-31144456344 16442394 28 Seth A. Chandra P. Chouhan N. S. Thakkar A. S. A first-in-man study of sirolimus-eluting, biodegradable polymer coated cobalt chromium stent in real life patients Indian Heart Journal 2012 64 6 547 552 10.1016/j.ihj.2012.07.011 2-s2.0-84879746065 23253405 29 Thakkar A. S. Abhyankar A. D. Dani S. I. Systemic exposure of sirolimus after coronary stent implantation in patients with de novo coronary lesions: Supralimus-Core® pharmacokinetic study Indian Heart Journal 2012 64 3 273 279 10.1016/s0019-4832(12)60086-8 2-s2.0-84870936091 22664810 30 Abhyankar A. D. Thakkar A. S. In vivo assessment of stent recoil of biodegradable polymer-coated cobalt-chromium sirolimus-eluting coronary stent system Indian Heart Journal 2012 64 6 541 546 10.1016/j.ihj.2012.07.005 2-s2.0-84879756942 23253404 31 Abhyankar A. D. Prajapati J. Reddy S. Early vascular healing with biodegradable polymer coated sirolimus-eluting coronary stent implantation: assessed by optical coherence tomography results at 4-month follow-up Minerva Cardioangiologica 2013 61 3 313 322 2-s2.0-84879741288 23681134 32 Chandwani P. Abhyankar A. D. Prajapati J. S. Porwal S. C. Thakkar A. S. Clinical performance of the cobalt-chromium biodegradable polymer coated sirolimus-eluting stent in an unselected real-world population International Journal of Clinical Medicine 2014 5 206 215 10.4236/ijcm.2014.55034 33 Chandwani P. Prajapati J. Porwal S. Khambhati B. Thakkar A. Evaluation of clinical outcomes in patients undergoing dual vessel percutaneous coronary intervention using sirolimus-eluting coronary stent system in India Journal of Clinical and Diagnostic Research 2015 9 2 OC05 OC09 10.7860/jcdr/2015/10766.5525 2-s2.0-84922225652 34 Banker D. Ahuja A. Sanadhya H. Clinical performance of biodegradable polymer-coated sirolimus-eluting stents in unselected real-world population with coronary artery disease: results from the multicenter CORE Registry Minerva Cardioangiologica 2016 64 1 9 14 25907696 35 Pocock S. J. Lansky A. J. Mehran R. Angiographic surrogate end points in drug-eluting stent trials: a systematic evaluation based on individual patient data from 11 randomized, controlled trials Journal of the American College of Cardiology 2008 51 1 23 32 10.1016/j.jacc.2007.07.084 2-s2.0-37449021807 18174032 36 Chevalier B. Silber S. Park S. J. Randomized comparison of the nobori biolimus a9-eluting coronary stent with the taxus liberte paclitaxel-eluting coronary stent in patients with stenosis in native coronary arteries: the nobori 1 trial—phase 2 Circulation: Cardiovascular Interventions 2009 2 3 188 195 20031715 37 Serruys P. W. Ong A. T. Piek J. J. A randomized comparison of a durable polymer everolimus-eluting stent with a bare metal coronary stent: the spirit first trial EuroIntervention 2005 1 1 58 65 19758878 38 Costa R. A. Lansky A. J. Abizaid A. Angiographic results of the first human experience with the biolimus A9 drug-eluting stent for de novo coronary lesions American Journal of Cardiology 2006 98 4 443 446 10.1016/j.amjcard.2006.02.051 2-s2.0-33746859774 16893694 39 Meredith I. T. Verheye S. Weissman N. J. Six-month IVUS and two-year clinical outcomes in the EVOLVE FHU trial: a randomised evaluation of a novel bioabsorbable polymer-coated, everolimus-eluting stent EuroIntervention 2013 9 3 308 315 10.4244/eijv9i3a52 2-s2.0-84880870416 23872647 40 Barbato E. Salinger-Martinovic S. Sagic D. A first-in-man clinical evaluation of Ultimaster, a new drugeluting coronary stent system: Century study EuroIntervention 2015 11 5 541 548 10.4244/eijy14m08-06 2-s2.0-84943786480 25136883 41 Waltenberger J. Brachmann J. van der Heyden J. Real-world experience with a novel biodegradable polymer sirolimus-eluting stent: twelve-month results of the BIOFLOW-III registry EuroIntervention 2016 11 10 1106 1110 10.4244/EIJY15M03_08 25782184 42 Shetty R. Pai U. Shetty K. Khambhati B. Thakkar A. Evaluation of clinical and angiographic outcomes with bioabsorbable polymer coated sirolimus-eluting stent in de novo coronary artery disease: results of the MANIPAL-FLEX first-in-man study EuroIntervention 2015

PMC005xxxxxx/PMC5002303.txt

==== Front Biomed Res IntBiomed Res IntBMRIBioMed Research International2314-61332314-6141Hindawi Publishing Corporation 10.1155/2016/1629236Research ArticleRole of miRNAs in Epicardial Adipose Tissue in CAD Patients with T2DM http://orcid.org/0000-0001-5340-3944Liu Yang 1 http://orcid.org/0000-0001-8900-7871Fu Wenbo 2 Lu Mu 1 Huai Shitao 1 Song Yaqin 1 http://orcid.org/0000-0001-8738-0488Wei Yutao 3 * 1Medicine Department, Shihezi University, Shihezi, Xinjiang 832000, China2Department of Cardiology, People's Hospital of Xinjiang Uygur Autonomous Region, Urumqi, Xinjiang 830000, China3Department of Thoracic and Cardiovascular Surgery, The First Affiliated Hospital of Shihezi University School of Medicine, Shihezi, Xinjiang 832008, China*Yutao Wei: wytfwb@126.comAcademic Editor: Vinicio A. de Jesus Perez 2016 14 8 2016 2016 162923617 3 2016 26 6 2016 Copyright © 2016 Yang Liu et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background. Epicardial adipose tissue (EAT) is identified as an atypical fat depot surrounding the heart with a putative role in the involvement of metabolic disorders, including obesity, type-2 diabetes mellitus, and atherosclerosis. We profiled miRNAs in EAT of metabolic patients with coronary artery disease (CAD) and type-2 diabetes mellitus (T2DM) versus metabolically healthy patients by microarray. Compared to metabolically healthy patients, we identified forty-two miRNAs that are differentially expressed in patients with CAD and T2DM from Xinjiang, China. Eleven miRNAs were selected as potential novel miRNAs according to P value and fold change. Then the potential novel miRNAs targeted genes were predicted via TargetScan, PicTar, and miRTarbase, and the function of the target genes was predicted via Gene Ontology (GO) analysis while the enriched KEGG pathway analyses of the miRNAs targeted genes were performed by bioinformatics software DAVID. Then protein-protein interaction networks of the targeted gene were conducted by online software STRING. Finally, using microarray, bioinformatics approaches revealed the possible molecular mechanisms pathogenesis of CAD and T2DM. A total of 11 differentially expressed miRNAs were identified and among them, hsa-miR-4687-3p drew specific attention. Bioinformatics analysis revealed that insulin signaling pathway is the central way involved in the progression of metabolic disorders. Conclusions. The current findings support the fact that miRNAs are involved in the pathogenesis of metabolic disorders in EAT of CAD patients with T2DM, and validation of the results of these miRNAs by independent and prospective study is certainly warranted. National Natural Science Foundation of China2014BB019Xinjiang Production and Construction Crops ==== Body 1. Introduction Coronary artery disease (CAD) remains one of the most common causes of morbidity and mortality in diabetic patients [1]. Thus improving the understanding of the etiology associated with CAD is highly important. Epicardial adipose tissue (EAT) is suggested to play an important role in the progression of metabolic syndrome [2]. Several findings implicate that EAT thickness may be a useful indicator for T2DM and obesity [2, 3]. Studies have shown that EAT generates several bioactive molecules, such as anti- and proinflammatory mediators and cytokines [4], which may significantly enhance paracrine effects on cardiac function or produce a variety of effects that affect many physiological processes [5]. Nowadays, beside the main factors including obesity, hypertension, and dyslipidemia, novel risk factors such as chronic low-grade inflammation, oxidative stress, and endothelial dysfunction are accepted as the decisive factors to highlight this increased cardiovascular risk in human beings [6, 7]. Experimental and clinical studies have suggested that EAT may cause CAD [8]. EAT, visceral fat depot of the heart, was found to be associated with CAD, T2DM, and other metabolic disorders [6, 9, 10]. EATs are metabolically active visceral fat deposits found around the heart, between the pericardium and myocardium [11], which are strongly associated with cardiovascular diseases (CVD) including CAD and the development of cardiac arrhythmias, predominantly due to the secretion of bioactive mediators and cytokines [12]. T2DM plays a key role in the development of CVD. EAT has great interplay with diabetic patients and has potential to influence CVD. Owing to its close proximity to the heart and coronary vasculature, EAT exerts a direct metabolic impact by secreting free fatty acids and proinflammatory factors and decreased anti-inflammatory adipokines, which promote CVD locally [9]. MicroRNAs (miRNAs) are a class of about 21–25 nucleotides in length and small noncoding RNA molecules with essential roles, of which any alteration leads to several conditions. They were first identified in 1993, and the term microRNA was created in 2001 [13]. The main functions of miRNAs are to downregulate the target gene expression in translational repression and cleavage of mRNA and in a wide range of biological processes [13]. Decisive regulatory functions exhibited by the miRNA are associated with various human diseases such as human cancer and heart disease. In addition to the link with cancer, microRNAs play a vital role in the control of cardiac-related diseases. For multiple forms of heart disease, including ventricular wall, maintenance of cardiac rhythm myocyte growth, and contractility, the misexpressions of miRNAs were shown to be necessary. However, in the literature, data regarding the relationship between miRNAs and EAT in CAD patients T2DM is scant. Hence, we aimed to investigate the role of miRNAs in EAT. We also sought to predict the targets of novel miRNAs in EAT in patients compared with the results that are obtained from control subjects [6]. The current study was carried out to assess the changes in the EAT levels of miRNAs in subjects suffering from T2DM and CAD compared to healthy control ones. In addition, bioinformatics analyses were carried out in order to find out how these possible miRNAs are associated with the incidence and pathogenesis of CAD on top of T2DM. In the present study, we depicted comparative miRNAs in “metabolically healthy” patients without metabolic disorders and in metabolic patients with CAD and T2DM. This strategy allowed us to identify a set of miRNAs characterizing EAT in health and disease as well as potential novel biological processes characterizing EAT in CAD patients with T2DM. 2. Methods 2.1. Subjects EAT samples were taken from 10 subjects of both genders, in the Department of Thoracic and Cardiovascular Surgery, the First Affiliated Hospital of Shihezi University School of Medicine, in which levels of miRNAs expression in EAT (using microarray) and routine parameters were measured. Subjects were divided into two groups, 5 in each group as follows: patients with T2DM and CAD and metabolically healthy control subjects without T2DM and CAD. Ten Asian patients were recruited, 5 underwent cardiac valve surgery (no evidences of CAD/T2DM/carotid atherosclerosis/metabolic syndrome), and 5 underwent coronary artery bypass graft surgery (CAD + T2DM group). The study protocol was approved by the Medical Ethics Committee of Shihezi University (School of Medicine, Xinjiang, China). Written informed consent was obtained from all subjects included in the study. This was a cross-sectional study and a review of medical records (including information on sex, age, height, weight, medications, disease duration, smoking, and history of other diseases) was undertaken. Control subjects were chosen from metabolically healthy individuals according to NCEP ATP III Metabolic syndrome criteria (two or less metabolic criteria; TG ≥ 1.7 mmol/L, blood pressure ≥ 130/85 mmHg, glucose ≥ 5.6 mmol/L, HDL-C: men < 1.03 mmol/L and women < 1.30 mmol/L, and waist circumference: men < 102 cm and women < 88 cm) as previous said [6, 14]. miRNAs microarray expression analyses were conducted on RNA extracted from perivascular EAT using the Human-MicroRNA Expression Kits. 2.2. RNA Extraction and Purification Total RNA, including the miRNAs, was extracted from the EAT and purified using mirVana™ miRNA Isolation Kit (Cat. number AM1561, Ambion, Austin, TX, US), following the manufacturer's instructions, and checked for a RNA integrity number (RIN) to inspect RNA integration by an Agilent Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, US). RIN ≥ 6.0 and 28S/18S ≥ 0.7 were used for the miRNA array analysis. 2.3. RNA Labeling miRNA molecular in total RNA was labeled by miRNA Complete Labeling and Hyb Kit (Cat. number 5190-0456, Agilent Technologies, Santa Clara, CA, US) following the manufacturer's instructions, labeling section. 2.4. Array Hybridization MiRNA microarray assays were performed using the Agilent Human miRNA (8 ∗ 60 K) V21.0 microarray platform (design ID: 70156) at Shanghai Biotechnology Co., Ltd. (Shanghai, China). Each slide was hybridized with 100 ng Cy3-labeled RNA using miRNA Complete Labeling and Hyb Kit (Cat. number 5190-0456, Agilent Technologies, Santa Clara, CA, US) in hybridization oven (Cat. number G2545A, Agilent Technologies, Santa Clara, CA, US) at 55°C, 20 rpm for 20 hours according to the manufacturer's instructions, hybridization section. After hybridization, slides were washed in staining dishes (Cat. number 121, Thermo Shandon, Waltham, MA, US) with Gene Expression Wash Buffer Kit (Cat. number 5188-5327, Agilent Technologies, Santa Clara, CA, US). 2.5. Data Acquisition Slides were scanned by Agilent Microarray Scanner (Cat. number G2565CA, Agilent Technologies, Santa Clara, CA, US) and Feature Extraction software 10.7 (Agilent Technologies, Santa Clara, CA, US) with default settings. Raw data were normalized by Quantile algorithm, Gene Spring Software 12.6 (Agilent Technologies, Santa Clara, CA, US). 2.6. Differential miRNAs Targeted Gene Prediction The differential miRNAs targets predicted by computer-aided algorithms were obtained from TargetScan, PicTar, and miRBase targets [15]. More detailed information can be acquired from online software (http://pictar.mdc-berlin.de/cgi-bin/new_PicTar_vertebrate.cgi; http://microrna.sanger.ac.uk/cgi-bin/targets/v5/search.pl; http://www.targetscan.org/). 2.7. The Interaction Network and Signaling Pathway Analysis of Differential microRNA and mRNA DAVID [16–18], a bioinformatics analysis software, is used for the analysis of the enriched KEGG (Kyoto Encyclopedia of Genes and Genomes) signaling pathway analysis for the interactions between microRNAs and mRNAs (http://david.abcc.ncifcrf.gov/). Online software Gene Ontology (http://geneontology.org/) was employed to perform GO enrichment analysis [19, 20]. The microRNA and mRNA of differential expression in patients were uploaded to DAVID and Gene Ontology for analysis. 2.8. Protein-Protein Interactions (PPI) Network Analysis A number of abnormal mRNAs were found in the interaction analysis between miRNAs and mRNA. Therefore, to further understand the function of microRNA in the network, the PPI analysis was performed in the protein products of mRNAs to find out the key proteins. The selected targeted genes were put into the STRING (Search Tool for the Retrieval of Interacting Genes) database (http://string-db.org/), a metaresource that collects most of the available information on protein-protein associations and scores and weights it and augments it with predicted interactions and with the results of automatic opuses-mining searches to match the interactions of proteins [21, 22]. 2.9. Statistical Analysis Descriptive statistics for each variable were determined. Results for continuous variables were demonstrated as mean ± standard deviation. Statistical significant difference between the groups was determined by the chi-square test for categorical variables and unpaired Student's t-test for continuous variables. Differentially expressed targeted genes were studied using bioinformatics analysis and statistical analysis allowed algorithm of the selected software. 3. Results 3.1. Baseline Characteristics of Patients The baseline characteristics of 5 patients and 5 control subjects were shown in Table 1. There were no differences with respect to the following variables between patients and control subjects, age, gender, waist circumference (WC), systolic blood pressure (SBP), diastolic blood pressure (DBP), and body mass index (BMI). Compared to “controls,” CAD patients with T2DM were characterized by significantly increased waist-hip ratio, LDL-C, and systemic inflammation, while HDL-C was decreased. 3.2. Data of Microarray In order to measure the miRNAs expression patterns that characterize EAT in CAD patients with T2DM from EAT in control group, we used the whole-genome miRNAs microarrays. Overall, the resulting signal intensity of miRNAs genes is statistically different at the Wilcoxon signed-rank test in EAT in both groups, thus underscoring the profound diversity of EAT. Unsupervised hierarchical clustering was presented in Figure 1. Compared to metabolically healthy patients, we identified forty-two miRNAs that are differentially expressed in patients with CAD and T2DM (26 downregulated; 16 upregulated, data was not shown). Eleven miRNAs were selected as potential novel miRNAs according to P value and fold change (Table 2: 6 were significantly downregulated; 5 were significantly upregulated in both subgroups of patients with P < 0.05; fold change > 2 times). 3.3. Results of Bioinformatics Analyses To depict the possible role of miRNAs in EAT, we selected these miRNAs as potential novel biomarkers that were significantly different (patients versus controls) in the overall population (P < 0.05; fold change > 2). To provide a framework for interpretation of our results, we then functionally clustered significant biological pathways using the bioinformatics analyses. 3.3.1. Bioinformatics Analyses of miRNAs Targeted Genes In order to investigate the possible regulation mechanisms of miRNAs in the process of CAD complicated with T2DM, we first utilized three online bioinformatics databases (TargetScan, PicTar, and miRBase targets) to select plausible targets and validated targets of miRNAs and finally obtained target genes for the following analysis; then we analyzed biological processes, molecular functions, and cellular components through Gene Ontology and enriched KEGG pathways by DAVID. The results showed that the predicted target genes mainly are enriched in the following biological processes: detection of chemical stimulus involved in sensory perception of smell, G-protein coupled receptor signaling pathway, translation, axon guidance, protein phosphorylation, and so forth (P < 0.001), they mainly are enriched in the following molecular functions: G-protein coupled receptor activity, olfactory receptor activity, protein binding, structural constituent of ribosome, and so forth (P < 0.001), and significant cellular components were extracellular region and Golgi apparatus (P < 0.001). Enriched KEGG pathways by DAVID displayed in KEGG pathway database showed that the predicted target genes of miRNAs were significantly enriched in the insulin signaling pathway (Figure 2), adipocytokine signaling pathway, MAPK signaling pathway, FoxO signaling pathway, and other signaling pathways (P < 0.05). 3.3.2. Protein-Protein Interaction (PPI) Network We analyzed the protein-protein interaction network of selecting miRNAs target genes using STRING 10 and removed the target genes which were linked to lax isolated nodes through data analysis; the result showed that the interaction existed in total 148 proteins targeted by the predicted genes, which together formed the target gene interaction network. The network consists of 148 nodes which represent 148 proteins and many lines with different colors that represent the types of evidence for the association. From the result we can see that PDPK1, PIK3R3, PPP1R3B, PRKAR1A, SOCS3, SREBF1, PPARGC1A, SHC4, MAPK1, GRB2, and MKNK2 played key roles in maintaining stability in the network, especially PIK3R3, MAPK1, and GRB2 whose connections were very close, so the protein encoded by them may be important downstream target proteins (Figure 3). 4. Discussion Better understanding of the biological characteristics can provide vital theoretical basis for the prevention and treatment of disease. The aim of this study was to investigate the profiles of miRNAs and the interaction network of novel microRNA and mRNA as well as related signaling pathway in EAT through analyzing the expression profile of microRNA and mRNA to provide novel insights in the biological characteristics in CAD patients with T2DM. According to newly added studies, it is clear that miRNAs are involved in the regulation of metabolic and inflammation functions, and the alterations in CAD and T2DM diseases have been analyzed through the identification and evaluation of miRNAs profiles in patients as in animal models [13, 23, 24]. microRNAs are key components of many cellular processes. Different studies have demonstrated that miRNA expression is tissue-specific, tightly regulated during embryogenesis, and overexpressed/underexpressed in many diseases, including CAD and T2DM pathologies [25]. They are easily detected in a quantitative way by real-time polymerase chain reaction (qRT-PCR) or microarrays and by other less frequently used identification methods, such as PCR-based restriction fragment length polymorphisms (PCR-RLFP), traditional northern blotting, direct sequencing using next generation sequencing (NGS), and platforms ligation based measurement [26]. Risk prediction for T2DM and CVD remains suboptimal even after the introduction of global risk assessment by various methods. This has prompted the search for additional biomarkers [27]. EAT is a source of several inflammatory mediators in high-risk cardiac patients [4]. Adipose tissue may function as an endocrine organ that contributes to an inflammatory burden in patients at risk of CVD. Obesity, adiposopathy, and insulin resistance induce EAT enlargement, inflammation, and dysfunction and trigger CAD [10]. EAT is an atypical fat depot surrounding the heart with a putative role in the development of atherosclerosis [28]. Given the close anatomic relationship between perivascular EAT and coronary arteries and the positive correlation between EAT and the presence of coronary atherosclerosis, several results point to EAT as a putative actor in CAD and/or T2DM [3, 10]. EAT thickness is an independent risk factor for CAD. EAT could thus be able to modulate heart and coronary artery pathophysiology, and mounting evidences point to EAT as a candidate player in pathophysiology of CAD. miRNAs act on multiple targets and complex pathogenesis, thus representing candidate regulators of adipocyte differentiation, metabolic homeostasis, and inflammation. miRNAs have also been described as differentially modulated in adipose tissue during metabolic disease, thus being considered candidate biomarkers for metabolic disorders, CAD, T2DM, and putative targets for therapy [28]. We present evidence that a profile of miRNAs was dysregulated in EAT in CAD patients with T2DM compared to metabolic health patients, supporting the concept that miRNAs in EAT are involved in pathogenesis of CAD and/or T2DM. The current study identified a total of 11 differentially expressed miRNAs, and among them, hsa-miR-4687-3p drew specific attention for the largest targeted genes which were identified. Particularly, bioinformatics analysis revealed that insulin signaling pathway is the central way involved in the progression of metabolic disorders. Previous studies indicated that the circulating levels of miR-133a can be used as a predictor for diagnosing CAD, since increased miR-133a level may be used to predict both the presence and severity of coronary lesions in CAD patients [29]. Evidence from a recent study had showed that miR-370 was significantly increased in patients with T2DM and CAD and CAD only and patients with T2DM [23]. This was also concomitant with another study that showed the use of certain miRNA imprints including miR-370, in the screening of patients at risk for developing CAD [30]. Because of the highly metabolic paracrine and endocrine functions of EAT, it has been proposed to play a role in the pathogenesis of CVD by releasing proinflammatory and proatherogenic factors. EAT plays important roles in CAD, not only in its location, but also by its blood supply. EAT derives its blood supply from coronary circulation. There is a functional and anatomic relationship between EAT and muscular components of the heart as these components share the same coronary blood supply. The release of proinflammatory and proatherogenic factors into the circulation advancing CVD is more significantly linked to the progression of CAD [10]. It is the close anatomical relationship between EAT and the coronary arteries combined with its biologically active properties that participates in the pathogenesis of diabetic coronary atherosclerosis [31]. EAT-specific miRNAs are miR-196b-5p, miR-196a-5p (a promoter of brown adipogenesis), miR-18a-3p (a member of the miR-17/92 cluster that promotes adipocyte differentiation), and miR-10a-3p (an anti-inflammatory agent) which were analyzed by one article that also supported our study. In EAT of CAD patients, miR-135b-3p (a direct target of inflammatory pathways) was found upregulated, while miR-455-3p (a driver of during brown adipocyte differentiation), miR-193b-3p (promoting adiponectin secretion in human adipocytes), and Let-7a-3p and miR-127-3p (negative modulators of inflammatory pathways) were found downregulated [32, 33]. These findings are in accordance with our results that miRNAs are involved in the pathogenesis of metabolic disorders in EAT of CAD patients with T2DM. Overall, we were able to depict a novel miRNA signature of EAT in CAD patients with T2DM characterized by dysregulated miRNAs profile which are probably involved in pathogenesis of CAD. The results of the current study support the hypothesis that miRNA expression is deregulated in epicardial adipose tissue in patients suffering from T2DM and CAD disease. We provided a comprehensive potential novel miRNAs expression signature of EAT in CAD patients with T2DM, and we showed that the targets of the miRNAs are necessarily associated with the pathogenesis of CAD and T2DM. Acknowledgments The authors would like to acknowledge the help from the First Hospital Affiliated to Medical College of Shihezi University. The work was supported in part by NFSC 81460059 from the National Science Foundation of China and doctor/scientific Grant 2014BB019 from Xinjiang Production and Construction Crops. Disclosure Yang Liu and Wenbo Fu should be considered co-first authors. Competing Interests The authors declare that they have no competing interests regarding the publication of this paper. Authors' Contributions Yang Liu and Wenbo Fu contributed equally to this work. Figure 1 Unsupervised hierarchical clustering (heat map). Heat map generated by hierarchical clustering for differentially expressed miRNAs in the EAT from CAD with T2DM patients versus control subjects. Hierarchical clustering for differentially expressed miRNAs in CAD with T2DM (n = 4) versus control (n = 5) (P < 0.05 and fold change > 2 times). Columns display the clustering of EAT samples; rows show the clustering of genes. The expression intensity of each miRNA in each sample varies from red to green, which indicates relative high or low expression, respectively. Expression clusters representing different patterns of upregulation to downregulation are depicted on the side of figure. Figure 2 Insulin signaling pathway. ERK1/2(MAPK1), GRB2, PKA (PRKAR), P13K (PIK3), and PDK1/2 (PDPK1) are the main target genes of differential miRNAs. Figure 3 Protein-protein interaction (PPI) network of target proteins of the selecting miRNAs. Table 1 Baseline characteristics of patients.   Non-CAD + T2DM (n = 5) CAD + T2DM (n = 5) P value Age (years) 61.8 ± 5.2 54.6 ± 7.0 0.392 BMI (kg/m2) 29.47 ± 5.83 28.63 ± 4.26 0.213 WC (cm) 99.64 ± 10.67 101.23 ± 10.34 0.492 Waist-hip ratio 0.96 ± 0.14 0.99 ± 0.09 <0.001 FPG (mg/dL) 106.5 (98.5–115.5) 116.0 (100.5–138.0) 0.103 Total cholesterol 166.48 ± 37.10 159.55 ± 39.41 0.335 LDL-C (mg/dL) 93.51 ± 26.59 98.44 ± 23.14 <0.01 HDL-C (mg/dL) 40.66 ± 10.08 34.39 ± 9.64 <0.01 Triglyceride (mg/dL) 139 (105.5–223.5) 154 (106.0–234.0) 0.276 hsCRP (mg/dL) 1.03 ± 2.31 3.15 ± 5.22 <0.001 Adiponectin (μg/mL) 12.37 ± 6.55 9.16 ± 4.78 <0.05 Fibrinogen (mg/dL) 523.0 (453.0–638.0) 603.5 (510.5–766.5) <0.01 FPG: fasting plasma glucose, LDL-C: low density lipoprotein cholesterol, HDL-C: high density lipoprotein cholesterol, and hsCRP: hypersensitive C reactive protein. Adiponectin, hsCRP, and fibrinogen indicate systemic inflammation. Table 2 Disregulated miRNAs (CAD + T2DM versus control). miRNA P value Fold change hsa-miR-4429 0.04437405 0.16922769 hsa-miR-409-3p 0.012379586 0.184890055 hsa-miR-6802-5p 0.047496135 0.305662149 hsa-miR-5703 0.008804377 0.40973895 hsa-miR-630 0.020813173 0.43854972 hsa-miR-4687-3p 0.034377945 0.44430727 hsa-miR-3651 0.04317499 2.063333894 hsa-miR-574-3p 0.025198143 2.156938357 hsa-miR-619-5p 0.00179165 2.179715733 hsa-miR-664b-3p 0.04815202 2.325766896 hsa-miR-146b-5p 0.044983658 2.450112684 Lists of the deregulated miRNAs between CAD + T2DM and control. Fold change < 0.5 indicated downregulation significantly and fold change > 2 indicated upregulation significantly. ==== Refs 1 Coutinho M. Gerstein H. C. Wang Y. Yusuf S. The relationship between glucose and incident cardiovascular events. A metaregression analysis of published data from 20 studies of 95,783 individuals followed for 12.4 years Diabetes Care 1999 22 2 233 240 10.2337/diacare.22.2.233 2-s2.0-0032902850 10333939 2 Song D. K. Hong Y. S. Lee H. Oh J.-Y. Sung Y.-A. Kim Y. Increased epicardial adipose tissue thickness in type 2 diabetes mellitus and Obesity Diabetes and Metabolism Journal 2015 39 5 405 413 10.4093/dmj.2015.39.5.405 2-s2.0-84947429178 26566498 3 Mohar D. S. Salcedo J. Hoang K. C. Epicardial adipose tissue volume as a marker of coronary artery disease severity in patients with diabetes independent of coronary artery calcium: Findings from the CTRAD Study Diabetes Research and Clinical Practice 2014 106 2 228 235 10.1016/j.diabres.2014.08.021 2-s2.0-84920551969 25262111 4 Mazurek T. Zhang L. Zalewski A. Human epicardial adipose tissue is a source of inflammatory mediators Circulation 2003 108 20 2460 2466 10.1161/01.cir.0000099542.57313.c5 2-s2.0-0344393607 14581396 5 Iacobellis G. Leonetti F. Epicardial adipose tissue and insulin resistance in obese subjects Journal of Clinical Endocrinology and Metabolism 2005 90 11 6300 6302 10.1210/jc.2005-1087 2-s2.0-27744581763 16091479 6 Akbas E. M. Hamur H. Demirtas L. Predictors of epicardial adipose tissue in patients with type 2 diabetes mellitus Diabetology and Metabolic Syndrome 2014 6 1, article 55 10.1186/1758-5996-6-55 2-s2.0-84900823801 7 Baker A. R. Silva N. F. D. Quinn D. W. Human epicardial adipose tissue expresses a pathogenic profile of adipocytokines in patients with cardiovascular disease Cardiovascular Diabetology 2006 5, article 1 2-s2.0-34447618577 8 Xu Y. Cheng X. Hong K. Huang C. Wan L. How to interpret epicardial adipose tissue as a cause of coronary artery disease: a meta-analysis Coronary Artery Disease 2012 23 4 227 233 10.1097/mca.0b013e328351ab2c 2-s2.0-84862777149 22361934 9 Noyes A. M. Dua K. Devadoss R. Chhabra L. Cardiac adipose tissue and its relationship to diabetes mellitus and cardiovascular disease World Journal of Diabetes 2014 5 6 868 876 10.4239/wjd.v5.i6.868 25512789 10 Iacobellis G. Bianco A. C. Epicardial adipose tissue: emerging physiological, pathophysiological and clinical features Trends in Endocrinology and Metabolism 2011 22 11 450 457 10.1016/j.tem.2011.07.003 2-s2.0-82455199186 21852149 11 Iacobellis G. Corradi D. Sharma A. M. Epicardial adipose tissue: anatomic, biomolecular and clinical relationships with the heart Nature Clinical Practice Cardiovascular Medicine 2005 2 10 536 543 10.1038/ncpcardio0319 2-s2.0-27944453528 16186852 12 Wang C.-P. Hsu H.-L. Hung W.-C. Increased Epicardial Adipose Tissue (EAT) volume in type 2 diabetes mellitus and association with metabolic syndrome and severity of coronary atherosclerosis Clinical Endocrinology 2009 70 6 876 882 10.1111/j.1365-2265.2008.03411.x 2-s2.0-66149112904 18778397 13 Srinivasan S. Selvan S. T. Archunan G. Gulyas B. Padmanabhan P. MicroRNAs -the next generation therapeutic targets in human diseases Theranostics 2013 3 12 930 942 10.7150/thno.7026 2-s2.0-84904285776 24396504 14 Grundy S. M. Cleeman J. I. Merz C. N. B. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III Guidelines Journal of the American College of Cardiology 2004 44 3 720 732 10.1016/j.jacc.2004.07.001 2-s2.0-4544275379 15358046 15 Fu H. L. Wu D. P. Wang X. F. Altered miRNA expression is associated with differentiation, invasion, and metastasis of esophageal squamous cell carcinoma (ESCC) in patients from Huaian, China Cell Biochemistry and Biophysics 2013 67 2 657 668 10.1007/s12013-013-9554-3 2-s2.0-84886596921 23516093 16 Huang D. W. Sherman B. T. Lempicki R. A. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists Nucleic Acids Research 2009 37 1 1 13 10.1093/nar/gkn923 2-s2.0-58549112996 19033363 17 Huang D. W. Sherman B. T. Tan Q. DAVID bioinformatics resources: expanded annotation database and novel algorithms to better extract biology from large gene lists Nucleic Acids Research 2007 35 2 W169 W175 10.1093/nar/gkm415 2-s2.0-34547589578 17576678 18 Huang D. W. Sherman B. T. Tan Q. The DAVID gene functional classification tool: a novel biological module-centric algorithm to functionally analyze large gene lists Genome Biology 2007 8 9, article R183 10.1186/gb-2007-8-9-r183 2-s2.0-41349115432 19 Boyle E. I. Weng S. Gollub J. GO::TermFinder—open source software for accessing Gene Ontology information and finding significantly enriched Gene Ontology terms associated with a list of genes Bioinformatics 2004 20 18 3710 3715 10.1093/bioinformatics/bth456 2-s2.0-12344269924 15297299 20 Cheng J. Sun S. Tracy A. NetAffix gene ontology mining tool: a visual approach for microarray data analysis Bioinformatics 2004 20 9 1462 1463 10.1093/bioinformatics/bth087 2-s2.0-3142750599 14962933 21 Jensen L. J. Kuhn M. Stark M. STRING 8-a global view on proteins and their functional interactions in 630 organisms Nucleic Acids Research 2009 37 1 D412 D416 10.1093/nar/gkn760 2-s2.0-58149193234 18940858 22 Szklarczyk D. Franceschini A. Kuhn M. The STRING database in 2011: functional interaction networks of proteins, globally integrated and scored Nucleic Acids Research 2011 39 1 D561 D568 10.1093/nar/gkq973 2-s2.0-78651324347 21045058 23 Motawae T. M. Ismail M. F. Shabayek M. I. Seleem M. M. MicroRNAs 9 and 370 association with biochemical markers in T2D and CAD complication of T2D PLoS ONE 2015 10 5, article e0126957 10.1371/journal.pone.0126957 2-s2.0-84929340153 24 Karolina D. S. Tavintharan S. Armugam A. Circulating miRNA profiles in patients with metabolic syndrome Journal of Clinical Endocrinology and Metabolism 2012 97 12 E2271 E2276 10.1210/jc.2012-1996 2-s2.0-84870746025 23032062 25 Bronze-Da-Rocha E. MicroRNAs expression profiles in cardiovascular diseases BioMed Research International 2014 2014 23 985408 10.1155/2014/985408 2-s2.0-84903647962 26 Etheridge A. Lee I. Hood L. Galas D. Wang K. Extracellular microRNA: a new source of biomarkers Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesi 2011 717 1-2 85 90 10.1016/j.mrfmmm.2011.03.004 2-s2.0-82755187564 27 Herder C. Karakas M. Koenig W. Biomarkers for the prediction of type 2 diabetes and cardiovascular disease Clinical Pharmacology and Therapeutics 2011 90 1 52 66 10.1038/clpt.2011.93 2-s2.0-79959378842 21654741 28 Vacca M. Di Eusanio M. Cariello M. Integrative miRNA and whole-genome analyses of epicardial adipose tissue in patients with coronary atherosclerosis Cardiovascular Research 2016 109 2 228 239 10.1093/cvr/cvv266 26645979 29 Wang F. Long G. Zhao C. Plasma microRNA-133a is a new marker for both acute myocardial infarction and underlying coronary artery stenosis Journal of Translational Medicine 2013 11 1, article 222 10.1186/1479-5876-11-222 2-s2.0-84884374796 30 Hoekstra M. van der Lans C. A. C. Halvorsen B. The peripheral blood mononuclear cell microRNA signature of coronary artery disease Biochemical and Biophysical Research Communications 2010 394 3 792 797 10.1016/j.bbrc.2010.03.075 2-s2.0-77950518455 20230787 31 Kremen J. Dolinkova M. Krajickova J. Increased subcutaneous and epicardial adipose tissue production of proinflammatory cytokines in cardiac surgery patients: possible role in postoperative insulin resistance The Journal of Clinical Endocrinology & Metabolism 2006 91 11 4620 4627 10.1210/jc.2006-1044 2-s2.0-33751537132 16895955 32 Marchington J. M. Mattacks C. A. Pond C. M. Adipose tissue in the mammalian heart and pericardium: structure, foetal development and biochemical properties Comparative Biochemistry and Physiology—Part B 1989 94 2 225 232 10.1016/0305-0491(89)90337-4 2-s2.0-0024820494 33 Gaborit B. Venteclef N. Ancel P. Human epicardial adipose tissue has a specific transcriptomic signature depending on its anatomical peri-atrial, peri-ventricular, or peri-coronary location Cardiovascular Research 2015 108 1 62 73 10.1093/cvr/cvv208 2-s2.0-84943416994 26239655

PMC005xxxxxx/PMC5002304.txt

==== Front Biomed Res IntBiomed Res IntBMRIBioMed Research International2314-61332314-6141Hindawi Publishing Corporation 10.1155/2016/2571080Review ArticleEffect of Ulinastatin in the Treatment of Postperative Cognitive Dysfunction: Review of Current Literature http://orcid.org/0000-0002-8238-7194Lv Zheng-tao http://orcid.org/0000-0002-6800-2788Huang Jun-ming Zhang Jin-ming Zhang Jia-ming Guo Jin-feng http://orcid.org/0000-0002-0729-8212Chen An-min * Department of Orthopedics, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei 430030, China*An-min Chen: anminchen@hust.edu.cnAcademic Editor: Fabrizio Montecucco 2016 14 8 2016 2016 257108012 3 2016 8 6 2016 9 6 2016 Copyright © 2016 Zheng-tao Lv et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background. Ulinastatin, identified as a urinary trypsin inhibitor, has been widely used in patients with inflammatory disorders. However, little is known about its effect on postoperative cognitive dysfunction (POCD). The aim of our current work is to review the current body of literature. Methods. A systematic literature search in PubMed and EMBASE was performed to identify randomized controlled trials. Incidence of POCD, MMSE score, and laboratory indicators (IL-6, TNF-α, CRP, and S100β) were selected as outcomes. Results. Five RCTs involving 461 elderly patients that underwent surgical operations were identified. The meta-analysis suggested no statistical difference of incidence of POCD between ulinastatin and control groups on postoperative day 1; but ulinastatin could significantly decrease the incidence of POCD on postoperative day 3 and day 7 when compared with control treatment. Ulinastatin was effective in improving the MMSE score on day 1, day 3, and day 7 after operation. IL-6 and S100β concentrations were lower up to postoperative day 2. The incidences of postoperative complications in ulinastatin groups were lower than control. Conclusion. Ulinastatin administration was effective in treating early POCD (postoperative day 3 and day 7) and reducing IL-6 and S100β concentrations within two days after operations. Studies with larger-scale and rigorous design are urgently needed. National High Technology Research and Development Program of China2012AA02A612National Natural Science Foundation of China81472082811716968137191581572094 ==== Body 1. Introduction Postoperative cognitive dysfunction (POCD) is a common situation that may occur after any sort of surgery and defined by a drop in cognitive domain on a set of neuropsychological tests from before to after surgery; it broadly refers to a “deterioration in cognition temporally associated with surgical operation” [1]. POCD impacts a wide variety of cognitive performance, for instance, memory, information processing, and executive function. The main subjective complaint is deterioration in memory, and some patients even find it hard to manage their jobs. The pathophysiology and etiology of POCD are relatively unknown, and higher age and preexisting cognitive impairment have been identified to date as risk factors [2–4]. Animal studies have suggested excessive neuroinflammation after surgery [5]; failure to effectively control inflammation may involve development of POCD [6, 7]. Using different preclinical models of noncentral nervous system surgery, neuroinflammation has been repeatedly associated with behavioral dysfunction and memory deficits. Upregulation of systemic inflammatory mediators and cytokines, including tumor necrosis factor-alpha (TNF-α), interleukin- (IL-) 1, interleukin- (IL-) 6, and C-reactive protein (CRP), has been shown to activate bone-marrow derived macrophages and contribute to the overall brain pathology after aseptic trauma [6–9]. Elevated serum levels of S100β, which indicate continuous tissue damage, have been reported to correlate with neurological deterioration after cardiac surgery and with poor likelihood of survival after hypoxia [10]. Ulinastatin not only have function in blocking the protease pathway, but have anti-inflammation proper in vitro as well [11]. In recent years, ulinastatin was wildly used in the treatment of a variety of severe diseases, such as disseminated intravascular pancreatitis, shock, and disseminated or diffuse intravascular coagulation [12, 13]. Some reports also support the role of protective in acute lung injury (ALI) and acute respiratory distress syndrome (ARDS) and multiple organ dysfunction syndrome (MODS) [14, 15]. The protective mechanism of ulinastatin may be attributed partly to the suppression of NF-Kappa B pathway [16] and mitogen-activated protein kinases (MAPKs) [17]. The suppression of aforementioned pathways could downregulate the production of proinflammatory cytokines, including TNF-α, IL-1, IL-6, and CRP [17–19]. Surgery is associated with a central neuroinflammatory response in humans. The inflammatory response may associate with incidence of POCD. Ulinastatin can significantly downregulate occurrence of inflammation. In addition, ulinastatin reported a possible protection of brain in gene level. Ulinastatin has certain therapeutic effects on cerebral ischemic reperfusion injury through upregulation expression of nerve growth factor and brain-derived neurotrophic factor and downregulation apoptosis of oligodendrocytes [20, 21]. However, little is known about its beneficial effect on POCD. Thus, the aim of our current systematic review and meta-analysis was to evaluate the clinical effect of ulinastatin on POCD based on randomized controlled trials (RCTs). 2. Methods and Materials This systematic review was performed in accordance to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [22]. 2.1. Search Strategy A comprehensive literature search was conducted in EMBASE and PubMed. Two electronic databases were searched from their inception date to the latest issue (January 2016), without language restriction. A combination of medical subject headings (MeSH) and free terms was used based on the specifications of each database. The search strategy for PubMed was as follow: (“urinastatin” [Supplementary Concept] or ulinastatin or UTI68 or urinary trypsin inhibitor or MR20 or miraclid) and ((Post-operative cognit∗  or Postoperative cognit∗  or POCD) and (surgery or operation) and (cognit∗  OR intelligence OR MMSE OR Mini Mental OR dementia OR Alzheim∗  OR mild cognitive impairment OR MCI)). The bibliographies of related systematic reviews and clinical guidelines were also searched. In addition, the reference section for each study was also searched. 2.2. Inclusion and Exclusion Criteria We made our inclusion and exclusion criteria in adherence to the PICOS principle. P: subjects enrolled in our systematic review were patients undergoing surgical operations and no restriction on race, age, and gender was imposed; I: patients in ulinastatin groups were treated with ulinastatin intravenously before or/and after surgeries; C: control groups received a placebo administration of normal saline; the volume of normal saline must be the same as ulinastatin administration; O: the primary outcome measures included incidence of POCD and MMSE score; secondary outcomes were IL-6, TNF-α, CRP, and S100β; S: study design was restricted to RCTs. Case reports, case series, book chapters, and editorials were excluded. 2.3. Data Extraction Two investigators (Zheng-tao Lv and Jun-ming Huang) screened each article independently and were blinded to the findings of the other reviewer. Following the prespecified inclusion criteria, two reviewers performed a rigorous screening to identify eligible articles. Data were collected from these selected articles using a standardized data collection sheet, which included first author, country, year of the publication, study design, cohort sizes, demographic characteristics of participants in different groups, details of intervention and control, and main outcomes. Discrepancies between two reviewers was resolved through discussion until a general consensus could be reached. The third review author (Jin-ming Zhang) was sought for opinions if a consensus could not be reached. 2.4. Risk of Bias Assessment To assess the risk of bias among our included studies, the Cochrane Collaboration's tool was utilized, which was based on seven items: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other sources of bias [23]. Two reviewers (Zheng-tao Lv and Jun-ming Huang) judged the risk of bias among studies independently; the results were compared afterwards. In case of disagreements regarding the risk of bias judgement, discussion was conducted until a consensus was reached. 2.5. Data Synthesis For the incidence of POCD, risk ratio (RR) and the associated 95% confidence interval (95% CI) were calculated using the Rev Man 5.3 (Copenhagen: the Nordic Cochrane Centre, the Cochrane Collaboration, 2014). Mean difference (MD) and the associated 95% CI were calculated for continuous variables using the same methodology. Before the combination of data from individual study, the chi-squared test and the Higgins I 2 test were used to assess the heterogeneity among studies (P > 0.1 and I 2 indicate acceptable heterogeneity). The fixed-effect model was used for statistical analysis if no obvious heterogeneity existed; random-effect model was employed if apparent heterogeneity existed. Metaregression was performed to find the possible source of heterogeneity. Begg's rank correlation test and Egger's linear regression test were used to evaluate the publication bias if the number of included studies was greater than ten. 3. Results 3.1. Literature Search An initial literature search yielded a total of 22 potential relevant citations including 7 from PubMed and 15 from EMBASE; 5 duplicates were deleted. At the stage of titles and abstracts screen, 12 articles were excluded because they were not related with ulinastatin in the treatment of POCD, and the remaining 5 articles were retrieved for a full-text review. Finally, 5 studies met our predetermined inclusion criteria (Figure 1). 3.2. Study Characteristics All the included studies were conducted in China and published from 2010 to 2015; each study was undertaken in a single center. These five RCTs [24–28] included a total of 461 elderly patients that underwent surgical operations: 242 in the ulinastatin groups and 219 in the control groups. Three two-arm parallel RCTs [24, 25, 27] enrolled patients that underwent orthopedic operations; a prospective double-blind RCT [26] recruited participants scheduled for abdominal surgeries. Ge et al. [28] randomly assigned patients into three groups to investigate the clinical effect of ulinastatin on POCD: high dose ulinastatin group (16000 units/kg, i.v.), low dose ulinastatin group (8000 units/kg, i.v.), and control group (normal saline). The characteristics of included studies were summarized in Table 1. 3.3. Definition of POCD Xu and colleagues assessed the neuropsychological state of enrolled patients using a brief battery of neuropsychological test. A postoperative deficit in any test was defined by a decline of 20% or more from the preoperative value of that test; any patient demonstrating a deficit in 2 or more tests was considered as having POCD [29]. Ge and co-workers diagnosed POCD according to the same diagnostic criteria. The other three studies determined the incidence of POCD using MMSE scale. POCD was defined if the difference between post- and preoperative MMSE score was equal or larger than the standard deviation of preoperative MMSE scores [30]. 3.4. Risk of Bias Assessment All the studies included the suggested randomization; four studies reported the method of random sequence generation. Only one study reported procedure of allocation concealment and double-blinding: study drugs were prepared by the hospital pharmacy in identical containers marked with the name of the project, the investigator's name, and consecutive numbers. In this study, patients and investigators were blinded to the infusion. Two studies reported number of drop-outs; the reasons for withdraw in different groups were similar. When it comes to selective reporting bias, all studies were judged low risk of bias because we only included studies which reported incidence of POCD, MMSE score. The judgement of risk of bias was presented in Figures 2 and 3. 3.5. Incidence of POCD All the included studies measured incidence of POCD as outcome assessment. As there was no obvious heterogeneity, fixed-effect model was utilized for statistical analysis. Subgroup analysis was conducted according to the different timing of neuropsychological tests. The meta-analysis showed no statistical difference of incidence of POCD between ulinastatin and control groups on postoperative day 1 (RR 0.71, 95% CI 0.28, 1.81; P = 0.48); but ulinastatin could significantly decrease the incidence of POCD on postoperative day 3 (RR 0.12, 95% CI 0.05, 0.31; I 2 = 0; P < 0.0001) and day 7 (RR 0.37, 95% CI 0.23, 0.61; I 2 = 11%; P < 0.0001) when compared with control treatment (Figure 4). 3.6. MMSE Score Two studies [25, 27] measured MMSE score at different time point after surgeries. Subgroup analysis based on different timing of MMSE score measurement was conducted. The combination of data showed that ulinastatin could further improve the MMSE score on day 1 (RR 1.80, 95% CI 1.43, 2.17; P < 0.00001), day 3 (RR 1.86, 95% CI 1.54, 2.18; P < 0.00001), and day 7 (RR 1.10, 95% CI 0.77, 1.43; P < 0.00001) after operation (Figure 5). 3.7. Secondary Outcomes In addition to the incidence of POCD and MMSE score, IL-6, TNF-α, CRP, and S100β concentrations were also measured by our included studies. Baseline similarities in ulinastatin and control groups were described in all the studies. Except for TNF-α, all these aforementioned laboratory indicators increased significantly within the first 24 postoperative hours. On day 3 after operation, there seemed to be no significant differences of IL-6, TNF-α, and S100β. IL-6 and S100β concentrations were significantly lower in ulinastatin groups than control groups from the end of operation to postoperative day 2; there were no significant differences of IL-6 and S100β on postoperative day 3. Regarding the changes of CRP and TNF-α, studies yielded conflicting results. The descriptions of secondary outcomes were listed in Table 2 in detail. 3.8. Other Outcomes Two studies [24, 26] reported postoperative complications, such as nausea and vomiting. In the study conducted by Xu et al., the incidence of nausea and vomiting, arrhythmia, delirium, infection, allergic reaction, wound dehiscence, and myocardial ischemia was lower in ulinastatin group than control group, but the differences were not statistically significant. Ge and coworkers reported significantly lower incidence of deep vein thrombosis in ulinastatin group than control on the third day after hip joint replacement. 3.9. Publication Bias Publication bias was explored via funnel plots (Figures 6 and 7). Both funnel plots presented asymmetry, indicating publication bias. 4. Discussion To the best of our knowledge, this is the first systematic review and meta-analysis to assess the clinical effect of ulinastatin in the treatment of patients with early POCD. Five RCTs involving a total of 461 patients were identified by our current work. Based on the findings of this study, ulinastatin administration was effective in reducing the incidence of POCD and improving MMSE score. Compared with control treatment, incidences of all types of postoperative complications were lower in ulinastatin groups. However, the effect of ulinastatin in reducing proinflammatory cytokines (IL-6 and TNF-α), CRP, and S100β remained to be further elucidated. POCD refers to a deterioration in cognition noted to occur after surgery and anesthesia. It is a subtle form of cognitive decline that can occur after surgery and affect cognitive performance, especially in the elderly, although few young patients have been described as cognitive declines as well [31]. Surgical trauma can increase level of inflammatory cytokine IL-6, TNF-α, and level of CRP. These factors with a wide bioactivity can cross blood brain barrier, promote brain cell permeability, and cause an inflammatory reaction in the central nervous system, thereby affecting the functioning of synaptic connections, resulting in damage in cognitive function. Our included studies observed a significant increase of IL-6 after operation, and the concentration of IL-6 returned to normal on the third day after operation. The incidence of POCD and concentration of IL-6 were significantly lower than control within the first two days after operation; this suggested that ulinastatin may attenuate POCD by inhibiting the release of IL-6. Regarding the changes of TNF-α and CRP, the results among studies were inconsistent. Ge and colleagues [28] found that TNF-α were increased on postoperation day 1; both high dose and low dose administration of ulinastatin could inhibit the release of TNF-α. Xu et al. [26] did not find a postoperative increase of TNF-α and CRP; they attributed this to the populations differences. S100β is a calcium-binding protein, mainly in stellate cells and Schwann cells, as glial marker protein, brain-specific proteins. After surgical operation, increased serum S100β values were reported to correlate with poor neurological outcome [26]. S100β was usually elevated in the blood and cerebrospinal fluid following nervous system damage due to a functional disturbance of membrane integrity and/or increased permeability of the blood brain barrier. Serum S100β concentrations were reported to be significantly increased within two days after operations, and it fell back to normal on the third day following operations. Xu et al. and Kang et al. [25, 26] found that ulinastatin could further reduce the concentration of S100β in serum than control in the first two days after operation. However, Ge and coworkers [28] found that S100β was significantly lower in ulinastatin group only on the postoperative 6 hours; no statistically significant difference of S100β was observed on other time-points studied. The biological half-life of S100β is approximately 30 minutes. Therefore, persistently increased levels of S100β in the blood indicate continuous release of this protein from damaged tissue. The discrepancies of postoperative S100β among our included studies can be explained by inconsistent method of ulinastatin administration and duration of operations. MMSE is a 30-point scale that measures global cognitive function, with higher scores indicating better function, with scores <24 suggestive of cognitive impairment. Two studies measured MMSE score on postoperative day 1, day 3, and day 7. The findings suggested that ulinastatin was effective in improving the postoperative MMSE scores. In a word, ulinastatin could attenuate POCD, but whether ulinastatin played a pivotal part in reducing inflammatory mediators remained unclear, and the underlying mechanisms need to be validated by additional studies. There were several limitations in our systematic review. Firstly, only five RCTs involving 461 patients that underwent surgical operations were selected by us; the sample-sizes of these studies were relatively small. Four studies did not mention blinding procedures, which might lead to exaggeration of conclusions drawn by these trials. Secondly, the MMSE remains one of the most commonly used instruments to assess cognitive outcomes, but it has been criticized for some shortcomings such as being less sensitive to milder cognitive impairments in older adults. Thirdly, all the neuropsychological tests and laboratory indicators were detected within a week after operations; the long-term benefits of ulinastatin on cognitive impairments could not be confirmed. Only one study reported the incidence of POCD one month after operation (high dose ulinastatin 12.9%, low dose ulinastatin 16.7%, and control 28.1%). Future studies with longer follow-ups to validate the beneficial effect of ulinastatin are needed. Besides, only two included studies reported duration of operations; considering that some inflammatory mediators were directly associated with duration of operations, the results and conclusions may be sequentially confounded. Lastly, the funnel plots indicated obvious publication bias; this could be explained by the fact that all these trials were conducted in China and all these articles were published in Chinese academic journals; these works have only been performed in a Chinese population, clinical studies within western culture to evaluate the effect of ulinastatin are encouraged. 5. Conclusion In summary, ulinastatin administration showed remarkable effect in reducing the incidence of POCD and improving MMSE score. Ulinastatin could significantly reduce the concentrations of IL-6 and S100β from the end of operation to postoperative day 2; results regarding the changes in CRP and TNF-α remained debatable. Further studies are need not only to determine the potential benefit but to understand the mechanisms involved in the mitigation of POCD by ulinastatin. Acknowledgments The present study was supported by the National High Technology Research and Development Program of China (863 Program) [no. 2012AA02A612] and the National Natural Science Foundation of China [no. 81472082; no. 81171696; no. 81371915; no. 81572094]. Competing Interests The authors have declared no conflict of interests. Authors' Contributions Zheng-tao Lv and Jun-ming Huang contributed equally to this study. Figure 1 Flowchart of literature search and study selection. Figure 2 Risk of bias summary: review authors' judgements about each risk of bias item for each included study. Figure 3 Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies. Figure 4 Forest plot of ulinastatin versus control: incidence of POCD. Figure 5 Forest plot of ulinastatin versus control: MMSE score. Figure 6 Funnel plot of ulinastatin versus control: incidence of POCD. Figure 7 Funnel plot of ulinastatin versus control: MMSE score. Table 1 Characteristics of included studies. Study Study design Population Intervention and control Outcomes Xu et al., 2013 China [26] Prospective, double-blind, 2-arm RCT Abdominal surgery under intravenous general anesthesia; U: 40 patients, 75.6 ± 7.2 years; C: 40 patients, 74.1 ± 8.1 years U: ulinastatin 10000 units/kg diluted in normal saline to a volume of 100 mL (i.v.), over a period of 30 min before surgical incision and 5000 units/kg after surgery on days 1–3; C: 100 mL normal saline (i.v.) Incidence of POCD; IL-6, TNF-α, CRP, S100β Ge et al., 2011 China [24] 2-arm RCT Hip joint replacement under combined spinal-epidural anesthesia; U: 80 patients, 72.8 ± 7.5 years; C: 80 patients, 75.0 ± 7.1 years U: ulinastatin 10000 units/kg diluted in normal saline to a volume of 50 mL (i.v.), over a period of 30 min before surgical incision and 5000 units/kg after surgery on days 1–3; C: 50 mL normal saline (i.v.) Incidence of POCD Kang et al., 2010 China [25] 2-arm RCT Hip joint replacement under combined spinal-epidural anesthesia; U: 40 patients, 75.0 ± 7.81 years; C: 40 patients, 72.8 ± 7.25 years U: ulinastatin 10000 units/kg diluted in normal saline to a volume of 50 mL (i.v.), over a period of 30 min before surgical incision and 5000 units/kg after surgery on days 1–3; C: 50 mL normal saline (i.v.) Incidence of POCD, MMSE score; S100β Shan et al., 2015 China [27] 2-arm RCT Hip fracture under combined spinal-epidural anesthesia; U: 21 patients, 78 ± 2 years; C: 27 patients, 75 ± 1 years U: ulinastatin 5000 units/kg diluted in normal saline to a volume of 100 mL (i.v.), before surgical incision and 5000 units/kg immediately after surgery; C: 100 mL normal saline (i.v.) Incidence of POCD, MMSE score; CRP Ge et al., 2015 China [28] 3-arm RCT Coronary artery bypass grafting under intravenous general anesthesia; U1: 31 patients, 69.1 ± 4.8 years; U2: 30 patients, 68.9 ± 4.7 years; C: 32 patients, 67.7 ± 5.4 years U1: ulinastatin 16000 units/kg diluted in normal saline to a volume of 60 mL (i.v.) before anesthesia induction; U2: ulinastatin 8000 units/kg diluted in normal saline to a volume of 60 mL (i.v.) before anesthesia induction; C: 60 mL normal saline (i.v.) Incidence of POCD; IL-6, TNF-α, CRP, S100β Note: POCD: postoperative cognitive dysfunction; U: ulinastatin group; C: control group; i.v.: intravenously; RCT: randomized controlled trials. Table 2 Secondary outcomes reported by included studies. Study IL-6 (pg/mL) TNF-α (pg/mL) CRP (mg/L) S100β (ug/L) U C U C U C U C Xu et al., 2013 [26] 7.1 ± 0.1a 8.2 ± 0.2a 870 ± 490a 890 ± 590a 6.8 ± 3.2a 7.4 ± 3.4a 0.039 ± 0.012a 0.040 ± 0.011a 55.2 ± 5.1b∗# 98.3 ± 4.4b∗ 1210 ± 450b 1380 ± 860b 108.3 ± 4.5b∗ 109.8 ± 5.3b∗ 0.097 ± 0.014b∗# 0.129 ± 0.034b∗ 46.2 ± 4.8e∗# 72.2 ± 3.8e∗ 1070 ± 540e 1190 ± 750e 78.6 ± 3.6e∗ 85.7 ± 5.1e∗ 0.086 ± 0.016e∗# 0.141 ± 0.029e∗ 21.4 ± 7.3f∗# 45.3 ± 6.3f∗ 950 ± 510f 960 ± 460f 54.6 ± 5.1f∗ 65.3 ± 3.6f∗ 0.057 ± 0.019f∗# 0.089 ± 0.038f∗ 8.2 ± 0.3g 9.3 ± 0.4g 880 ± 490g 890 ± 510g 7.4 ± 4.1g 8.9 ± 4.3g 0.042 ± 0.017g 0.047 ± 0.018g Kang et al., 2010 [25] — — — — — — 0.040 ± 0.013a 0.041 ± 0.012a 0.095 ± 0.021b∗# 0.125 ± 0.031b∗ 0.116 ± 0.017c∗# 0.178 ± 0.036c∗ 0.087 ± 0.019e∗# 0.142 ± 0.038e∗ 0.043 ± 0.012g 0.048 ± 0.015g Shan et al., 2015 [27] — — — — 26 ± 5a 32 ± 5a — — 64 ± 10g∗# 124 ± 7g∗ Ge et al., 2015 (1) [28] 36.10 ± 5.48a 34.92 ± 4.68a 29.67 ± 4.17a 30.84 ± 3.98a — — 250 ± 30a 250 ± 40a 49.66 ± 5.89b∗# 62.90 ± 7.23b∗ 37.93 ± 6.80b∗# 44.09 ± 11.35b∗ 770 ± 180b∗ 810 ± 230b∗ 65.14 ± 10.86d∗# 90.63 ± 12.06d∗ 51.92 ± 6.39d∗# 71.26 ± 11.33d∗ 620 ± 160d∗# 770 ± 210d∗ 48.03 ± 6.01e∗# 61.20 ± 6.17e∗ 62.55 ± 12.07e∗# 80.98 ± 15.33e∗ 430 ± 90e∗ 470 ± 100e∗ Ge et al., 2015 (2) [28] 34.67 ± 4.77a 34.92 ± 4.68a 30.24 ± 4.05a 30.84 ± 3.98a — — 240 ± 50a 250 ± 40a 48.56 ± 6.25b∗# 62.90 ± 7.23b∗ 38.17 ± 5.70b∗# 44.09 ± 11.35b∗ 750 ± 170b∗ 810 ± 230b∗ 68.16 ± 9.05d∗# 90.63 ± 12.06d∗ 50.42 ± 3.27d∗# 71.26 ± 11.33d∗ 590 ± 180d∗# 770 ± 210d∗ 47.02 ± 6.73e∗# 61.20 ± 6.17e∗ 61.30 ± 11.91e∗# 80.98 ± 15.33e∗ 440 ± 100e∗ 470 ± 100e∗ Note: U: ulinastatin group; C: control group; apreoperative; bat the end of operation; cthree hours after operation; dsix hours after operation; eone day after operation; ftwo days after operation; gthree days after operation; ∗ P < 0.05 from preoperation in group U and group C (statistically significant); # P < 0.05 from group C (statistically significant). Ge et al., 2015 (1): high-dose ulinastatin (16000 U/kg); Ge et al., 2015 (2): low-dose ulinastatin (8000 U/kg). ==== Refs 1 Deiner S. Silverstein J. H. Postoperative delirium and cognitive dysfunction British Journal of Anaesthesia 2009 103 supplement 1 i41 i46 10.1093/bja/aep291 2-s2.0-77449118492 20007989 2 Tsai T. L. Sands L. P. Leung J. M. An update on postoperative cognitive dysfunction Advances in Anesthesia 2010 28 269 284 10.1016/j.aan.2010.09.003 2-s2.0-78649383955 21151735 3 van Dijk D. Keizer A. M. A. Diephuis J. C. Durand C. Vos L. J. Hijman R. Neurocognitive dysfunction after coronary artery bypass surgery: a systematic review Journal of Thoracic and Cardiovascular Surgery 2000 120 4 632 639 10.1067/mtc.2000.108901 2-s2.0-0033770945 11003741 4 Monk T. G. Price C. C. Postoperative cognitive disorders Current Opinion in Critical Care 2011 17 4 376 381 10.1097/MCC.0b013e328348bece 2-s2.0-79960835349 21716111 5 Terrando N. Monaco C. Ma D. Foxwell B. M. J. Feldmannc M. Maze M. Tumor necrosis factor-a triggers a cytokine cascade yielding postoperative cognitive decline Proceedings of the National Academy of Sciences of the United States of America 2010 107 47 20518 20522 10.1073/pnas.1014557107 2-s2.0-78650580820 21041647 6 Terrando N. Eriksson L. I. Ryu J. K. Resolving postoperative neuroinflammation and cognitive decline Annals of Neurology 2011 70 6 986 995 10.1002/ana.22664 2-s2.0-84255194757 22190370 7 Su X. Feng X. Terrando N. Dysfunction of inflammation-resolving pathways is associated with exaggerated postoperative cognitive decline in a rat model of the metabolic syndrome Molecular Medicine 2012 18 12 1481 1490 10.2119/molmed.2012.00351 2-s2.0-84873506932 23296426 8 Degos V. Vacas S. Han Z. Depletion of bone marrow-derived macrophages perturbs the innate immune response to surgery and reduces postoperative memory dysfunction Anesthesiology 2013 118 3 527 536 10.1097/ALN.0b013e3182834d94 2-s2.0-84874629777 23426204 9 Zhang Y.-H. Guo X.-H. Zhang Q.-M. Yan G.-T. Wang T.-L. Serum CRP and urinary trypsin inhibitor implicate postoperative cognitive dysfunction especially in elderly patients International Journal of Neuroscience 2015 125 7 501 506 10.3109/00207454.2014.949341 2-s2.0-84938697146 25105909 10 Stalnacke B. M. Sojka P. S100B for diagnosis and prognosis of sequelae following minor head injury. Contradictory results according to studies Lakartidningen 2008 105 24-25 1840 1845 18619007 11 Inoue K.-I. Takano H. Shimada A. Urinary trypsin inhibitor protects against systemic inflammation induced by lipopolysaccharide Molecular Pharmacology 2005 67 3 673 680 10.1124/mol.104.005967 2-s2.0-14944350369 15576631 12 Inoue K.-I. Takano H. Urinary trypsin inhibitor as a therapeutic option for endotoxin-related inflammatory disorders Expert Opinion on Investigational Drugs 2010 19 4 513 520 10.1517/13543781003649533 2-s2.0-77952492404 20367192 13 Chen C.-C. Wang S.-S. Lee F.-Y. Action of antiproteases on the inflammatory response in acute pancreatitis Journal of the Pancreas 2007 8 4, supplement 488 494 2-s2.0-34547929954 17625305 14 Sun R. Li Y. Chen W. Zhang F. Li T. Total ginsenosides synergize with ulinastatin against septic acute lung injury and acute respir atory distress syndrome International Journal of Clinical and Experimental Pathology 2015 8 6 7385 7390 2-s2.0-84938630484 26261640 15 Atal S. S. Atal S. Ulinastatin—a newer potential therapeutic option for multiple organ dysfunction syndrome Journal of Basic and Clinical Physiology and Pharmacology 2016 27 2 10.1515/jbcpp-2015-0003 16 Li Y. Zhao L. Fu H. Wu Y. Wang T. Ulinastatin suppresses lipopolysaccharide induced neuro-inflammation through the downregulation of nuclear factor-κ B in SD rat hippocampal astrocyte Biochemical and Biophysical Research Communications 2015 458 4 763 770 10.1016/j.bbrc.2015.01.155 2-s2.0-84925465418 25681771 17 Kim S. J. Yoo K. Y. Jeong C. W. Urinary trypsin inhibitors afford cardioprotective effects through activation of PI3K-Akt and ERK signal transduction and inhibition of p38 MAPK and JNK Cardiology 2009 114 4 264 270 10.1159/000234321 2-s2.0-68549101881 19684396 18 Li W. Qiu X. Jiang H. Zhi Y. Fu J. Liu J. Ulinastatin inhibits the inflammation of LPS-induced acute lung injury in mice via regulation of AMPK/NF-κ B pathway International Immunopharmacology 2015 29 2 560 567 10.1016/j.intimp.2015.09.028 2-s2.0-84948457052 26481965 19 Zhang X. Liu F. Liu H. Urinary trypsin inhibitor attenuates lipopolysaccharide-induced acute lung injury by blocking the activation of p38 mitogen-activated protein kinase Inflammation Research 2011 60 6 569 575 10.1007/s00011-010-0305-2 2-s2.0-79956224306 21246393 20 Shu Y. Yang Y. Qiu W. Neuroprotection by ulinastatin in experimental autoimmune encephalomyelitis Neurochemical Research 2011 36 11 1969 1977 10.1007/s11064-011-0520-4 2-s2.0-80055021848 21667278 21 Koga Y. Fujita M. Tsuruta R. Urinary trypsin inhibitor suppresses excessive superoxide anion radical generation in blood, oxidative stress, early inflammation, and endothelial injury in forebrain ischemia/reperfusion rats Neurological Research 2010 32 9 925 932 10.1179/016164110X12645013515133 2-s2.0-77957583751 20223106 22 Moher D. Liberati A. Tetzlaff J. Altman D. G. Group P. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement PLoS Medicine 2009 6 7 e1000097 10.1371/journal.pmed.1000097 23 Higgins J. P. T. Green S. Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0 2011 The Cochrane Collaboration http://handbook.cochrane.org/ 24 Ge Y.-Y. Xu Y. Cheng J.-Q. Chen J.-M. Xu R.-M. Jiang Y.-D. Case-control study on the effect of ulinastatin on postoperative complications in elderly patients undergoing hip joint replacement China Journal of Orthopaedics and Traumatology 2011 24 6 459 462 2-s2.0-79961177490 21786544 25 Kang Y.-M. Ge Y.-Y. Cheng J.-Q. Effect of ulinastatin on post-operative cognition disorders in elderly patients undergoing hip joint replacement China Journal of Orthopaedics and Traumatology 2010 23 5 355 358 2-s2.0-77954916650 20575290 26 Xu L.-L. Hu Z.-Y. Shen J.-J. A preliminary study of the effects of ulinastatin on early postoperative cognition function in patients undergoing abdominal surgery Neuroscience Letters 2013 541 15 19 10.1016/j.neulet.2013.02.008 2-s2.0-84876411360 23422155 27 Shan X. Zhang X. Li X. Xu G. Li Z. Effect of ulinastatin on postoperative cognitive function in the elderly with fracture National Medical Journal of China 2015 95 20 1586 1589 10.3760/cma.j.issn.0376-2491.2015.20.009 2-s2.0-84940475376 26463607 28 Ge Y.-L. Shi H.-W. Zhu W.-F. Effects of ulinastatin on cognitive function in patients with coronary artery bypass grafting Zhejiang Da Xue Xue Bao Yi Xue Ban 2015 44 5 532 538 26713528 29 Newman S. P. Analysis and interpretation of neuropsychologic tests in cardiac surgery The Annals of Thoracic Surgery 1995 59 5 1351 1355 10.1016/0003-4975(95)00215-7 2-s2.0-0028933052 7733767 30 Boos G. L. Soares L. F. De Oliveira Filho G. R. Postoperative cognitive dysfunction: prevalence and associated factors Revista Brasileira de Anestesiologia 2005 55 5 517 524 2-s2.0-25444523775 19468642 31 Johnson T. Monk T. Rasmussen L. S. Postoperative cognitive dysfunction in middle-aged patients Anesthesiology 2002 96 6 1351 1357 10.1097/00000542-200206000-00014 2-s2.0-18444363330 12170047

PMC005xxxxxx/PMC5002305.txt

==== Front Stem Cells IntStem Cells IntSCIStem Cells International1687-966X1687-9678Hindawi Publishing Corporation 10.1155/2016/3604203Review ArticleRecent Advances and Future Direction in Lyophilisation and Desiccation of Mesenchymal Stem Cells Bissoyi Akalabya 1 http://orcid.org/0000-0001-8735-479XKumar Awanish 2 http://orcid.org/0000-0002-9427-5739Rizvanov Albert A. 3 * http://orcid.org/0000-0003-4686-2303Nesmelov Alexander 3 Gusev Oleg 3 4 Patra Pradeep Kumar 5 http://orcid.org/0000-0003-4738-8324Bit Arindam 1 * 1Department of Biomedical Engineering, National Institute of Technology, Raipur 492010, India2Department of Biotechnology, National Institute of Technology, Raipur 492010, India3Institute of Fundamental Medicine and Biology, Kazan Federal University, Kazan 420000, Russia4Preventive Medicine & Diagnosis Innovation Program (PMI), Division of Genomic Technologies, RIKEN, Yokohama 2300045, Japan5Department of Biochemistry, Pt. JNM Medical College, Raipur 492001, India*Albert A. Rizvanov: albert.rizvanov@kpfu.ru and *Arindam Bit: arinbit.bme@nitrr.ac.inAcademic Editor: Andrea Ballini 2016 14 8 2016 2016 360420318 4 2016 3 7 2016 Copyright © 2016 Akalabya Bissoyi et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Mesenchymal Stem Cells (MSCs) are a promising mammalian cell type as they can be used for the reconstruction of human tissues and organs. MSCs are shown to form bone, cartilage, fat, and muscle-like cells under specific cultivation conditions. Current technology of MSCs cryopreservation has significant disadvantages. Alternative technologies of mammalian cells preservation through lyophilisation or desiccation (air-drying) are among the upcoming domains of investigation in the field of cryobiology. Different protectants and their combinations were studied in this context. Loading of the protectant in the live cell can be a challenging issue but recent studies have shown encouraging results. This paper deals with a review of the protectants, methods of their delivery, and physical boundary conditions adopted for the desiccation and lyophilisation of mammalian cells, including MSCs. A hybrid technique combining both methods is also proposed as a promising way of MSCs dry preservation. Russian Science Foundation grant for international group14-44-00022National grant of Department of Science and TechnologyYSS/2015/000618 ==== Body 1. Introduction Mesenchymal Stem Cells (MSCs) are multipotent stromal cells able to differentiate into different cell types, including chondrocytes, adipocytes, and osteoblasts [1–3]. MSCs have potential utility in cell therapy and regenerative medicine, with applications relating to tissue engineering and as vehicles for gene therapy. Advantages of their usage include high plasticity, regenerative, and immunosuppressive properties and tropisms toward inflamed, hypoxic, and cancerous sites [2]. Additionally, the typical usage of MSCs for the patient was derived from autologous transplantation. It is biologically safe and free of any ethical issues associated with the source of cell. The usage of MSCs requires their preservation in a way allowing them to be rapidly available for an application. The commonly used approach for MSCs storage is cryopreservation using liquid nitrogen [4]. This technology can ensure a high viability of stored cells but the storage, transfer, and prevention of contamination are complicated and relatively expensive. Lyophilisation (freeze-drying) and desiccation (air-drying) combined with the application of specific protective compounds emerge as a promising strategies for the storage of mammalian cells, including MSCs. As these technologies allow the storage of dried samples at ambient temperature, they can greatly simplify storage and distribution of samples, thereby decreasing the preservation cost. However, the usage of these technologies requires a revision of existing protocols in order to enhance the cell viability rates. For this purpose, we need an improved understanding of processes underlying the cell survival during water loss and in the dry state. This paper discusses the fundamental principles, mechanisms, and advantages of the lyophilisation and desiccation usage for MSCs preservation. It also briefly describes an application and possible crosstalk of these technologies, usage of different lyoprotectants, and physiological factors involved in the cell response. 2. Lyophilisation and Desiccation Lyophilisation is a process of drying of the frozen sample when frozen water is removed in two steps: primary drying (sublimation), followed by secondary drying (desorption). Since lyophilisation is performed at low temperature, it is used to stabilize and prolong shelf life of thermolabile products and those products which are otherwise unstable in aqueous state and hence need to be dried. The principle involved is sublimation of water at temperature and pressure below its triple point, that is, 611 Pascal and 0.0098 degree Celsius. Sublimation takes place after the sample is frozen by supplying heat through conduction or radiation or both. The driving force behind the sublimation or removal of water is the water vapour concentration gradient between the drying front and the condenser [5]. Similarly to the cryopreservation, different poly-hydroxy compounds such as sugars, polyalcohols, and their derivatives can be used in lyophilisation to protect the products that are sensitive to occurring dehydration. These compounds are called lyoprotectants and are reviewed below. The diagram summarizing the processes beyond the cell lyophilisation in the presence of nonreducing disaccharide trehalose is depicted on (Figure 1). Lyophilisation was reported to ensure the viability up to 70% of MSCs without addition of protectants [6]. Desiccation is a nature-driven process, widely used by plants and lower animals as a way to survive the water deficiency. In response to desiccation, such organisms switch to the specific state of suspended animation, anhydrobiosis, allowing them to revive when water returns [7]. Organisms capable of anhydrobiosis were found among a variety of taxa, including bacteria, yeast, invertebrates, and plants [8]. The common process underlying anhydrobiosis is an accumulation of sugars such as trehalose or sucrose [7]. In the dry state these sugars form a highly viscous glass-like medium that has been shown to be necessary for stabilization of cell components [9]. In addition, anhydrobiotes were shown to utilize a variety of protective proteins such as LEA (Late Embryogenesis Abandoned) and heat-shock proteins (HSPs), proteins of antioxidant system, or transport proteins [10]. Successful short-term storage of human fibroblast was achieved [11], but successful MSCs desiccation is not reported yet. 3. Protectants in Lyophilisation and Desiccation A variety of protectants were tried for the dry preservation of mammalian cells. Among them, nonreducing disaccharides trehalose and sucrose are the most popular because they are used by almost every anhydrobiotic animal. There are two main hypotheses explaining the protective effect of sugars in desiccation: (i) water-replacement hypothesis and (ii) vitrification hypothesis [9]. The water-replacement hypothesis proposes that sugars replace water interacting with polar or charged groups of biomolecules via direct hydrogen bonds. Thereby sugars stabilize the native structure of phospholipid membranes or proteins in the absence of water. The vitrification hypothesis assumes that sugars ensure a glass-like state of both cytoplasm and extracellular medium, preventing denaturation or mechanical disruption of membranes and cellular components. Proteins, such as LEA, can also play an important role in the formation of vitrified medium during desiccation [9]. Protectants naturally used by anhydrobiotic animals were successfully used for lyophilisation. Nonreducing sugars were preferred as protectants for desiccated objects since they are less reactive than reducing monosaccharides or disaccharides like maltose. Maltose has proven to be an effective lyoprotectant for DNA [12], but its higher reactivity should be taken into account for the dry storage. Similarly, nonreducing disaccharide sucrose was shown to be more susceptible to hydrolysis into reducing monosaccharides than trehalose [7]. This indicates that trehalose might be a preferable over sucrose as a protectant in applications requiring the storage of dry samples. Interestingly, widely used cryoprotectants such as dimethyl sulfoxide (DMSO) or glycerol did not find a usage as protectants during desiccation, due to fundamental differences between these modes of preservation. 4. Advantages of Trehalose over Other Protectants Trehalose is widely assumed as the protectant superior to others in the desiccation context. The hydration radius of trehalose is 2.5 times greater than that of sucrose, which implies that 2.5 times the concentration of sucrose is required to provide the same protein protection properties as trehalose [13]. Trehalose interacts more strongly with both water and proteins [14] and is able to displace water molecules which are bound to carbonyls at the phospholipid bilayer of cell membranes unlike to sucrose [15]. Trehalose has higher glass transition temperature (T g) than sucrose which means that when a small amount of water is added, T g for sucrose goes below the storage temperature and hence the sample dried with sucrose degrades more rapidly as compared to trehalose dried product (whose T g remains above storage temperature). In addition, antioxidant effect of trehalose was shown [16]. Hence, due to all these properties of trehalose, it can be considered to be superior to the sucrose as an effective protectant for the dry storage (Table 1). 5. Synergistic Enhancers Different combinations of protective agents were tried for the desiccation of cells, since it is clear from existing studies that anhydrobiotic animals typically use the interplay of several mechanisms. Similarly to the case of trehalose, these combinations were mostly inspired by their naturally existing prototypes in different anhydrobiotic organisms. Study shows that more than 40 mM trehalose could effectively maintain the viability of MSC more than 92% for 7 days at 4°C storage temperature [17]. Expression of small heat-shock proteins (HSPs) is tightly linked to the stress tolerance, including the desiccation tolerance in African chironomid Polypedilum vanderplanki [18]. Probably, chaperone activity of these proteins can mitigate deleterious effects of water loss. Transfection of embryonic cells of human kidney with the gene Hsp26 protein coupled with trehalose application exhibits a sharp rise in survival after desiccation in comparison to trehalose alone [19]. This improved cell survival clearly demonstrates a useful synergy of trehalose and HSP for the recovery of cells undergoing desiccation and rehydration. Expression level of Hsp26 protein can be further increased in order to improve this method of cell drying [19]. Another promising way to improve desiccation tolerance of cells is the use of LEA proteins. LEA proteins stabilize vitrified sugar glasses which are important in the dried state [20] and protect desiccation-liable proteins from deactivation and aggregation [21]. Additionally, LEA proteins were shown to interact with biological membranes [22]; protein studied in this work was only afterwards identified as LEA protein [23]. Recently, LEA protein from African chironomid Polypedilum vanderplanki was reported to ensure the same level of protection as intracellular trehalose, as judged by membrane integrity after rehydration [24]. Since existing LEA proteins differ in their characteristics and localization [24, 25], their combination with each other or with other protectants is a promising way of engineering of cells tolerant to desiccation. Combination of trehalose with other molecules has also been tested for an increase in cell viability compared to usage of trehalose alone. A study involving desiccation of bovine sperm using sorbitol as a substitute for trehalose proves that in bovine sperm during dehydration it can be used as an excellent osmolyte and it also offers improvement in motility of sperm during desiccation. Sorbitol was found to be enhanced protection on permeating through the plasma membrane, whose mechanism is yet to understand properly [26]. Glycerol also has a significant role in combination with trehalose in achieving higher postrehydration membrane integrity in the desiccation of adipose tissue derived adult stem cells [27]. Recent studies involve the components of stress alleviating pathways such as chelators. For instance, sperm exhibited improved desiccation tolerance and enhanced survival rates when they were coupled with either sucrose or intracellular trehalose along with 1 mM Desferal to the desiccation buffer. Although sperm motility at postdrying stage did not show significant improvement, an increase in membrane integrity of sperm was observed [28]. Furthermore, recent study shows that sucrose, trehalose, and raffinose pretreatment improve postthaw viability of MSCs after cryopreservation [29]. Synergistic effect of trehalose and Arbutin (antioxidant) associated with an induction of HSPs in human MSCs was also shown [30]. These findings may highlight the correlation between the enhancer molecules and desiccation tolerance in mammalian cells which should be taken into account to derive a standard protocol. 6. Trehalose Delivery Presence of protectants such as trehalose on both sides of the cell membrane is important for adequate protection of intracellular components and cell membrane. However, the key protectants trehalose and sucrose are not membrane permeable. In typical case, MSCs do not have transporters for those protectants. The possible use of native membrane pores such as P2X7 [35] is very limited because they are specific to only a few cell types. Human MSCs are also capable of loading trehalose from the extracellular space by a fluid-phase endocytosis [34] but this process is inefficient and cell type dependent [37]. Thus the presence of trehalose in the MSCs should be achieved artificially. The number of proteins were shown to transport trehalose or other protective sugars into the cells, such as MalEFGK2, MAL11/AGT1, or TRET1 [38]. The advantages of TRET1 transporter are its high-capacity, specificity to the trehalose, neutral pH optimum, and expression from one gene [38]. Stable expression of TRET1 in Chinese hamster ovary (CHO) cells causes an astonishing increase by 170% of the cell viability after desiccation [38]. However, the use of this transporter may not be suitable for the MSCs preservation since this approach requires the genetic modification of the cells. The same reason may block the use of trehalose synthesis inside the cells despite of successful expression of trehalose producing genes in human fibroblast cells [32]. Direct methods of trehalose loading have also been used for the delivery of trehalose. Electropermeabilization was first tried as a trehalose delivery mechanism by Shirakashi et al. in mouse myeloma cells in isotonic trehalose substituted medium [31]. Microinjection was also successfully used for the trehalose loading in human oocytes, an attempt to improve their cryosurvival [39]. However, the poration of the cell membrane allows nonspecific transport of molecules and ions. Resulting alteration of transmembrane ionic balance may lead to significant cell damage. A novel technique of trehalose loading involves the use of basic amino acid rich cell penetrating peptides (CPPs), such as peptide KRKRWHW [39]. This peptide was designed to deliver trehalose into mammalian cells on the base of molecular simulations. Trehalose, as a cargo coupled with the KRKRWHW peptide through hydrogen bond and π–π bond, was successfully loaded into the Mouse Embryonic Fibroblasts (MEFs) [36]. This CPP is able to efficiently deliver trehalose into mammalian cells with low cytotoxicity even at high concentrations. Therefore, this CPP may be very helpful for improving the tolerance of cells during desiccation. Similarly, Toner and his group used a genetically engineered mutant of Staphylococcus aureus   α-hemolysin to create pores and to load trehalose into the cytoplasm of fibroblast cells [33]. In this study, trehalose was used as cryoprotectant. Recently, a successful trehalose loading into cells was achieved through modification of its molecule [32]. Intracellular concentration of trehalose was sufficient for applications in biopreservation while the impact on cellular viability and function was negligible. Different ways of trehalose delivery into mammalian cells are summarized in Table 2. 7. Role of Physical Factors and Cell Culture Condition The physical conditions of the macroenvironment and microenvironment of cells play a major role in the desiccation tolerance of desiccated cells. The beneficial aspect of this approach is an elimination of genetic engineering of the cells. A thorough study of the effects of various physical conditions should be carried out to make conclusions for optimization of favourable limits for all the physical parameters. Primarily, vacuum conditions improve the cell tolerance to desiccation in case of lyophilisation compared to air-drying. In case of MSCs, lyophilisation can ensure the cell viability up to 70% even without trehalose [6]. However, results of this study cannot be directly extrapolated to the cell storage because the cell viability was studied only at 2 h after lyophilisation. Electromagnetic cryopreservation is undergoing a lot of research as a way of cryopreservation of MSCs, since static and oscillating electric fields and magnetic fields influence the ice formation [40]. Dental pulp stem cells preserved in a programmed freezer using magnetic field show around 70% viability when recovered, suggesting that magnetic freezing may be an alternate method for MSC preservation [41]. We have studied also the effect of drying rate on cell viability. Cultured cells were dried in tissue culture plates in the airflow of a biosafety cabinet. The drastically increased cell-death was observed in an absence of trehalose (data not shown). Ceasing laminar airflow caused a reduction of the rate of desiccation and an increase of cell recovery. Interestingly, clearly different cell viability was observed among different wells and within different areas of the same well. A better cell survival was observed at the well periphery, whereas reverse occurred at the centre of the well (data not shown). In a different set of experiments we demonstrated that MSC-like cells from third molar tooth germs maintained their biological properties, including expression of MSC surface antigens CD29, CD73, CD90, CD105, and CD166, expression of pluripotency associated genes, proliferation, and differentiation ability [42]. Importantly, cryopreserved cells displayed neuroprotective effects in a model of neuroblastoma SH-SY5Y cells exposed to oxidative and chemotherapeutic stress conditions. Light causes deleterious effect on the cells during desiccation, probably because of the increase of the free radicals concentration. Fluorescence light was shown to generate free radicals within hamster of human cells, thereby facilitating oxidative DNA lesion and single-strand breaks. Therefore, cells desiccated and maintained at dark condition have greater viability than those exposed to the fluorescence light. In order to optimize the ability of cell culture to withstand desiccation, the effects of desiccation temperature and cell culture confluency were also investigated. Optimal desiccation tolerance was found at higher cell density. Cell aggregation decreases their tolerance to desiccation. The temperature of operation for cell desiccation greatly influences the viability of cells, and the optimal survival point was found to be slightly lower than room temperature (20°). Recovery of desiccated cells was sufficient at this optimum temperature only. Therefore, it can be summarized that vacuum encapsulates the cells during the process of desiccation thereby improving the cell viability. Also, it had been found that cells can survive without addition of carbohydrates (or polyols) if the slow process of desiccation is followed by vacuum storage. A consistent outcome of the procedure can be obtained by defining a desiccation process which preserves cellular structures. Even in this case, gradual loss of viability may be found because of destruction with time of the desiccated state due to release of free radicals [11]. 8. Application of Lyophilisation and Desiccation and Molecular Mechanism Involved Survival of cells in dry and desiccated states is few of the most fascinating phenomena of nature and is lot to explore. The ability of MSCs to withstand these conditions for sustaining high degree of viability could have applications for wide spectrum of biological sciences, including tissue engineering. From above discussion it can be concluded that the lyophilisation and desiccation are two competing formulators for long-term storage of MSC. Both these technologies have been quite successfully applied in a wide range of fields including tissue engineering. Preservation of MSCs through desiccation in the dry state is used to meet the growing demand for MSCs storage and transport in regenerative medicine. Desiccation provides significant advantages over lyophilisation because of simplicity and higher energy efficiency. However desiccation may cause cell stress such as changes in cell volume, osmotic pressure, shrinkage of membrane/cell organelles, changes in activity of enzyme, metabolism downregulation, increased salt concentrations, cell viscosity changes, and stress protein production [43]. Lyophilisation seems to be more safe technology in this respect since it helps to avoid damage maintaining sample frozen throughout drying. Lower cell damage and loss of activity and higher degree of dehydration are the principle advantages of lyophilisation. Lyophilisation should be used when complete rehydration needed and product are heat liable, unstable, or high valuable. However, lyophilisation may also cause problems due to changes in liquid phase, solubility, and water freezing [44]. Thus, both processes have their own pros and cons and different application. The potential of MSCs in regenerative medicine increases the importance of development of optimized protocols for their desiccation and lyophilisation. Uptake of protectants by MSCs is required for successful desiccation or lyophilisation. However, uptake of protectant by cells can cause cellular response specific to the cell type. Since HSPs are able to improve the cell viability and protect cells from apoptosis during freeze-drying [19], some MSCs cells may be more resistant because of their ability to produce higher HSP levels. Similarly, MSCs can differ in antioxidant response to the activation of stress sensing mechanisms under desiccation or lyophilisation. Since antioxidants have protective potential for a dry storage [30], this difference should also be taken into account. 9. Crosstalk between the Process of Lyophilisation and Desiccation Thermodynamic aspects of the above two processes of preservation of MSC are interestingly competitive. Each process can involve quantization of temperature and pressure for transition from one state to another in order to improve cellular viability. Figure 2 shows a typical thermodynamic diagram for fluidic matter (specifically extracellular and intracellular fluid). In the figure, black arrowed path indicates the normal process of lyophilisation which involves vitrification followed by creation of extensive vacuum at a temperature of −79 degrees centigrade. This process is expensive as well as time consuming. On the other hand, desiccation involves greater cellular mortality due to rapid dehydration of cellular masses by overshooting the base temperature of the cellular medium to a higher point. A brief description of the heat transfer accounted during the process of lyophilisation is described below. Bioheat equation for heat transfer in a system of suspended cells in ultrathin straw (UTS) undergoing for slow cooling or vitrification can be realised by zero-dimensional heat transfer equation and it is given as (1) 1r∂∂rλr∂T∂r+∂∂zλ∂T∂z+q˙=ρcp∂T∂t, where λ, c p, and q˙ refers to thermal conductivity, specific heat, and metabolic heat source, respectively. Latent heat thus released is considered as ineffective, because insignificant volume ratio of ice to the maximum crystallisable ice x is (1 × 10−3). Since the diameter of cellular suspension is much smaller than the length of UTS, therefore, heat transfer along the axis can be neglected. Additionally, thermal property within the UTS can be assumed to be uniform and temperature-independent. Thus, the above equation can be simplified as (2) 1r∂∂rr∂T∂r=1α∂T∂t, where α is the thermal diffusivity. Let the initial temperature of the suspension be T 0 and let the final temperature be T ∞. Let the dimensionless form of temperature be expressed as (3) θ∗=T−T∞T0−T∞. Therefore, the transient temperature distribution can be written as (4) θ∗=∑n=1∞Cnexp⁡−ζn2F0J0ζnrr0, where (5) F0=αtr02,Cn=2ζnJ1ζnJ02ζn+J12ζn, where J 0 and J 1 are Bessel functions and are positive roots of the transcendental equation defined by ζ n: (6) ζnJ1ζnJ0ζn=2hr0k, where heat transfer coefficient is denoted by “h” and often it evaluates the quality of the cryopreservation system. The production of volume fraction of ice (x) is governed by the equation (7) dxdt=κa1xTm−Texp⁡−QRT, where (8) κ=Lπλ2νTmrf,a1x=x2/31−x, where T m, Q, R, L, and r f refer to temperature at the end of the freezing process, activation energy, gas constant, latent heat, and radius of ice when x is equal to 1, respectively. The thickness of transition layer between liquid and solid (“mushy zone”) is λ, and n is the kinematic viscosity. In view of the above thermodynamics describing processes of lyophilisation and desiccation, it has been observed that viability of cell is a vital issue. It results in a requirement of crosstalk between the above two mentioned processes. An approach of crosstalk has also been explained by the three-point thermodynamic diagram. Path defined by red colour as (1→2→3→4→5) is the projected path for the process of desiccation followed by lyophilisation. Sample is at room temperature at point 1; then, it is mildly heated with a gradual decrease in pressure to reach a state of point 2. Point 2 is defined as step-down transition point: slope and intercepts of this point are optimized in order to achieve maximum pressure drop, with minimum increase in temperature and maximum dehydration of the cell. Thus maximum cellular viability can be maintained. Once this state is achieved, the cell mass will be allowed to freezing state while maintaining the temperature to be constant, and point 3 will be reached. Point 3 is a critical point of freezing at low pressure when crystalline state of fluid present in intracellular and extracellular regions will be converted to glassy state. Then, desiccation will be followed by freeze-drying to reach point 4. Since vitrification of cell is done at lower pressure, it might happen that vitrification temperature will be much higher than −80 degrees centigrade. Therefore, pressure drop required to achieve vapour state from freeze state is much higher, reducing the cost of the combined process. Last step is sample dehydration by rapid temperature increase (up to room temperature) at constant pressure to reach point 5. This entire path of crosstalk is able to ensure two vital aspects of preservation of cells: (i) enhancement of cellular viability and (ii) decreasing lyophilisation (freeze-drying) cost. The scheme of hypothesized processes in the cells undergoing desiccation-lyophilisation in accordance with proposed method is depicted in the following (Figure 3). Cell preservation using crosstalk technique explained in the previous section can be achieved in a controlled chamber. Rate of dehydration of cell can be monitored continuously by a ultrasonic transducer measuring the humidity in the chamber. Cell viability can be continuously monitored by a droplet based remote microscopic system and real-time processing of obtained images. Output of this image-processing system can be used to control temperature and pressure in the chamber. 10. Conclusion Enhanced measures of protection are required for successful anhydrobiotic engineering of MSCs since existing protocols are not able to ensure robust cell recovery. Although the lyophilisation and desiccation are found to be efficient in some cases, the limitations of individual methods impart certain rigidity of their implementation. At the same time, maintenance of the dried cells viability in long-term storage is another critical issue which needs to be addressed. Continuous monitoring and control of preservation process can ensure successful MSCs preservation, but optimization for all the physical parameters is also needed. Acknowledgments The authors are thankful to National Institute of Technology, Raipur (CG), India, for providing facility for this work. This study was supported by Russian Science Foundation grant for international group (14-44-00022) as well as National grant of Department of Science and Technology (YSS/2015/000618) New Delhi, India. Competing Interests The authors declare that there are no competing interests regarding the publication of this paper. Authors' Contributions All authors contributed equally in this paper. Akalabya Bissoyi and Awanish Kumar had made equal contribution to this work. Figure 1 Molecular mechanics involved in cell lyophilisation. Figure 2 Phase diagram for cells under lyophilisation and desiccation. Figure 3 Hypothesized physiological changes being observed in cells following red path for desiccation-lyophilisation phase. Table 1 Advantages of trehalose over sucrose as a lyoprotectant. Sucrose Trehalose Vulnerable to hydrolysis under acidic conditions Not prone to hydrolysis under acidic conditions Smaller hydration radius; thus large amount of sucrose is required for lyophilisation Hydration radius 2.5 times that of sucrose, thus requiring 2.5 times less trehalose for lyophilisation as compared to sucrose Does not protect cells and proteins from oxidative damage Protects cells and proteins from oxidative damage Less interaction with water and hence does not displace water molecules bound to carbonyls at the phospholipid bilayer of the cell membrane Interacts more strongly with water and thus able to displace water molecules bound to carbonyls at the phospholipid bilayer of cell membranes Lower transition temperature High transition temperature Table 2 Different methods to deliver trehalose into cells. Method Explanation Reference Electroporation Murine myeloma cells were loaded with trehalose by electroporation, then freeze-dried, and rehydrated [31] Genetically engineering the cells Human primary fibroblasts were transfected with otsA and otsB genes from E. coli, encoding for trehalose synthase [32] Genetically engineered pores Genetically engineered mutant of alpha-hemolysin from Staphylococcus aureus was used to create pores in the cellular membrane of 3T3 fibroblasts and human keratinocytes. Resulting nonselective pore is equipped with a metal-actuated switch that is sensitive to extracellular zinc concentrations, thus permitting controlled loading of trehalose [33] Fluid phase endocytosis Human MSCs were loaded by trehalose up to 30 mM internal concentration at usual cultivation conditions for 24 h at 37°C in the presence of MSC medium with 0–125 mM trehalose [34] Endogenous cell surface receptor TF-1 cells were permeabilized using an endogenous protein P2X7 [35] Cell penetrating peptides (CPPs) Trehalose was coupled with CPP and incubated with mouse embryonic fibroblast cells at usual cultivation conditions [36] ==== Refs 1 Caplan A. I. Bruder S. P. Mesenchymal stem cells: building blocks for molecular medicine in the 21st century Trends in Molecular Medicine 2001 7 6 259 264 10.1016/s1471-4914(01)02016-0 2-s2.0-0034975453 11378515 2 Park J. S. Suryaprakash S. Lao Y.-H. Leong K. W. Engineering mesenchymal stem cells for regenerative medicine and drug delivery Methods 2015 84 3 16 10.1016/j.ymeth.2015.03.002 2-s2.0-84935851134 25770356 3 Bissoyi A. Pramanik K. Role of the apoptosis pathway in cryopreservation-induced cell death in mesenchymal stem cells derived from umbilical cord blood Biopreservation and Biobanking 2014 12 4 246 254 10.1089/bio.2014.0005 2-s2.0-84927589643 25162461 4 Haack-Sørensen M. Kastrup J. Cryopreservation and revival of mesenchymal stromal cells Methods in Molecular Biology 2011 698 161 174 10.1007/978-1-60761-999-4_13 2-s2.0-79959778739 21431518 5 Nireesha G. R. Divya L. Sowmya C. Lyophilization/freeze drying—an review International Journal of novel Trends in Pharmaceutical Sciences 2013 3 4 87 98 6 Zhang S.-Z. Qian H. Wang Z. Preliminary study on the freeze-drying of human bone marrow-derived mesenchymal stem cells Journal of Zhejiang University: Science B 2010 11 11 889 894 10.1631/jzus.b1000184 2-s2.0-78649459193 21043058 7 Crowe J. H. Anhydrobiosis: an unsolved problem Plant, Cell and Environment 2014 37 7 1491 1493 10.1111/pce.12304 2-s2.0-84901746756 8 Keilin D. The leeuwenhoek lecture: the problem of anabiosis or latent life: history and current concept Proceedings of the Royal Society B: Biological Sciences 1959 150 939 149 191 10.1098/rspb.1959.0013 9 Sakurai M. Furuki T. Akao K.-I. Vitrification is essential for anhydrobiosis in an African chironomid, Polypedilum vanderplanki Proceedings of the National Academy of Sciences of the United States of America 2008 105 13 5093 5098 10.1073/pnas.0706197105 2-s2.0-42449157545 18362351 10 Gusev O. Suetsugu Y. Cornette R. Comparative genome sequencing reveals genomic signature of extreme desiccation tolerance in the anhydrobiotic midge Nature Communications 2014 5, article 4784 10.1038/ncomms5784 2-s2.0-84919651803 11 Puhlev I. Guo N. Brown D. R. Levine F. Desiccation tolerance in human cells Cryobiology 2001 42 3 207 217 10.1006/cryo.2001.2324 2-s2.0-0034796520 11578120 12 Quaak S. G. L. Haanen J. B. A. G. Beijnen J. H. Nuijen B. Naked plasmid DNA formulation: effect of different disaccharides on stability after lyophilisation AAPS PharmSciTech 2010 11 1 344 350 10.1208/s12249-010-9391-2 2-s2.0-77950918447 20204715 13 Sola-Penna M. Roberto Meyer-Fernandes J. Stabilization against thermal inactivation promoted by sugars on enzyme structure and function: why is trehalose more effective than other sugars? Archives of Biochemistry and Biophysics 1998 360 1 10 14 10.1006/abbi.1998.0906 2-s2.0-0032441482 9826423 14 Lerbret A. Bordat P. Affouard F. Hédoux A. Guinet Y. Descamps M. How do trehalose, maltose, and sucrose influence some structural and dynamical properties of lysozyme? Insight from molecular dynamics simulations Journal of Physical Chemistry B 2007 111 31 9410 9420 10.1021/jp071946z 2-s2.0-34548036050 15 Del C Luzardo M. Amalfa F. Nuñez A. M. Díaz S. Biondi De Lopez A. C. Disalvo E. A. Effect of trehalose and sucrose on the hydration and dipole potential of lipid bilayers Biophysical Journal 2000 78 5 2452 2458 10.1016/s0006-3495(00)76789-0 2-s2.0-0342368957 10777741 16 França M. B. Panek A. D. Eleutherio E. C. A. Oxidative stress and its effects during dehydration Comparative Biochemistry and Physiology—A Molecular and Integrative Physiology 2007 146 4 621 631 10.1016/j.cbpa.2006.02.030 2-s2.0-33847401984 17 Di G. Wang J. Liu M. Wu C.-T. Han Y. Duan H. Development and evaluation of a trehalose-contained solution formula to preserve hUC-MSCs at 4°C Journal of Cellular Physiology 2012 227 3 879 884 10.1002/jcp.23066 2-s2.0-84155192360 22170612 18 Gusev O. Cornette R. Kikawada T. Okuda T. Expression of heat shock protein-coding genes associated with anhydrobiosis in an African chironomid Polypedilum vanderplanki Cell Stress & Chaperones 2011 16 1 81 90 10.1007/s12192-010-0223-9 2-s2.0-79951671917 20809134 19 Ma X. Jamil K. MacRae T. H. A small stress protein acts synergistically with trehalose to confer desiccation tolerance on mammalian cells Cryobiology 2005 51 1 15 28 10.1016/j.cryobiol.2005.04.007 2-s2.0-23444443329 15963489 20 Hand S. C. Menze M. A. Toner M. Boswell L. Moore D. LEA proteins during water stress: not just for plants anymore Annual Review of Physiology 2011 73 1 115 134 10.1146/annurev-physiol-012110-142203 2-s2.0-79951782916 21 Hatanaka R. Hagiwara-Komoda Y. Furuki T. An abundant LEA protein in the anhydrobiotic midge, PvLEA4, acts as a molecular shield by limiting growth of aggregating protein particles Insect Biochemistry and Molecular Biology 2013 43 11 1055 1067 10.1016/j.ibmb.2013.08.004 2-s2.0-84884402606 23978448 22 Steponkus P. L. Uemura M. Joseph R. A. Gilmour S. J. Thomashow M. F. Mode of action of the COR15a gene on the freezing tolerance of Arabidopsis thaliana Proceedings of the National Academy of Sciences of the United States of America 1998 95 24 14570 14575 10.1073/pnas.95.24.14570 2-s2.0-0032564408 9826741 23 Wise M. J. LEAping to conclusions: a computational reanalysis of late embryogenesis abundant proteins and their possible roles BMC Bioinformatics 2003 4, article 52 10.1186/1471-2105-4-52 2-s2.0-2942549269 24 Li S. Chakraborty N. Borcar A. Menze M. A. Toner M. Hand S. C. Late embryogenesis abundant proteins protect human hepatoma cells during acute desiccation Proceedings of the National Academy of Sciences of the United States of America 2012 109 51 20859 20864 10.1073/pnas.1214893109 2-s2.0-84871374515 23185012 25 Hatanaka R. Gusev O. Cornette R. Diversity of the expression profiles of late embryogenesis abundant (LEA) protein encoding genes in the anhydrobiotic midge Polypedilum vanderplanki Planta 2015 242 2 451 459 10.1007/s00425-015-2284-6 2-s2.0-84937517151 25822798 26 Sitaula R. Fowler A. Toner M. Bhowmick S. A study of the effect of sorbitol on osmotic tolerance during partial desiccation of bovine sperm Cryobiology 2010 60 3 331 336 10.1016/j.cryobiol.2010.03.004 2-s2.0-77952615563 20233588 27 Mittal S. Devireddy R. V. Desiccation tolerance of adult stem cells in the presence of trehalose and glycerol The Open Biotechnology Journal 2008 2 1 211 218 10.2174/1874070700802010211 28 Sitaula R. Elmoazzen H. Toner M. Bhowmick S. Desiccation tolerance in bovine sperm: a study of the effect of intracellular sugars and the supplemental roles of an antioxidant and a chelator Cryobiology 2009 58 3 322 330 10.1016/j.cryobiol.2009.03.002 2-s2.0-67349205052 19318090 29 Petrenko Y. A. Rogulska O. Y. Mutsenko V. V. Petrenko A. Y. A sugar pretreatment as a new approach to the Me2SO- and xeno-free cryopreservation of human mesenchymal stromal cells Cryo Letters 2014 35 3 239 246 2-s2.0-84906912799 24997842 30 Jamil K. Crowe J. H. Tablin F. Oliver A. E. Arbutin enhances recovery and osteogenic differentiation in dried and rehydrated human mesenchymal stem cells Cell Preservation Technology 2005 3 4 244 255 10.1089/cpt.2005.3.244 2-s2.0-30344473587 31 Shirakashi R. Köstner C. M. Müller K. J. Kürschner M. Zimmermann U. Sukhorukov V. L. Intracellular delivery of trehalose into mammalian cells by electropermeabilization Journal of Membrane Biology 2002 189 1 45 54 10.1007/s00232-002-1003-y 2-s2.0-0036731526 12202951 32 Guo N. Puhlev I. Brown D. R. Mansbridge J. Levine F. Trehalose expression confers desiccation tolerance on human cells Nature Biotechnology 2000 18 2 168 171 10.1038/72616 2-s2.0-0033950708 33 Eroglu A. Russo M. J. Bieganski R. Intracellular trehalose improves the survival of cryopreserved mammalian cells Nature Biotechnology 2000 18 2 163 167 10.1038/72608 2-s2.0-0033951109 34 Oliver A. E. Jamil K. Crowe J. H. Tablin F. Loading human mesenchymal stem cells with trehalose by fluid-phase endocytosis Cell Preservation Technology 2004 2 1 35 49 10.1089/153834404322708745 2-s2.0-0347360092 35 Elliott G. D. Liu X. H. Cusick J. L. Trehalose uptake through P2X7 purinergic channels provides dehydration protection Cryobiology 2006 52 1 114 127 10.1016/j.cryobiol.2005.10.009 16338230 36 Wei Y. Li C. Zhang L. Xu X. Design of novel cell penetrating peptides for the delivery of trehalose into mammalian cells Biochimica et Biophysica Acta (BBA)—Biomembranes 2014 1838 7 1911 1920 10.1016/j.bbamem.2014.02.011 2-s2.0-84898759691 24583082 37 Abazari A. Meimetis L. G. Budin G. Bale S. S. Weissleder R. Toner M. Engineered trehalose permeable to mammalian cells PLoS ONE 2015 10 6 e0130323 10.1371/journal.pone.0130323 2-s2.0-84938675673 38 Chakraborty N. Menze M. A. Elmoazzen H. Trehalose transporter from African chironomid larvae improves desiccation tolerance of Chinese hamster ovary cells Cryobiology 2012 64 2 91 96 10.1016/j.cryobiol.2011.11.007 2-s2.0-84856608197 22155480 39 Eroglu A. Toner M. Toth T. L. Beneficial effect of microinjected trehalose on the cryosurvival of human oocytes Fertility and Sterility 2002 77 1 152 158 10.1016/S0015-0282(01)02959-4 2-s2.0-0036156006 11779606 40 Wowk B. Electric and magnetic fields in cryopreservation Cryobiology 2012 64 3 301 303 10.1016/j.cryobiol.2012.02.003 2-s2.0-84861230212 22330639 41 Lee S.-Y. Huang G.-W. Shiung J.-N. Magnetic cryopreservation for dental pulp stem cells Cells Tissues Organs 2012 196 1 23 33 10.1159/000331247 2-s2.0-84862817467 22285908 42 Yalvaç M. E. Ramazanoglu M. Tekguc M. Human tooth germ stem cells preserve neuro-protective effects after long-term cryo-preservation Current Neurovascular Research 2010 7 1 49 58 10.2174/156720210790820181 2-s2.0-77949677147 20158462 43 Potts M. Slaughter S. M. Hunneke F.-U. Garst J. F. Helm R. F. Desiccation tolerance of prokaryotes: application of principles to human cells Integrative and Comparative Biology 2005 45 5 800 809 10.1093/icb/45.5.800 2-s2.0-30744437405 21676831 44 Tang M. Wolkers W. F. Crowe J. H. Tablin F. Freeze-dried rehydrated human blood platelets regulate intracellular pH Transfusion 2006 46 6 1029 1037 10.1111/j.1537-2995.2006.00838.x 2-s2.0-33744463370 16734821

PMC005xxxxxx/PMC5002306.txt

==== Front Comput Intell NeurosciComput Intell NeurosciCINComputational Intelligence and Neuroscience1687-52651687-5273Hindawi Publishing Corporation 10.1155/2016/6459873Research ArticleA Long-Term Prediction Model of Beijing Haze Episodes Using Time Series Analysis http://orcid.org/0000-0002-1613-843XYang Xiaoping 1 http://orcid.org/0000-0002-0056-8750Zhang Zhongxia 1 Zhang Zhongqiu 2 Sun Liren 1 Xu Cui 1 http://orcid.org/0000-0001-8503-2535Yu Li 1 * 1School of Information, Renmin University of China, Beijing 100872, China2School of Computer Science, Northeastern University, Shenyang 110819, China*Li Yu: buaayuli@ruc.edu.cnAcademic Editor: Francisco Martínez-Álvarez 2016 14 8 2016 2016 645987320 4 2016 28 6 2016 10 7 2016 Copyright © 2016 Xiaoping Yang et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.The rapid industrial development has led to the intermittent outbreak of pm2.5 or haze in developing countries, which has brought about great environmental issues, especially in big cities such as Beijing and New Delhi. We investigated the factors and mechanisms of haze change and present a long-term prediction model of Beijing haze episodes using time series analysis. We construct a dynamic structural measurement model of daily haze increment and reduce the model to a vector autoregressive model. Typical case studies on 886 continuous days indicate that our model performs very well on next day's Air Quality Index (AQI) prediction, and in severely polluted cases (AQI ≥ 300) the accuracy rate of AQI prediction even reaches up to 87.8%. The experiment of one-week prediction shows that our model has excellent sensitivity when a sudden haze burst or dissipation happens, which results in good long-term stability on the accuracy of the next 3–7 days' AQI prediction. National Natural Science Foundation of China71271209Beijing Municipal Natural Science Foundation4132052Humanity and Social Science Youth Foundation of Ministry of Education of China11YJC630268 ==== Body 1. Introduction Industry of developing countries is mainly centralized around big cities, accompanied by a large population, consumption, and pollution. Together with Tianjin city and Hebei province, Northern China has become one of the most prosperous and polluted areas on Earth. By 2013, the transient population of Beijing was 37.5 million, and the intermittent outbreak of air pollution has greatly impacted every citizen's life: physiological diseases [1, 2], depression, and poor visibility in traffic [3, 4]. The main component of haze is pm2.5 (particulate matters less than 2.5 μm in aerodynamic diameter), and the concentration of pollution is described with Air Quality Index (AQI, the concentration of pm2.5). The Chinese Government began to monitor and record pm2.5 concentrations for major cities since 2013 [5]. According to the report of Quan et al. [6], the AQI reached 600 in Beijing during the haze event in January 2013. In recent years, more and more papers have referred to the haze episodes and the consequences in Northern China [7–11]. Researchers pointed out that, over the coming years, haze episodes would continue to burst frequently in Northern China [12]. This paper presents an AQI prediction model of Beijing based on time series analysis. We collected Beijing's AQI data of 29 continuous months since 2013 and constructed a dynamic structural prediction model. Statistical methods are used to obtain the maximum likelihood estimation of the prediction model. And both short-term and long-term experiments are carried out to test the accuracy and robustness of our model. The remainder of this paper is organized as follows. In Section 2, we introduce recent related work. Section 3 presents our prediction model and proves our model to be a vector autoregressive model. Experiments and evaluations are reported in Section 4. We conclude the paper in Section 5 with future works. 2. Related Work Generally, pm2.5, or haze, is born mainly through anthropogenic factors [13–16] and eliminated by natural diffusion. Several days after emission, secondary pm2.5 is produced through photochemical reactions among indiffusible pollutants. Secondary pm2.5 is the principal component in most severe haze episodes in China [17]. A typical way of haze prediction is to use pollutant emission data (CO, SO2, and NOx) in the simulation [5, 18]. Huang et al. [14] analyzed the chemical compositions of pm2.5 and used chemical mass balance to identify the emission sources. Other more complex models are proposed to introduce the atmospheric features, chemistry components, and transport factors [15]. But the more common case is that pollutant emission data usually increase or decrease synchronously with AQI. Sun [19] took population, car ownership, and GDP into consideration and proposed a statistical index system of average annual haze episode days. They found that although most factors contribute to predicting pm2.5, the annual average of NOx is negatively correlated with average severely polluted days. The paper [12] established a cubic exponential smoothing model by introducing dust emission into haze prediction. Liang et al. pointed out that there are various distribution and transmission patterns of pm2.5 [20]. In fact, Wang et al. mentioned that the government control policy should be considered in model simulations [9]. Many researches use backpropagation neural network as the simulation model [19, 21]. Statistical time series analysis is rarely used in haze prediction, so long-term haze prediction is difficult for current methods to accomplish [22]. Multiple linear regression models also perform well on daily scale prediction [23, 24]. However, the test data of existing researches is not ample; for example, [21] tested the prediction accuracy on only 3 days. Besides, Zhang et al. pointed out that pm2.5 accumulation in previous days significantly affects the present daily pm2.5 concentration, which should also be a concern in the modeling process [22]. Considering the above points, this paper presents a new AQI prediction model integrated with natural factor, humanity factor, and self-evolution factor. 3. The Prediction Model of Beijing's Daily AQI 3.1. The Parameters and Architecture of the Prediction Model The change of daily pm2.5 concentration depends on two factors: daily overall production of pm2.5 by human activities P t and daily overall natural diffusion or overall natural accumulation of pm2.5 C t. The production of haze depends a lot on the control policies of the government toward the emission of industry fuels I t. The diffusion of haze mainly depends on the airflow W t. Besides, complex chemical changes could occur between pm2.5 and other pollutants; thus, previous day's pm2.5 concentration also affects the AQI, which could be seen as the evolution result of previous day's pm2.5 and is represented by Y t. Apparently, P t − C t could be directly observed. P t is generated by a semimanual method. P t is mainly related to daily human activities, and we calculate P t from AQI sequences of no less than five consecutive sunny and windless days. Special circumstances are also considered. In winter, P t will be larger because the heating system is on. The car usage restrictions and temporary stoppage of factories during Beijing APEC 2014 are also taken into consideration. C t is then calculated as P t − (P t − C t). Sometimes, C t is greater than zero, which means pm2.5 accumulates because of nonhuman factors. Thus, the daily net growth of pm2.5 (P t − C t) is a function of the evolution result Y t, the industry control index I t, and the forecast of wind power W t. Consider this problem as a dynamic structural model, and our model can be described as (1) Pt−Ct=β0+β1Yt+β2Wt+β3It+β4Pt−1−Ct−1+μtD. Parameters β 1, β 2, and β 3, respectively, represent the effect caused by the pm2.5 of the previous day, the wind power, and the industry control index. The net growth of previous day's pm2.5 partly affects present day's pm2.5 and partly affects the next day's pm2.5. The parameter β 4 represents this “partial adjustment.” The disturbance μ t D represents other factors which affect present day's pm2.5. 3.2. Complexity Reduction of the Prediction Model In order to facilitate the research and modeling process, we have proved that this model could be reduced to a vector autoregressive model. Proposition 1 . Formula (1) is a vector autoregressive model. Proof Assume that there exists sequence autocorrelation in formula (1). The autocorrelation is (2) μtD=ρμt−1D+υtD in which υ t D is white noise. Here, we apply the Cochrane-Orcutt iteration to rewrite formula (2): (3) 1−ρLμtD=υtD, where L is the lag operator (LV t ≡ V t−1), which can convert the last phase to current value in a time series. The next work is to find the most satisfying value of ρ through successive iteration method. Specifically, this method uses residual error to estimate the unknown ρ. Assume that we use previous p days' AQI to predict present day's AQI. Multiply (1 − ρL) on both sides of formula (1); the expansion formula will be as follows: (4) Pt=k1+β120Ct+β130It+β140Yt+β150Wt+β111Pt−1+β121Ct−1+β131It−1+β141Yt−1+β151Wt−1+β112Pt−2+β122Ct−2+β132It−2+β142Yt−2+β152Wt−2+⋯+β11pPt−p+β11pCt−p+β11pIt−p+β11pYt−p+β11pWt−p+υtD. In the substitution process, many assumptions are neglected. But the ordinary least square method (OLS estimation) should not be used in the estimation of formula (4), because OLS can only illustrate the relationship between daily pm2.5 production and the policy control index, the accumulation of history pm2.5, and the wind power. The net growth of previous day's pm2.5 is only one reason of the correlation of these variables. The government could make policies to control pm2.5 production of industry to obtain “satisfying” daily production of pm2.5; that is, I t is an endogenous variable. And the policy control index depends on present day's and previous p days' accumulation of history pm2.5, the wind power, the daily production of pm2.5, and daily diffusion of pm2.5: (5) It=k3+β310Pt+β320Ct+β330Yt+β340Wt+β311Pt−1+β321Ct−1+β331Yt−1+β341Wt−1+β351It−1+β312Pt−2+β322Ct−2+β332Yt−2+β342Wt−2+β352It−2+⋯+β31pPt−p+β32PCt−p+β33PYt−p+β34PWt−p+β35PIt−p+υtC, where υ t C represents the influence brought about by other policies. The net growths of previous days' pm2.5 and policy control index also have an effect on daily accumulation of pm2.5: (6) Yt=k4+β410Pt+β420Ct+β430It+β440Wt+β411Pt−1+β421Ct−1+β431It−1+β441Wt−1+β451Yt−1+β412Pt−2+β422Ct−2+β432It−2+β442Wt−2+β452Yt−2+⋯+β41pPt−p+β42pCt−p+β43pIt−p+β44pWt−p+β45pYt−p+υtA, where υ t A represents other factors that influence daily accumulation of pm2.5. Analogized from formulas (4), (5), and (6), C t and W t can both be written in a similar form. Join formulas (4), (5), and (6) together, and rewrite them into vector form: (7) B0xt=K+B1xt−1+B2xt−2+⋯+Bpxt−p+υt in which (8) xt=Pt,Ct,It,Yt,WtT,υt=υtD,υtS,υtC,υtA,υtHT,K=k1,k2,k3,k4,k5,B0=1−β120−β130−β140−β150−β2101−β230−β240−β250−β310−β3201−β340−β350−β410−β420−β4301−β450−β510−β520−β530−β5401. In B 0, the parameters in the 1st, 2nd, 3rd, 4th, and 5th row, respectively, relate P t, C t, I t, Y t, and W t to the other variables. Every B s is a 5∗5 matrix. Premultiply formula (7) by B 0 −1 (the inverse matrix of B 0): (9) xt=c+Ψ1xt−1+Ψ2xt−2+⋯+Ψpxt−p+εt in which (10) c=B0−1K,Ψs=B0−1Bs,εt=B0−1υt. This is the standard form of vector autoregressive model. So it is proved that our prediction model (formula (1)) is in fact a vector autoregressive model. The regression parameters of our haze prediction model can be obtained as follows. Let (11) −Γ=KB1B2⋯Bp,yt=1xt−1xt−2⋯xt−pT. The dynamic structural system (formula (7)) isss (12) B0xt=−Γxt+υt. Assume that the disturbance terms are not sequence correlated or correlated to each other, which means (13) EυtυτT=D,t=τ,0,t≠τ. D is a main diagonal matrix. Formula (12) could be written as (14) xt=ΦTyt+εt in which (15) ΦT=B0−1−Γ,εt=B0−1υt. Let Ω be the variance-covariance matrix of ε t: (16) Ω=EεtεtT=B0−1EυtυtTB0−1T=B0−1DB0−1T. Suppose B 0 is a lower triangular matrix, in which all main diagonal elements are assigned 1, and D is a main diagonal matrix. The parameters (B 0, Γ, D) can be obtained through the maximum likelihood estimation of complete information. The maximum likelihood estimation of Ω can be obtained by the variance-covariance matrix of regression residual. Finally, B⌢0-1 and D are calculated through triangular decomposition of Ω⌢; thus, Γ can be evaluated. Above all, the prediction model of Beijing AQI has considered factors including industry emission and policy control, together with the chemical changes of previous days' pollution accumulation and the diffusion conditions. This model also takes the correlations between these factors into consideration and introduces time series haze features into the dynamic structural model. The policy control index is simulated by the record of 4 severe haze episodes during this period. The diffusion is evaluated by weather record of daily wind power. 4. Model Evaluations We collected the daily AQI and daily weather information from 28 Oct. 2013 to 31 Mar. 2016. This complete sequence is used to test the accuracy of the prediction model. The next day's AQI prediction experiment (Section 4.1) and long-term AQI prediction experiment (Section 4.2) are both implemented. The next day's AQI prediction is evaluated from two perspectives: the accuracy of daily prediction and the accuracy of statistical analysis. 4.1. Next Day's AQI Prediction The next day's weather forecast information is applied in next day's AQI prediction. The observed and predicted daily mean AQI in Beijing are illustrated in Figure 1. The simulation result shows that the predicted value matched the observed value very well on the whole sequence of 886 days. Sometimes, there is severe deviation from the observed value; for example, on 19 Feb. 2014, the observed AQI was 89, while our model gives a prediction of 209, with an offset of 135%. But the fact is, in the afternoon of 19 Feb., the wind of Beijing suddenly changed from northeasterly to southwesterly, and by 19:00 the AQI has reached already up to 170, which could be interpreted as our model successfully forecasted a severe haze outbreak several hours in advance; in the coming 7 days, the average daily AQI of Beijing is 305. The occasional occurrence of this “foreseeing” phenomenon is caused by coarse time granularity (daily), and this phenomenon is marked with red ellipse in Figure 1. These marks indicate that our model could “foresee” the sharp change of both outbreaks and diffusions. Most haze outbreaks and diffusions could be accurately simulated; some could be foreseen but could never be delayed. Figure 2(a) is a scatter diagram of daily AQI, including both observed value and predicted value. Most points lie close to y = x (the red line). But some points lie in a queue at the bottom part, which means the observed AQI exceeds 200, while the predicted value is less than 50. There are altogether 15 such outliers, 7 of which “foresee” haze diffusion, while the other 8 bug points could not be well interpreted. All the 15 points are checked and listed in Table 1. “✓” means a “foreseeing” phenomenon, and “?” represents bug points. Figure 2(b) is a scatter diagram of annual AQI (sum of daily AQI in a certain year). Our data covers only 2 months of 2013 and 3 months of 2016, so, in this diagram, these 2 points lie in the lower left corner. The pie chart in Figure 3 shows the distribution of prediction accuracy. The deviation of predicted and observed AQI is obtained through the following formula: (17) Devt=PredictedValue−ObservedValueObservedValue×100%. Figure 3 shows that 55% predictions match the observed values very well (<20% deviation). The purple part is mainly caused by the “foreseeing” phenomenon. Most samples of the red part come from less-polluted days. For example, on 12 Jan. 2016, the AQI prediction is 40 while the observed AQI is 29, which makes a deviation of 37.9%. In fact, statistics also indicate that our model performs better in worse air conditions (Figure 4). A sample is correctly predicted if the deviation of a sample is less than 20% or the predicted air quality level matches the observed level. 4.2. Long-Term AQI Prediction In the long-term prediction, we use history haze data sequence and weather forecast information to predict the next 7 days' AQI. A sample is correctly predicted if the deviation of a sample is less than 20% or the predicted air quality level matches the observed level. From 26 Dec. 2015 to 31 Mar. 2016, we predict the AQI in the next 7 days and check the accuracy of n-day predictions. Figure 5 shows the accuracy of long-term prediction in the 91 days' experiment. Figure 5 shows that the accuracy stays stable on the next 3, 4, 5, 6, and 7 days' AQI prediction, which indicates that our model has excellent robustness on the task of long-term prediction. The next day's prediction accuracy surprisingly reaches 79.1%, which is far better than the experiment in Section 4.1. The reason is that, during the 91 days, 6 haze episodes attacked Beijing. These frequent attacks did contribute a lot to the overall performance because our model is very sensitive to sudden changes of AQI, including outbreaks and diffusions (Section 4.1; Figure 4). Figures 6 and 7 show several haze episodes during the 91 days. Both figures show a pm2.5 change process of more than 2 weeks. Figure 6 also shows a “foreseeing” phenomenon caused by coarse time granularity, marked by a red ellipse. 5. Conclusion and Future Work This paper presented a dynamic structural model to predict Beijing's daily AQI. This model integrated natural factor, humanity factor, and self-evolution factor into the time series model. This dynamic structural measurement model of daily haze increment is proven to be a vector autoregressive model. Experiments reflected two highlights of this model. First, our model is very sensitive to and performs very well on predicting sudden changes of AQI, including both outbreaks and diffusions. Second, the model has great robustness on the task of long-term AQI prediction. Lastly, limited by the coarse time granularity, our model sometimes “foresees” but never delays or misses any sudden changes of haze. Many researchers use simple backpropagation neural network to accomplish nonlinear prediction models. But since methods based on time series are proven to be effective in haze prediction modeling, we believe that recurrent neural networks give better performances in such a prediction task. Although the related factors are limited in existing models, the overfitting problem should still be concerned, because, in long-term prediction, a deviation could spread and be exaggerated in the following days' predictions. Acknowledgments This research was supported by the National Natural Science Foundation of China under Grant 71271209, Beijing Municipal Natural Science Foundation under Grant 4132052, and Humanity and Social Science Youth Foundation of Ministry of Education of China under Grant 11YJC630268. Competing Interests The authors declare that they have no competing interests. Figure 1 (a, b, c) Next day's AQI prediction on 886 continuous days. Figure 2 (a) Daily AQI of the 886 days. (b) Annual AQI from 2013 to 2016. Figure 3 The deviation of predicted and observed AQI. Figure 4 Prediction accuracies of different air qualities. Figure 5 The accuracy of long-term AQI prediction. Figure 6 Three haze episodes in Jan. 2016. Figure 7 Three haze episodes in Feb. 2016. Table 1 All the 15 outliers in Figure 2(a). Date of outlier Label Nov. 2, 2013 ✓ Dec. 7, 2013 ? Dec. 25, 2013 ✓ Feb. 14, 2014 ? Feb. 25, 2014 ? Mar. 26, 2014 ? Oct. 10, 2014 ✓ Oct. 11, 2014 ✓ Nov. 19, 2014 ? Nov. 20, 2014 ? Nov. 30, 2014 ✓ Dec. 9, 2014 ? Jan. 4, 2015 ? Jan. 15, 2015 ✓ Mar. 7, 2015 ✓ ==== Refs 1 Gao L. Jia G. Zhang R. Visual range trends in the Yangtze River Delta Region of China, 1981–2005 Journal of the Air and Waste Management Association 2011 61 8 843 849 10.3155/1047-3289.61.8.843 2-s2.0-80051744279 21874955 2 Liu Q. Baumgartner J. Zhang Y. Schauer J. J. Source apportionment of Beijing air pollution during a severe winter haze event and associated pro-inflammatory responses in lung epithelial cells Atmospheric Environment 2016 126 28 35 10.1016/j.atmosenv.2015.11.031 2-s2.0-84949197156 3 Tie X. Brasseur G. P. Zhao C. Chemical characterization of air pollution in Eastern China and the Eastern United States Atmospheric Environment 2006 40 14 2607 2625 10.1016/j.atmosenv.2005.11.059 2-s2.0-33645409847 4 Deng J. Wang T. Jiang Z. Characterization of visibility and its affecting factors over Nanjing, China Atmospheric Research 2011 101 3 681 691 10.1016/j.atmosres.2011.04.016 2-s2.0-79960145571 5 MEPC (Ministry of Environmental Protection of China) China Environment Yearbook 2014 Beijing, China China Environment Yearbook Press 6 Quan J. Tie X. Zhang Q. Characteristics of heavy aerosol pollution during the 2012-2013 winter in Beijing, China Atmospheric Environment 2014 88 83 89 10.1016/j.atmosenv.2014.01.058 2-s2.0-84894239955 7 Zhao X. J. Zhao P. S. Xu J. Analysis of a winter regional haze event and its formation mechanism in the North China Plain Atmospheric Chemistry & Physics 2013 13 11 5685 5696 10.5194/acp-13-5685-2013 2-s2.0-84887197034 8 Wang H. An J. Shen L. Mechanism for the formation and microphysical characteristics of submicron aerosol during heavy haze pollution episode in the Yangtze River Delta, China Science of the Total Environment 2014 490 501 508 10.1016/j.scitotenv.2014.05.009 2-s2.0-84901427335 24875262 9 Wang Y. Li L. Chen C. Source apportionment of fine particulate matter during autumn haze episodes in Shanghai, China Journal of Geophysical Research Atmospheres 2014 119 4 1903 1914 10.1002/2013jd019630 2-s2.0-84896909274 10 Ji D. Li L. Wang Y. The heaviest particulate air-pollution episodes occurred in northern China in January, 2013: insights gained from observation Atmospheric Environment 2014 92 546 556 10.1016/j.atmosenv.2014.04.048 2-s2.0-84901046102 11 Hu J. Wang Y. Ying Q. Zhang H. Spatial and temporal variability of PM2.5 and PM10 over the North China Plain and the Yangtze River Delta, China Atmospheric Environment 2014 95 598 609 10.1016/j.atmosenv.2014.07.019 2-s2.0-84904103448 12 Hou Q. Yang H. Analysis and forecasting of haze weather based on the cubic exponential smoothing model Environmental Protection Science 2014 6 73 77 13 Guo S. Hu M. Zamora M. L. Elucidating severe urban haze formation in China Proceedings of the National Academy of Sciences of the United States of America 2014 111 49 17373 17378 10.1073/pnas.1419604111 2-s2.0-84917691429 25422462 14 Huang R.-J. Zhang Y. Bozzetti C. High secondary aerosol contribution to particulate pollution during haze events in China Nature 2015 514 7521 218 222 10.1038/nature13774 2-s2.0-84908317345 25231863 15 Vieno M. Heal R M. Hallsworth S. The role of long-range transport and domestic emissions in determining atmospheric secondary inorganic particle concentrations across the UK Atmospheric Chemistry & Physics 2014 14 8435 8447 16 Kiesewetter G. Borkenkleefeld J. Schöpp W. Modelling street level PM 10 concentrations across Europe: source apportionment and possible futures Atmospheric Chemistry & Physics 2014 14 18315 18354 17 Zhang R. Wang L. Khalizov A. F. Formation of nanoparticles of blue haze enhanced by anthropogenic pollution Proceedings of the National Academy of Sciences of the United States of America 2009 106 42 17650 17654 10.1073/pnas.0910125106 2-s2.0-70449568307 19815498 18 Wooldridge J. M. Econometric Analysis of Cross Section and Panel Data 2010 2nd Cambridge, Mass, USA MIT Press MR2768559 19 Sun Y. Research on the Forecast of Haze Days Based on the Principal Component-BP Neural Network Model 2015 Jiangsu University 20 Liang X. Zou T. Guo B. Assessing Beijing's PM2.5 pollution: severity, weather impact, APEC and winter heating Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences 2015 471 2182 10.1098/rspa.2015.0257 2-s2.0-84946077874 21 Ai H. Shi Y. Study on prediction of haze based on BP neural network Computer Simulation 2015 32 402 405 22 Zhang X. Wu Y. Gu B. Characterization of haze episodes and factors contributing to their formation using a panel model Chemosphere 2016 149 320 327 10.1016/j.chemosphere.2016.01.090 26874060 23 Li L. Qian J. Ou C.-Q. Zhou Y.-X. Guo C. Guo Y. Spatial and temporal analysis of Air Pollution Index and its timescale-dependent relationship with meteorological factors in Guangzhou, China, 2001–2011 Environmental Pollution 2014 190 75 81 10.1016/j.envpol.2014.03.020 2-s2.0-84898637711 24732883 24 Tian G. Qiao Z. Xu X. Characteristics of particulate matter (PM10) and its relationship with meteorological factors during 2001-2012 in Beijing Environmental Pollution 2014 192 266 274 10.1016/j.envpol.2014.04.036 2-s2.0-84904068632 24857048

PMC005xxxxxx/PMC5002307.txt

==== Front Comput Intell NeurosciComput Intell NeurosciCINComputational Intelligence and Neuroscience1687-52651687-5273Hindawi Publishing Corporation 10.1155/2016/3253678Research ArticleA Model of Generating Visual Place Cells Based on Environment Perception and Similar Measure http://orcid.org/0000-0002-8384-7842Zhou Yang * http://orcid.org/0000-0002-6691-5225Wu Dewei Information and Navigation College, Air Force Engineering University, Xi'an, Shaanxi 710077, China*Yang Zhou: yydayl@sina.cnAcademic Editor: Michele Migliore 2016 14 8 2016 2016 325367826 4 2016 15 6 2016 11 7 2016 Copyright © 2016 Y. Zhou and D. Wu.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.It is an important content to generate visual place cells (VPCs) in the field of bioinspired navigation. By analyzing the firing characteristic of biological place cells and the existing methods for generating VPCs, a model of generating visual place cells based on environment perception and similar measure is abstracted in this paper. VPCs' generation process is divided into three phases, including environment perception, similar measure, and recruiting of a new place cell. According to this process, a specific method for generating VPCs is presented. External reference landmarks are obtained based on local invariant characteristics of image and a similar measure function is designed based on Euclidean distance and Gaussian function. Simulation validates the proposed method is available. The firing characteristic of the generated VPCs is similar to that of biological place cells, and VPCs' firing fields can be adjusted flexibly by changing the adjustment factor of firing field (AFFF) and firing rate's threshold (FRT). National Natural Science Foundation of China61273048 ==== Body 1. Introduction In 1971, O'Keefe and Dostrovsky [1] found that some neurons in rat's hippocampus exhibited location selectivity during rat free moving. They defined these cells as place cells. The activity of place cells is related to the spatial location. Once the animal is located at a relatively narrow region, the corresponding place cell can fire with a high rate. But in other regions, this place cell does not fire or fires with a low rate. This firing characteristic shows animal's spatial representation pattern and navigation mechanism to a certain extent, and it lays a biological foundation to the bioinspired navigation technology, for example, RatSLAM [2, 3], which is realized by simulating the navigation mechanism in hippocampus and it can provide the robot with a spatial awareness of entire environment. How to generate place cells? Brain neuroscience points out that the firing of place cells can be activated by the idiothetic information (e.g., self-motion) and allothetic information (e.g., visual) [4–7]. For the idiothetic activity mode, it can be realized through the transformation from the grid cells to the place cells. The existing models include the following: transformation model based on Fourier analysis [8], transformation model based on competitive learning [9], transformation model based on independent component analysis (ICA) [10], and so on. For the allothetic activity mode, it is achieved mainly through the process of obtaining landmarks, place code, and the calculation of place cell's firing rate. This paper mainly focuses on the second-activity mode based on visual information, and the generated place cells are called visual place cells (VPCs) [11, 12]. Doboli et al. [13] proposed an attractor model of the hippocampus. In their model, the external sensory input encodes distances to perceived landmarks as well as allocentric bearings to them, and VPC's external receptiveness is formed by a product of Gaussian functions. Gaussier et al. [11, 12, 14, 15] had done a lot of work in the field of generating VPCs. In their researches, VPC is defined by a spatial constellation of online learned visual view, which is obtained by the process of gradient processing, convolved with a difference of Gaussian filter (DoG), selection of focus point, and log-polar transformation. During the learning of a location, each local view in log-polar coordinate is learned as a landmark for the system; then, the landmarks' recognition information and their spatial localization in the visual field are merged in a product space to define a place code, and the VPC is generated after the recognition and association of the place code. DalleMole and Araújo [16] proposed a topological map of place cells. To produce a cognitive topological map of the environment, they acquire and organize knowledge associating it with places and define an activation function to determine the similarity between the current perception, the stored memories, and so on. Through the analysis of the current research status, we know that there is a great similarity in the existing VPCs' generation methods, and their differences mainly are reflected in the acquisition and representation of landmarks and the calculation of VPCs' firing rates. This paper abstracts a unified VPCs' generation model after fully considering their similarities and differences. The acquisition and representation of landmarks are included in the environment perception. The calculation of VPCs' firing rates is included in the similarity measure. As a result, the generation process of VPCs and the issues that need to be solved are stated more clearly, which can provide a complete research idea for the related researchers. Simultaneously, according to the abstracted model, a specific method for generating VPCs is presented, and it is validated and analyzed by the simulation. 2. VPCs' Generation Model The firing activity of biological place cells exhibits strong location selectivity. When the animal is located in the region represented by place cells, the corresponding place cells will fire with high rate. In biology, the field with the firing activity is called “place field” or “firing field.” Figure 1 shows the firing activity of place cells recorded by biological experiment [17], where the trajectory of rat is indicated by gray line, and each spike is plotted in red. In the theory of hippocampal cognitive map, place cells are believed to constitute the basic unit of this cognitive map. A single place cell can represent a specific location in the environment, and the firing activity of the whole place cells can describe and represent the entire environment. By analyzing and summarizing the existing VPCs' generation methods [10–16], VPCs' generation model is abstracted here based on environment perception and similarity measure, as shown in Figure 2. VPCs' generation process is divided into environment perception, similarity measure, and recruiting of a new VPC. The environment perception phase achieves the landmarks' acquisition and place code. Landmarks' acquisition resolves what information is used as reference and how to represent it. In this paper, a landmark is also described by two kinds of information as shown in [15]. One is the recognition information (What) and the other one is the location information (Where). What information and Where information are merged to get the representation information of a landmark. Place code is to define current location by the whole landmarks, and each generated VPC is associated with a place code. Similarity measure achieves the analysis of the similarity between current place code and the generated VPCs and then quantifies the VPCs' firing rates according to their similarity. The phase of recruiting a new VPC achieves the analysis of the firing status of the existing VPCs and flexibly recruits a new VPC. VPCs' generation process can be summarized as follows: during spatial exploration, the vehicle acquires the recognition information and location information of landmarks by especial environment perception approach and gets the place code according the landmarks' representation information. Then, VPCs' firing rates are calculated according to the similarity between their associated place codes and current place code. Finally, the comparison among the generated VPCs is implemented and the winner is compared with the given firing rate's threshold (FRT) (FRT is set as the condition to recruit a new VPC; it indicates the minimum firing rate that the generated VPCs should satisfy). If it is below FRT, the current place code is memorized and a new VPC is recruited to be associated with this memory. Next, a specific VPCs' generation method will be presented according to above-mentioned process; the details are as follows. Step 1 (acquire landmarks). Combining local invariant theory and our previous proposed model for landmarks' acquisition [18], the attention points which are obtained by the process of extraction of feature points, generation of saliency value, and selection of attention points in the visual image are used as landmarks. Considering we mainly focus on the generating VPCs, the landmarks' acquisition process is not discussed in detail here. Let the number of acquired landmarks at current moment be N(t). The recognition information and the saliency value of landmark i are denoted by r i(t) and s i(t), respectively. The distance and the orientation of landmark i relative to the vehicle are denoted by d i(t) and θ i(t), respectively. Then, the landmark is represented as follows: (1) lit=rit,sit,θit,dit,i=1,2,…,Nt, where l i(t) denotes the representation information of landmark i. θ i(t) denotes the orientation between north direction and the vector from vehicle to landmark i in clockwise direction. The schematic diagram of landmarks' orientation is shown in Figure 3. In this paper, the absolute direction is used as the reference direction to calculate landmarks' orientation. The reason is that if the reference direction (such as vehicle's running direction) may change at different location or different time, then the calculated landmarks' orientation relative to this kind of reference direction may be affected by the change of reference direction, which finally makes it difficult to measure the accurate relation between landmarks and vehicle at the same location. Step 2 (construct place code). The whole representation information of landmarks at current moment is combined to get current place code, denoted by C(t); namely, (2) Ct=l1t,l2t,…,lNtt. Step 3 (design similarity measure function to quantify VPCs' firing rates). In our paper, the similarity between VPCs' place codes and current place code is used to evaluate VPCs' firing rates. The similarity measure function is designed based on Euclidean distance and Gaussian function; the formula is as follows: (3) fkt=∑i=1Nkwie−wddit−dik2/σd2+wθθit−θik2/σθ2, where f k is the firing rate of VPC k. N k is the number of matched landmarks between VPC's place code and current place code. σ d 2 and σ θ 2 are defined as the adjustment factor of firing field (AFFF). They denote the influence on VPCs' firing fields from the distance and orientation, respectively. The bigger the AFFF, the larger the firing fields with high firing rate. w d = {0,1} and w θ = {0,1} denote whether the distance and orientation are used as the condition of similarity measure, respectively. If the answer is yes, its value is set to one, otherwise it is set to zero. w i denotes the contribution to calculating VPCs' firing rates. w i can be calculated by the saliency value of landmarks. The formula is as follows: (4) wi=si∑i=1Ntsi. From (4), we know that the bigger the landmark's saliency value (that is to say, the better the landmark's robustness), the bigger its contribution to calculating VPC's firing rate. Step 4 (recruit a new place cell). First, VPCs' firing rates are compared with each other. Then, the winner is compared with the given FRT. If the winner is below FRT, current place code is memorized, and a new VPC is recruited to be associated with this place code. Through the above four steps, the vehicle can generate VPCs during its spatial exploration. Next, the proposed method will be validated and analyzed by simulation. 3. Results and Analysis 3.1. Realization of the Method The simulation conditions are set as follows:(1) The vehicle's running environment is defined in a rectangular space, and its size is set to 40 m × 40 m. There are one hundred random distribution landmarks in the given space, as shown in Figure 4. (2) A recognition distance is set to select the landmarks which are associated with current vehicle's location. Its role is similar to the method used to acquire landmarks. Suppose that the vehicle can recognize the landmarks whose distance is between 10 m and 15 m relative to vehicle (in actual environment, this condition can be removed, and the acquisition of landmarks is implemented by specific algorithm). Besides, the saliency values of selected landmarks are equal. (3) The vehicle runs randomly in the defined space. Its maximal running speed is set to 5 m/s, and the location updating period is set to 1 s. Vehicle' speed remains unchanged in each period, but it changes randomly in different periods. Besides, the running direction is changed when the vehicle arrives at the boundary, and the relation between the new direction and the original direction obeys the reflection theorem. For the single spatial exploration, the total location updating step is set to 4000. (4) w d and w θ are set to 1. σ d 2 is set to 25. σ θ 2 is set to 100. FRT is set to 0.2. Figure 5 shows the firing status of partial VPCs after a single spatial exploration. Figure 6 shows the firing status after the overlapping of the whole generated firing fields. For the simulation results, vehicle's trajectory is indicated by black line. The firing rate indicated by dark red and dark blue is corresponding to the highest and lowest firing rate, respectively. Simulation results show that the behavior of the generated VPCs is similar to that of biological place cells. A single VPC's firing field is corresponding to a restricted region, and the overlapping of the firing fields of the whole VPCs can cover the entire space. Therefore, the proposed method is available. The generated VPCs can well simulate the firing activity of biological place cells. 3.2. Influence on VPCs from the Parameters Next, the influence on VPCs from AFFF, FRT, recognition distance, and location updating step will be discussed. Figure 7 shows the firing status of VPCs in different AFFFs. Each simulation is carried out at the same trajectory. The other parameters are set according to the simulation conditions of Section 3.1. Simulation results show that AFFF affects the firing status of generated VPCs. When the AFFF increases, the single VPC's firing field extends and simultaneously VPCs' number used to represent the spatial environment decreases. That is to say, the setting of AFFF can affect the spatial representation precision. The smaller the AFFF, the higher the representation precision, but it also increases the memory cost. Therefore, the AFFF can be set flexibly according to the demand of representation precision and memory cost. Figures 8 and 9 show the firing status of VPCs and the generated VPCs' number in different given FRTs, respectively. FRT is divided into four different values, including 0.1, 0.2, 0.3, and 0.4. Five different exploration trajectories are implemented in the same FRT. The other parameters are set according to the simulation conditions of Section 3.1. Simulation results show that FRT affects VPCs' firing field and the generated VPCs' number. The bigger the FRT, the smaller the single firing field, and simultaneously the generated VPCs' number increases. Thus, the spatial representation precision and the memory cost can also be adjusted by setting different FRTs. Figure 10 shows the generated VPCs' number in different recognition distances. The recognition distance is divided into three different intervals, including 5 m–10 m, 5 m–15 m, and 5 m–20 m. Five different exploration trajectories are implemented in the same recognition distance. The other parameters are set according to the simulation conditions of Section 3.1. Simulation results show that the recognition distance affects the generated VPCs' number. The longer the recognition distance, the less the generated VPCs' number. What is more, the number of obtained landmarks increases with the increasing of the recognition distance, so we can also get that the generated VPCs' number will decrease when more landmarks are used to calculate VPCs' firing rates. Figure 11 shows the generated VPCs' number in different location updating steps. Five different exploration trajectories are implemented in the same condition. The other parameters are set according to the simulation conditions of Section 3.1. Simulation results show that the more the location updating step, that is to say, the more the complete exploration of the environment, the more the generated VPCs' number, but the differences of generated VPCs' number will decrease or disappear when the higher number of location updating steps is implemented. 4. Conclusions Combining the existing method to generate VPCs, this paper abstracts a model of generating VPCs based on environment perception and similar measure. In the model, the acquisition and representation of landmarks are included in the environment perception, and the calculation of VPCs' firing rates is included in the similarity measure, which can provide clear and complete process to generate VPCs. Simulation results show that the firing characteristic of generated VPCs is similar to that of biological place cells. VPCs' firing fields are corresponding to local regions, and the overlapping of the whole firing fields can cover the explored space. Simultaneously, the firing fields and the generated VPCs' number can be adjusted by setting different AFFFs and FRTs. In the next researches, we will discuss the proposed model in the actual environment and especially analyze the influence on the generation results of VPCs from landmarks' distribution and saliency value. In the actual environment, the distribution of landmarks is usually complex. For some locations, a lot of landmarks may be perceived by the vehicle, but, for other locations, the perceived landmarks may be very few. Besides, the landmarks' saliency values should be calculated by specific algorithm. Acknowledgments This work is supported by National Natural Science Foundation of China (NSFC Grant no. 61273048). Competing Interests The authors declare that they have no competing interests. Figure 1 Firing pattern of place cells [17]. Figure 2 VPCs' generation model. Figure 3 Schematic diagram of landmarks' orientation. Figure 4 Landmarks in the given space. Figure 5 Firing status of partial VPCs. Figure 6 Firing status after the overlapping of the whole firing fields (generated VPCs' number is 50). Figure 7 Firing status of VPCs in different AFFFs. Figure 8 Firing status of VPCs in different FRTs. Figure 9 Generated VPCs' number in different FRTs (FRT is denoted by T). Figure 10 Generated VPCs' number in different recognition distances. Figure 11 Generated VPCs' number in different location updating steps. ==== Refs 1 O'Keefe J. Dostrovsky J. The hippocampus as a spatial map Brain Research 1971 34 1 171 175 10.1016/0006-8993(71)90358-1 2-s2.0-0015145985 5124915 2 Milford M. Wyeth G. Persistent navigation and mapping using a biologically inspired slam system International Journal of Robotics Research 2010 29 9 1131 1153 10.1177/0278364909340592 2-s2.0-77955596781 3 Ball D. Heath S. Wiles J. Wyeth G. Corke P. Milford M. OpenRatSLAM: an open source brain-based SLAM system Autonomous Robots 2013 34 3 149 176 10.1007/s10514-012-9317-9 2-s2.0-84877875441 4 Kesner R. P. Rolls E. T. A computational theory of hippocampal function, and tests of the theory: new developments Neuroscience and Biobehavioral Reviews 2015 48 92 147 10.1016/j.neubiorev.2014.11.009 2-s2.0-84914165973 25446947 5 Bush D. Barry C. Burgess N. What do grid cells contribute to place cell firing? Cell 2014 37 3 136 145 6 Muessig L. Hauser J. Wills T. J. Cacucci F. A developmental switch in place cell accuracy coincides with grid cell maturation Neuron 2015 86 5 1167 1173 10.1016/j.neuron.2015.05.011 2-s2.0-84930531470 26050036 7 Sheynikhovich D. Grèzes F. King J.-R. Arleo A. Villa A. E. P. Duch W. Érdi P. Masulli F. Palm G. Exploratory behaviour depends on multisensory integration during spatial learning Artificial Neural Networks and Machine Learning—ICANN 2012 2012 7552 296 303 Lecture Notes in Computer Science 10.1007/978-3-642-33269-2_38 8 Solstad T. Moser E. I. Einevoll G. T. From grid cells to place cells: a mathematical model Hippocampus 2006 16 12 1026 1031 10.1002/hipo.20244 2-s2.0-33846074376 17094145 9 Rolls E. T. Stringer S. M. Elliot T. Entorhinal cortex grid cells can map to hippocampal place cells by competitive learning Network: Computation in Neural Systems 2006 17 4 447 465 10.1080/09548980601064846 2-s2.0-33845456327 10 Franzius M. Vollgraf R. Wiskott L. From grids to places Journal of Computational Neuroscience 2007 22 3 297 299 10.1007/s10827-006-0013-7 MR2309028 2-s2.0-34249870109 17195112 11 Jauffret A. Cuperlier N. Gaussier P. Tarroux P. Ziemke T. Balkenius C. Hallam J. Multimodal integration of visual place cells and grid cells for navigation tasks of a real robot LNAI 7426 Proceedings of the International Conference on Simulation of Adaptive Behavior (SAB '12) 2012 136 145 12 Jauffret A. Cuperlier N. Gaussier P. From grid cells and visual place cells to multimodal place cell: a new robotic architecture Frontiers in Neurorobotics 2015 9 1, article 1 1 22 10.3389/fnbot.2015.00001 2-s2.0-84948986369 25904862 13 Doboli S. Minai A. A. Best P. J. A computational model of the interaction between external and internal cues for the control of hippocampal place cells Neurocomputing 2003 52–54 371 379 10.1016/s0925-2312(02)00843-3 2-s2.0-0037781045 14 Giovannangeli C. Gaussier P. Désilles G. Robust mapless outdoor, vision-based navigation Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS '06) October 2006 Beijing, China 3293 3300 10.1109/iros.2006.282501 2-s2.0-34250629985 15 Giovannangeli C. Gaussier P. Autonomous vision-based navigation: goal-oriented action planning by transient states prediction, cognitive map building, and sensory-motor learning Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS '08) September 2008 Nice, France Acropolis Convention Center 22 26 16 DalleMole V. L. Araújo A. F. R. Diamantaras K. Duch W. Iliadis L. S. A novel topological map of place cells for autonomous robots 6353 Proceedings of the 20th International Conference on Artificial Neural Networks, part II (ICANN '10) 2010 Springer 296 306 Lecture Notes in Computer Science (LNCS) 17 Moser E. I. Kropff E. Moser M.-B. Place cells, grid cells, and the brain's spatial representation system Annual Review of Neuroscience 2008 31 69 89 10.1146/annurev.neuro.31.061307.090723 2-s2.0-45949087462 18 Zhou Y. Wu D. Du J. Li W. A computational model for landmarks acquisition in positioning Journal of Intelligent and Robotic Systems: Theory and Applications 2016 82 3 537 553 10.1007/s10846-015-0276-1 2-s2.0-84944706399

PMC005xxxxxx/PMC5002317.txt

==== Front Eur Rev Aging Phys ActEur Rev Aging Phys ActEuropean Review of Aging and Physical Activity1813-72531861-6909BioMed Central London 16710.1186/s11556-016-0167-xResearch ArticleImpact of a randomized possible selves experiment on new retirees’ physical activity and identity Perras Mélanie G. M. melanie.perras@uottawa.ca 1Strachan Shaelyn M. shaelyn.strachan@umanitoba.ca 2Fortier Michelle S. mfortier@uottawa.ca 1Dufault Brenden brenden.dufault@umanitoba.ca 31 School of Human Kinetics, University of Ottawa, 125 University Private, Ottawa, Ontario K1N 6N5 Canada 2 Faculty of Kinesiology and Recreation Management, University of Manitoba, Winnipeg, MB R3T 2N2 Canada 3 George and Fay Yee Centre for Healthcare Innovation, College of Medicine, University of Manitoba, 735 McDermot Ave, Winnipeg, MB R3E 0T6 Canada 27 8 2016 27 8 2016 2016 13 1 723 2 2016 9 8 2016 © The Author(s) 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Retirement is not always associated with greater engagement in physical activity. Previous interventions informed by possible selves, a type of future-oriented self-representation, proved useful to increase physical activity in young adults. We thus wanted to explore if a similar intervention would yield favorable outcomes in new retirees. We also examined whether possible selves could help increase identity relative to the physical activity context. Identity circumscribes the meanings which help individuals define who they are in a given role (i.e., what it means to be a physically active person). The strength of identification as a physically active person increases when individuals endorse these meanings more strongly. Possible selves may be tied to identity as they allow individuals to imagine themselves as physically active, which has been argued to incite changes to one’s sense of self. Hence, the overall aim of this study was to determine whether a possible selves intervention would increase physical activity behaviour and physical activity identity in a group of newly-retired individuals. Methods A total of 294 participants were randomized into one of three groups: (a) a repeated group with three possible selves image generation exposures, (b) a one-time group with one possible selves image generation exposure, or (c) a control group. Participants completed self-report measures at baseline and follow-up assessments were taken at weeks 4, 8, and 12 of the study. The measures for the outcomes of interest were the Godin Leisure Time Exercise Questionnaire and the modified Exercise Identity Scale. Results Repeated measures mixed-effects models analyses with maximum likelihood estimation revealed no significant differences between groups on physical activity behaviour (p = 0.34) or physical activity identity (p = 0.97) at follow-up time points. However, a time effect was found for physical activity (p <.01) and physical activity identity (p <.01), which increased across time (baseline-to-12-week follow-up) in all three groups. Such a time effect (inconsequential to group assignment) suggests that the observed increases in physical activity and identity cannot be attributed to an exposure to a possible selves intervention. Conclusions While the intervention failed to significantly increase physical activity identity and physical activity in newly retired individuals, we suggest future research directions for interventions targeting new retired individuals. Keywords InterventionPossible selvesIdentityPhysical activityRetireeshttp://dx.doi.org/http://dx.doi.org/10.13039/100008572University of Ottawaissue-copyright-statement© The Author(s) 2016 ==== Body Background Retirement and self From mid-to-late life, people’s physical activity levels decline [1]. This decrease may result in older adults missing out on the benefits associated with regular physical activity [2]. Retirement may present older adults with an opportunity to thwart these age-related declines by affording them increased time to be physically active [3, 4]. However, research examining the association between retirement and self-reported physical activity reveals that for many retirees, this life transition does not lead to increased physical activity [5–8]. Researchers highlight the need to understand the “personal factors” [8] related to physical activity among retirees as they could be employed in interventions aimed at helping retirees transition into a physically active retirement [8, 9]. Self-perceptions are a category of personal factors that are central in understanding physical activity behaviour and behaviour change [10, 11] and may be especially relevant to new retirees. Indeed, the transition from work to retirement is viewed as a period for self-exploration and enhancement during which retirees may self-reflect on their health and who they are (e.g., [12–14]). As such, we turn to one type of self-perceptions: possible selves. Possible selves Possible selves are future-oriented self-representations which circumscribe the thoughts people have about their future and their potential related to specific roles [15]. Since possible selves are anchored in the future - and thus are not confined by what is factual in the present – they represent malleable elements of the self [16]. Possible selves are not just any aspect of one’s imagination, rather they usually hone in on individuals’ enduring hopes and fears for the future [15]. As per these authors, possible selves provide motivation and guidance for change. First, they provide a vivid snapshot of what is possible for someone (good or bad) which serves as a goal to be pursued or avoided. Secondly, possible selves provide a context to evaluate or interpret current behaviour. Possible selves and physical activity The possible selves construct has been employed in physical activity/exercise research. A cross-sectional study by Harju and Reed [17] found, in a sample of undergraduate students, that the supposed attainment of a possible self in the exercise purview positively coalesced with exercise behaviour, fitness, and self-efficacy for exercise. Additional cross-sectional findings from Whaley [18] demonstrate how the reporting of possible selves pertaining to body image can distinguish inactive and active, middle-age women. Furthermore, Whaley and Schrider [19] showed that older adults participating in a 10-week exercise program had well-articulated possible selves related to physical activity. Recently, possible selves have been assessed relative to the physical activity levels of new retirees [20]; these authors confirmed that newly retired individuals’ possible selves pertaining to physical activity were positively associated with concurrent physical activity. And this has been recently confirmed in a prospective study [21]. As such, physical activity possible selves appear to be positively related to aspects of physical activity. Possible selves have also been leveraged in interventions designed to increase physical activity. Ouellette, Hessling, Gibbons, Reis-Bergan, and Gerrard [22] found that university students who thought and wrote about their exercise participation from a future perspective (i.e., possible self) reported increased exercise behaviour four weeks later. Subsequently, Murru and Martin Ginis [23] found that inactive university students who read a script to help generate images of themselves in the future as either an exerciser or an inactive person reported greater physical activity four and six weeks post manipulation than participants in the control group. Similarly, Strachan, Marcotte, Giller, Brunet and Schellenberg [24] randomized insufficiently-active adults (aged 18–64) to either a self-enhancing possible selves intervention (consistent with [23]) or an enhanced “self-regulatory” possible selves intervention (participants’ possible self image included the self-regulatory steps they would take to achieve their possible self). Participants in both conditions reported significantly higher physical activity than their control counterparts (at 4- and 8-week follow-up for the self-regulatory condition and at 8 weeks for the self-enhancing condition). These studies suggest that a one-time focus on images, including possible selves, can impact physical activity behaviour. Given the promising results of these studies, a possible selves intervention may represent a reasonable tool through which we can attempt to increase physical activity among retirees and by the same token, heed recommendations by Ouellette et al. [22] to examine the impact of such interventions on older adults. Possible selves and identity While possible selves interventions can lead to increases in physical activity, they also hold the potential to have another important impact, that of increasing the extent to which individuals identify with physical activity in the present (physical activity identity). Indeed, researchers argue that the consideration of possible selves can affect and lead to enduring changes in current self-views (i.e., identity; [15, 25, 26]). Given that Markus [27] and Burke [28] have both commented that identities are resistant to change, individuals may require repeated occasions to reflect on a physically-active possible self for identity to be impacted. This assertion is consistent with that of Hoyle and Sherrill [25] who argue that the chronic activation of a possible self can impact current self-representations (i.e., identity). No research to date has examined the effects of a physical activity possible selves intervention on physical activity identity or considered whether repeated exposure to the possible self image is necessary for a change in identity to occur. If imagining oneself as a physically active retiree could encourage people to identify with physical activity, this consequence may represent an important way in which sustained physical activity behaviour change could be fostered [29]. Indeed, the ameliorative influence of identities on physical activity/exercise is well documented (for a review, see [30]). Strength of exercise identity also positively relates to self-efficacy for exercise, exercise frequency, and the articulation and carrying out of exercise intentions [31–33]. Furthermore, the positive relationship between physical activity identity and related behaviour is demonstrated in middle-aged and older adults [34–36] as well as in recent retirees [20, 21]. Therefore, possible selves interventions may lead to changes in physical activity identity, a known correlate of physical activity behaviour. Further, testing the utility of possible selves interventions at increasing identity should also be of interest to scholars interested in identity change. Experimental purposes and hypotheses The overall aim of this experiment was to determine if a possible selves intervention would increase physical activity identity and physical activity behaviour in a group of newly-retired individuals. One specific aim was to compare physical activity levels and identity across three conditions of an experimental design: a repeated possible selves intervention, a one-time possible selves intervention, and a control group. We hypothesized that participants in the repeated possible selves group would report higher levels of physical activity and identity over the course of the intervention than the control group (hypothesis 1) and the one-time possible selves group (hypothesis 2). The one-time condition was expected to report higher levels in physical activity identity and physical activity over the course of the intervention than the control group (hypothesis 3). Methods Participants A total of 294 newly retired men and women were randomized in the present study. Specific inclusion criteria included confirmation of retirement status (i.e., collecting pension/retirement-like income, and considering oneself “retired”) and confirmation of current health status (allowing for increased physical activity). To allow for a focus on newly-retired individuals, we restricted eligibility to retirees within three years of retirement. This cut-off is consistent with past research on new retirees’ leisure experiences (e.g., [37]). Further, we focused on recently-retired, as opposed to all retired individuals, based on the assumption that the physical activity patterns and self-perceptions related to this role of newly-retired individuals should be less entrenched than those of more “seasoned” retirees. Moreover, individuals exercising four times per week or more or reporting a strong physical activity identity were excluded as they were deemed not likely to benefit from the intervention. The average age of the participants in the sample (i.e., across conditions) was 63.4 years old. The majority of the sample was Caucasian (94 %), female (68.5 %), married (60 %), and highly educated (with almost 30 % of participants reporting having a degree higher than a bachelor’s degree). The body mass index (BMI) was calculated with self-reported weight and height data; the BMI for the whole sample was 29.74, corresponding to the overweight classification [38]. We present sociodemographic information per each condition in Table 1.Table 1 Participant demographics at randomization (N = 294) Group Characteristic Control (n = 88) One-time (n = 107) Repeated (n = 99) Age, mean (SD) 63.02 (4.84) 63.84 (4.59) 63.23 (4.71) Sexa  Women 65 (74.7) 70 (66.0) 63 (65.6)  Men 22 (25.3) 36 (34.0) 33 (34.4) Ethnicityb  Caucasian 86 (97.7) 94 (87.9) 95 (97.9)  Other 1 (1.10) 6 (5.5) 1 (1.1)  Aboriginal/Native 0 (0.00) 4 (3.7) 1 (1.00)  South Asian 1 (1.1) 3 (2.8) 0 (0.00) Marital statusc  Married 45 (52.30) 67 (63.2) 63 (64.9)  Divorced/Separated 16 (18.6) 15 (14.1) 15 (15.4)  Living w/ partner 9 (10.5) 11 (10.4) 7 (7.2)  Single/Never married 10 (11.6) 7 (6.6) 6 (6.2)  Widowed 6 (7.0) 6 (5.7) 6 (6.2) Educationd  University: above B.A. 28 (32.2) 30 (28.0) 26 (26.5)  University: B.A. 18 (20.7) 26 (24.3) 23 (23.5)  College/CEGEP 22 (25.3) 17 (15.9) 14 (14.3)  No post-secondary 8 (9.2) 19 (17.8) 17 (17.3)  Other 11 (12.6) 15 (14.0) 18 (18.3) Body mass indexe, mean (SD) 29.98 (6.37) 28.73 (5.98) 29.97 (5.63) Retirementf, y, mean (SD) 1.86 (1.15) 1.77 (1.45) 1.63 (1.38) Perceived healthg, mean (SD) 6.87 (.99) 7.09 (.88) 6.77 (.98) CFCh, mean (SD) 3.37 (.56) 3.61 (.62) 3.52 (.56) Imaging abilityi, mean (SD) 34.64 (7.06) 35.70 (6.87) 33.42 (8.87) Data are presented Count (%) unless indicated otherwise. a Five participants did not report their gender; 68.5 % of the sample was female overall. b Two participants did not report their ethnicity; 94.2 % of the sample was Caucasian overall. c Five participants did not report their marital status; 60.6 % of the sample was married overall. d Two participants did not report their education; 28.8 % of the sample had a degree higher than a BA. e Two participants did not report their height. The BMI of the sample overall was 29.74 (SD = 6.43). f The participants in the sample were retired for 1.75 years on average. g Eight participants missed one of three items which comprises the scale. Scores could range between 3 and 10. The average for the sample was 6.92 (SD = .96) overall.h Consideration of future consequences (CFC). Scores could range between 1 and 5. The average for the sample was 3.51 (SD = .59) overall. i Scores could range between 0 and 44. The average for the sample was 34.62 (SD = 7.69) overall Procedures All procedures were approved by the appropriate Research Ethics Board. In an effort to reach as many retirees as possible, we decided to deliver the intervention online through a web survey platform. In the health purview, internet-delivered interventions carry some notable advantages over their lab-based counterparts (for a discussion of these advantages, please see [39, 40]). Furthermore, a growing body of research is showing that internet-delivered interventions are acceptable and/or effective in increasing physical activity in older adults [41–46]. Participants were recruited via word-of-mouth, social media and online postings, local advertisements (i.e., public libraries and notice boards, newspaper), newsletters from organizations catering to older adults, and e-mail lists. All advertisements featured the secure URL link for the online survey platform (powered by Fluid Surveys). Interested individuals could visit the platform to determine eligibility, and if eligible, read and provide informed consent to participate in the study, and complete baseline measures. Following completion of baseline measures, participants were randomly assigned to one of three conditions: control (n = 88), one-time intervention (n = 107), and repeated intervention (n = 99). Randomization took place online via a single multiple choice question with nonsensical answer choices; each choice was tied to a different group assignment. To minimize order bias, the answer choices were shown in a random order every time the page was viewed. Follow-up data were collected at four, eight, and 12 weeks after baseline. At all times and for each follow-up time point, participants had the option of saving their unfinished questionnaire and completing it later (by e-mailing themselves a link to retrieve the questionnaire bookmarked at the last completed page). Intervention One-time possible selves intervention condition Immediately after completing baseline measures, participants randomized to the one-time possible selves intervention group completed a standard image generation task adapted from Murru and Martin Ginis [23]. Participants were invited to watch a short video embedded within the web survey platform. Before clicking on the video to start, participants were reminded to turn up the volume of their device since the video (through JavaScript) could not be stopped, paused, fast-forwarded, or played anew. Participants were also reminded of the definition of moderate-to-vigorous physical activity given its reference in the video. The video was comprised of a series of slides (i.e., Powerpoint slides saved as a Windows Media Video file) with text presented against a solid background. A narrator also read the text as follows:The following video is a very important component of the present study. We kindly ask you to watch and listen carefully. It is only a few minutes long. This video addresses how you see yourself in the future. We all think about the future to some extent. When doing so, we usually think about the kinds of experiences that are in store for us and the kinds of people we might possibly become. We are interested in your impression of yourself five (5) to ten (10) years from now. More specifically, we would like you to think about yourself in the future as a physically active retiree. You incorporate physical activity to your lifestyle on most days of the week at a moderate to vigorous intensity. Five (5) to ten (10) years from now you have the energy to carry out your daily tasks and all your personal goals for retirement. When you think about yourself five to ten years from now as a retiree who is physically active on a regular basis, what images come to mind? Please take a few minutes to imagine and think about this image before moving on to the next page. On the following pages, you will be asked to answer some questions about this image. This part of the intervention (text/narration) was 2.5 min long. Through presenting the text via video/narration we sought to solicit the participants’ attention and immersion in the image generation task (and not just a quick skimming of the text). Once the sentence Please take a few minutes to imagine and think about this image appeared on screen, participants were given two minutes to carry out this imagining/thinking task after which the screen turned black, prompting participants to move ahead to the next page. In all, the image generation task (intervention) took about 4.5 min. The task heeds recommendations advocated in other literature as per the ideal run time (i.e., five minutes) for a video component in an Internet-delivered intervention [47]. Following this task, and in line with Murru and Martin Ginis [23], participants answered seven open-ended questions to ensure elaboration upon the image (i.e., image’s appearance, energy level, attitude toward life, general health, relationships, achievements and any other comment/thought that came to mind), which also served as a manipulation check for the image generation task. Further, “compliance” checks were conducted by asking two questions about the content of the video (to confirm viewing) and by asking if (and to what extent) participants used the time provided to think about the image of themselves as a physically active retiree. Repeated possible selves intervention condition Participants randomized to the repeated intervention group completed the standard possible selves image generation task (as described above) on three separate occasions, each one week apart. The three intervention exposures were practically identical with the exception of a few minor changes made to the video (i.e., change in background colour, narrator) and to the compliance checks (different “content” questions were posed). All told, participants in the repeated condition completed the subsequent intervention exposures before the week 4 (time 2) follow-up data collection point. Control condition During the 12-week study duration, participants randomized to the control condition only completed the follow-up measures at four, eight, and 12 weeks. The participants did not receive any intervention-like materials. The “no treatment” option was predicated on arguments that finding intervention effects can be more difficult when an active comparator (i.e., some form of intervention) is given to control participants [48]. Measures Eligibility measures Eligibility was assessed and instantaneously determined online via the Fluid Surveys platform; we opted for one-item eligibility measures of physical activity and physical activity identity. The multi-item measures used during the actual intervention (described below) were not helpful for eligibility purposes because of the platform’s inability to calculate means or scores. Physical activity Prospective participants’ physical activity level was assessed via a single item consistent with Godin and colleagues’ work [49–51]. The item read as follows: How often have you participated in one or more moderate-to-vigorous physical activities for at least 30 min in one day during your free time in the last three months? Definitions of leisure time, moderate and vigorous physical activity were provided. Seven answer choices were displayed and ranged from: Never to four times or more per week. Potential participants selected one answer choice; those reporting exercising four times per week or more were excluded. Physical activity identity Prospective participants’ physical activity identity was assessed via a single item recently adapted for physical activity research by Carraro and Gaudreau [52] but originally developed by Aron, Aron, and Smollan [53]. Potential participants were shown a figure with seven pairs of increasingly intertwined circles numbered from 1 (low identification) to 7 (high identification) and were asked to select the number that corresponded to their relationship with physical activity (i.e., the extent to which they think physical activity is a part of who they are). Potential participants who selected 6 or 7 were excluded. Main measures The main measures of the study are described below. Sociodemographic information Information on age, sex, ethnicity, marital status, education level, body mass index, years in retirement, and perceived health [54] was collected and is presented in Table 1. Godin Leisure Time Exercise Questionnaire (GLTEQ) Using the GLTEQ [55] participants indicated the number of bouts of 15 min or more of physical activity performed at strenuous and moderate activity performed over a 7-day period. The GLTEQ is valid [56] and has been used in older adults [36, 57]. To encourage more accurate reporting, we included additional instructions that normalized the difficulties of being physically active and encouraged honest reporting as recommended and proven effective by Gagné and Godin [58]. For the present study, bouts of moderate and vigorous activity were summed to obtain moderate-to-vigorous physical activity bouts. Physical activity was assessed at baseline, 4, 8, and 12 weeks. Physical activity identity scale A slightly modified version of the Exercise Identity Scale (EIS; [59]) was used to assess physical activity identity. The original scale comprises nine items and uses a 7-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly agree). Sample items include: I consider myself an exerciser and When I describe myself to others, I usually include my involvement in exercise. The modifications made to the scale were consistent with Strachan et al. [36] in that the word “exercise” was exchanged for “physical activity” (e.g., I consider myself a physically active person). This change is based on research suggesting that older adults find the term “physically active person” more self-descriptive than “exerciser” [36, 60]. All other aspects of the original scale were retained. Anderson and Cychosz [59] reported a strong Cronbach’s alpha (α = .94) and test-retest reliability for the original version of this scale (r = .93). The modified scale used by Strachan et al. [36] also proved reliable (Cronbach’s α of .90). In this study, items of the scale proved reliable at all four time points – baseline, 4, 8, and 12 weeks. (α’s: .87; .88; .91; .91, respectively). Covariates Given the temporal (i.e., oriented toward the future) and imaginal (i.e., image generation task) aspects of the intervention, covariates pertaining to consideration for future consequences and imaging ability were entered in the analyses. Imaging Ability Questionnaire (IAQ) The image generation subscale of the IAQ [61] was used to assess how vividly people generate images. The full IAQ (with absorption subscale; 21 items) boasts good internal consistency (α = .93) and test-retest reliability (r = .92). The IAQ has been used in exercise imagery interventions as a control variable (e.g., [62]) and was used in this capacity in the present study. Participants indicated, on 11 items, the extent to which they can generate certain images (e.g., imagine clouds with a storm blowing up and flash of lightning) using a 5-point scale ranging from 0 (no image at all) and 4 (perfectly clear and vivid). The image generation subscale was reliable in the present sample (α = .91). Imaging ability was entered as a covariate in the analyses. Consideration of Future Consequences (CFC) The 12-item CFC scale [63] was used to measure the extent to which individuals consider the immediate and distal consequences of their actions and was included as a covariate in the analyses. The scale has been validated previously in a number of samples (α’s ranging between .80 and .86; test-retest reliability between .72 and .76). CFC has been previously shown to influence a possible selves intervention targeting physical activity [22]. This scale features a Likert-type scale which ranges from 1 (extremely uncharacteristic) to 5 (extremely characteristic). A sample item reads: I think it is important to take warnings about negative outcomes seriously even if the negative outcome will not occur for many years. The scale was reliable in the present sample (α = .78). Statistical analysis Repeated measures In order to test for significant changes in physical activity identity and physical activity between conditions and over time (baseline, 4, 8, and 12 weeks), we conducted repeated measures mixed-effects models analyses with maximum likelihood estimation using PROC MIXED of SAS for Windows version 9.3 (SAS Institute Inc., Cary NC). The decision to run these models was predicated on the nested nature of the data (i.e., time points nested within participants), which introduces within-subject correlation between repeated measures of the outcome. Mixed-effects models allow for statistical dependence between observations within subjects and incorporates it into the model estimation algorithm, thereby providing unbiased effect estimates and appropriate standard errors. Furthermore, mixed models are increasingly used in longitudinal studies as they allow the inclusion of participants with missing data [64]. As a result, all participants who, at minimum, completed the baseline measurement and the first follow-up measurement (i.e., time 2) were included in the analyses (n = 221). The three conditions (groups), time, and condition by time interaction were modeled as fixed effects. The two covariates – consideration for future consequences and imaging ability – were also entered as fixed effects. Physical activity identity was modeled as continuous using linear mixed-effects with an unstructured covariance type. Physical activity (bouts) is a count variable however, and was modeled via negative-binomial generalized linear mixed-effects with an autoregressive covariance structure. Time was modeled as a categorical effect. Various within-subjects correlation structures were explored for each model, beginning with an unstructured approach, which allows all correlations to be estimated freely. More parsimonious correlation structures were then compared via likelihood ratio tests, and on grounds of parsimony were chosen if they did not perform significantly worse than the more intensive unstructured approach. The significance level for tests was set at p < 0.05. Sample size calculations Sample size calculations were performed in G*Power [65]. A minimum sample size of 66 per group is needed to detect a small effect size (d = .20) at .80 power and at an alpha level of .05. This sample size is predicated on a repeated measured study design with three groups (i.e., control, one-time, and repeated) which are assessed at four time points (i.e., baseline, 4, 8, and 12 weeks). Given the inclusion of two covariates (i.e., CFC and imaging ability), a sample size of 80 is preferred. The present study was sufficiently powered to detect differences between groups across time. Results Retention and attrition analysis After completing eligibility, consent and baseline measures, a total of 294 participants were randomized into one of the three groups. The detailed flow chart of the study is featured in Fig. 1. We note that 21 participants chose to leave the study before completing post randomization activities related to the baseline time point. As such, 273 participants fully completed the baseline time point. At time 2 follow-up (four weeks post baseline), 237 participants continued participation and had filled out time 2 measures. At time 3 (eight weeks post baseline), 205 participants completed measures. Finally, 187 participants completed the final follow-up measures at time 4 (12 weeks post baseline). From randomization through to time 4 follow-up, we retained 63.61 % of the sample. There were no significant differences in socio-demographic variables (as itemized in Table 1) between participants who completed the entire study (n = 187) and participants who dropped out (anytime after randomization; n = 107). Attrition was not influenced by group assignment. Lastly, the completion of follow-up possible self image generation tasks for the “repeated” group was satisfactory. The second task was completed by 88.57 % of participants. The third and final task was completed by 72.86 % of participants.Fig. 1 Flow Chart of the Study. The exclusion of data for analytic purposes was predicated on participants’ reported date of retirement (i.e., exceeding three years; ten participants) and unsatisfactory completion of the possible self task/s (six participants) Manipulation checks A manipulation check was conducted on the data from participants in the two intervention conditions. The first and second authors conducted a review of participants’ answers to manipulation check questions described previously. They checked for non-responses and appropriateness of answers (i.e., comments on point). Further, the same authors looked at the “compliance” questions to determine if participants watched the video(s) and used the time provided to think about the image of themselves as a physically active retiree. These steps prompted the removal of all data for six participants due to lack of compliance with the experimental protocol. Further data inspection prompted the removal of all data for 10 participants by reason of ineligibility (i.e., years in retirement). As a result, the removal of 16 participants led to a final analytic sample of 221 participants, split across groups as follows: 76, control; 78, one-time, and, 67, repeated. Data management Prior to running the repeated measures analyses, patterns of missing data and data distribution were examined. We examined missing data against group assignment itself and all sociodemographic variables presented in Table 1. Through the expectation maximization algorithm of the missing data package in SPSS version 22, it was determined that missing data were missing completely at random (MCAR). No data distribution issues were detected. Intervention outcomes To test the hypotheses that the repeated possible selves group would report higher levels of physical activity identity and physical activity over the course of the intervention than the control group (hypothesis 1) and the one-time possible selves group (hypothesis 2), and that the one-time group would too report higher levels and greater changes in physical activity and physical activity identity over the course of the intervention than the control group (hypothesis 3), we ran mixed models analyses with restricted maximum likelihood estimates for the effects of time, group assignment, and the group*time interaction set as fixed factors. A statistically significant group*time interaction would indicate changes in outcomes over time are different in the three groups. Consideration for future consequence and imaging ability were entered as covariates. Physical activity identity For the physical activity identity outcome, an unstructured covariance matrix proved to be the best fit. No group*time interaction was found; only a time effect was detected (see Table 2). Simply put, the three groups displayed similar increases in their physical activity identity over time; the increase, however, is not attributable to group assignment (given the lack of interaction). The means along with confidence intervals for each group at every time point are also presented in Table 2. Analyses showed that CFC had a significant positive influence on the prediction of physical activity identity (p < .01; estimate = .62 [95 % CI 0.34, 0.89]). Imaging ability was not significantly related to the outcome (p = 0.57; estimate =-.07 [−0.30, 0.17]).Table 2 Means and 95 % CI for three intervention groups (analytic sample N = 221) Outcome measure Control n = 76 (mean and 95 % CI) One-time n = 78 (mean and 95 % CI) Repeated n = 67 (mean and 95 % CI) Repeated measures model PA Identity  Baseline 3.82 (3.51–4.13) 3.68 (3.40–3.97) 3.69 (3.37–4.00) ↑ over time in all groups  4 weeks 3.95 (3.66–4.25) 4.01 (3.74–4.29) 4.13 (3.83–4.43) Time (T): p <.01  8 weeks 3.96 (3.63–4.28) 4.01 (3.70–4.32) 3.89 (3.55–4.23) Group (G): p = 0.97  12 weeks 4.32 (3.99–4.64) 4.16 (3.85–4.47) 4.20 (3.86–4.53) T x G: p = 0.39 PA (bouts)  Baseline 2.41 (1.73–3.37) 3.31 (2.46–4.44) 2.87 (2.07–3.99) ↑ over time in all groups  4 weeks 3.48 (2.56–4.72) 4.87 (3.68–6.45) 4.23 (3.12–5.74) Time (T): p <.01  8 weeks 3.61 (2.64–4.94) 4.53 (3.36–6.12) 3.67 (2.62–5.13) Group (G): p = 0.34  12 weeks 3.86 (2.79–5.34) 4.29 (3.15–5.83) 3.87 (2.77–5.41) T x G: p = 0.97 PA physical activity; PA identity scores vary between 1 and 7. Covariates included: Consideration for future consequences (CFC) and imaging ability. For both PA identity and PA, CFC was significant (p <.01, p = 0.04, respectively) Physical activity For this outcome, negative binomial distribution proved the best fit for the count data (i.e., bouts of physical activity). Again, no group*time interaction was found; only a time effect was detected (see Table 2). The three groups displayed similar increases in their physical activity over time; the changes, however, are not attributable to group assignment (given the lack of interaction). The means along with confidence intervals for each group at every time point are presented in Table 2. The prediction of physical activity was also significantly influenced by CFC (p = .04; estimate = .25 [95 % CI .01, .50]). Imaging ability was not significantly related to the outcome (p = 0.07; estimate = −.19 [95 % CI -.39, .02]). Discussion The aim of this study was to determine whether a possible selves intervention would increase physical activity behaviour and identity in a group of newly-retired individuals. Using an experimental design, we compared changes in identity and physical activity across time and between three conditions: a one-time possible selves intervention, a repeated possible selves intervention, and a control group. We hypothesized that differences over time in identity and physical activity would be found between participants in the repeated possible selves group as compared to the control group (hypothesis 1) and the one-time possible selves group (hypothesis 2). Participants in the one-time group were expected to report greater differences over time in both outcomes when compared to the control group (hypothesis 3). All hypothesized group differences did not materialize. However, over time, all three groups reported marginally higher levels of identity and physical activity irrespective of group assignment. Effects of the possible selves intervention on physical activity identity While identities are known to be rather stable (e.g., [28]), other authors [11] have argued that individuals can experience shifts and changes in self-views, sometimes in a matter of days. Further, Hoyle and Sherrill [25] argue that recurrently thinking about a possible self and its consequences can bring about changes in self-representations. From this theoretical perspective, we speculated that possible selves might bring about changes in identity and that repeated possible selves activation should have the greatest impact on identity. One possible interpretation of our findings of null effects of the intervention on identity is that a physical activity possible selves intervention is not an effective way to increase identity. Considering that no other possible selves intervention, to our knowledge, has attempted to increase identity in the physical activity purview, dismissing the possibility seems premature. Perhaps a hybrid and multifaceted approach is needed to truly impact identity. We offer a few possibilities of such approaches. An imagery intervention conducted by Cooke, Duncan, Hall, and Rodgers [unpublished observations] shares with our study, the idea that the act of imagining oneself as an exerciser can be a fruitful way to increase identity. In their 36-week exercise program for female exercise initiates, individuals who underwent an exercise imagery intervention in addition to exercise reported significantly stronger exercise role identity at 9 weeks than those within an exercise plus attention control condition. The findings from this study lead to two pertinent observations. First, the guided imagery intervention used in the Cooke et al. study and our possible selves image generation task shared similarities in that both involved evoking images of the self as an exerciser. Giacobbi et al. [66] have even commented on the close ties between mental imagery and possible selves. However, the imagery intervention used by Cooke et al. (i.e., script and lab procedures) provided more elaborate and multifaceted imagery sessions (e.g., more frequent exposure; conducted within the laboratory) than the present possible selves protocol; a more elaborate possible selves image generation task may have led to increases in identity in our study. Second, the study by Cooke et al. included, in addition to the imagery intervention, a physical activity component. As noted by Cooke et al., and others (e.g., [27, 67]), physical activity participation alone may not coalesce with increased identity and the addition of the imagery component in the study by Cooke et al. appeared to help increase identity. Indeed, previous research has pointed to the likely reciprocal relationship between physical activity and identity [19, 52, 60, 68]. Increases in identity are often tied to greater behavioural output - wherein physical activity levels increased significantly. Herein, the image generation tasks were not part of an exercise intervention or program. Future research should consider whether a possible selves intervention combined with a physical activity intervention/program will yield gains in identity and which one has the greatest effect. Effects of the possible selves intervention on physical activity Markus and Nurius [15] argue that possible selves encourage behaviour change by providing motivation and guidance in the form of standards for comparison. Indeed, possible selves have been used with success in a few physical activity/exercise interventions [22–24]. We are therefore surprised that, in the present study, participants exposed to a physical activity possible selves intervention did not report more physical activity than controls. Methodological considerations offer potential explanations for why we failed to find group differences in physical activity patterns, where others did. Ouellette et al. [22] found that university students who participated in a exerciser possible selves intervention, and scored high on consideration of future consequences, increased exercise behaviour at four weeks post intervention. However, their study did not include a control group, so it is unclear if the effect is attributable to the intervention. Interestingly, without a control group, our findings would also suggest an effect of the possible selves interventions, as we too found changes in physical activity amongst all participants. Murru and Martin-Ginis [23] found that university students exposed to a possible selves image generation task exercised more at 4 and 8 weeks follow-up than those exposed to a control activity. These researchers employed a more stringent eligibility criterion than we did (i.e., less than three versus less than four 30 min bouts of physical activity per week). Our slightly more active participants had less room to increase their physical activity making it more difficult for us to detect a change due to the intervention. Recently, Gunnell, Crocker, Mack, and Zumbo [69] also failed to find a possible selves intervention effect on physical activity in a community sample in which no limits were imposed on baseline physical activity levels. These findings suggest that physical activity possible selves interventions may work best for those who are inactive. Finally, in a study by Strachan, Marcotte, Giller, Brunet and Schellenberg [24] we note that participants were instructed to read the Canadian Physical Activity Guidelines [70] prior to partaking in one of two possible selves image generation tasks. Perhaps the exposure to the physical activity guidelines (i.e., messaging), coupled with the future perspective of the PSs task (i.e., seeing oneself engaging in regular physical, five to ten years from now) highlighted participants’ discrepant physical activity behaviour (or lack thereof), which in turn galvanized their behaviour. The purported impact of being presented with physical activity guidelines is, of course, speculative. The finding that all participants reported increased physical activity suggests that we cannot attribute physical activity increases to exposure to a possible selves intervention. All participants may have been disposed to a form of measurement reactivity [71]. For example, participants likely engaged in self-monitoring through self-reporting their physical activity at several time points via a questionnaire. If participants noticed low or reduced physical activity levels, this awareness may have energized them to increase their physical activity. In a similar vein, having participants complete questionnaires about physical activity (e.g., their intentions, thoughts, etc.) may have increased their chance of performing that behaviour – something known as question-behaviour effect [72]. This effect, documented in the physical activity/exercise purview, operates by making individuals’ underlying attitudes about a particular behaviour more accessible which, in turn, can lead to the behaviour being enacted [73, 74]. Our questionnaires, asked of all participants, included such measures. Considering that we cannot attribute the sample-wide increase in physical activity to our intervention activities, we speculate that some of these measurement aspects may be responsible for the small increase in physical activity reported across our sample. When considering the null results for both identity and physical activity, we ponder whether our approach, while inspired by previous studies, was appropriate for newly retired individuals. Hooker [75] opined: “Goals in later life may be less normatively structured than early and midlife [… and as such] their (older adults’) activities are the most likely to be motivated by their own personal agendas” (p. 111). Herein, and in line with past studies targeting younger adults (e.g., [22–24]), we utilized the same, broadly-defined possible self (i.e., physically active retiree) across groups partaking in the image generation tasks. Perhaps the one-size-fits-all nature of our intervention materials lacked personal meaning, thus dissuading participants from engaging fully in the tasks. Two recent reviews on the features of effective physical activity interventions in older adults [76, 77] point to the importance of personalizing of interventions. In the possible selves purview, this suggestion could take the form of allowing participants to articulate and or revise [16] their own physically active possible self for retirement. Such an individualized approach may be appropriate for retirees for whom generalized possible selves may lack relevance. In the same vein, McDonald, O’Brien, White, and Sniehotta [78] further confirm that “routes to retirement are highly individualistic”. Engagement in physical activity may thus be influenced by personal preferences, obligations, and other competing goals. Therefore, personalized interventions may prove more helpful in getting retirees physically active. Finally, we cannot dismiss the possibility that image generation may become more difficult for older adults. In their article on motor and exercise imagery, Kalicinski and Lobinger [79] review mixed evidence linking motor imagery ability deterioration due to age. However, it is “unclear whether this deterioration is selectively influenced by age” (p. 66). For now, though, linking age and difficulty in imagery tasks in the exercise purview is quite speculative. Strengths and limitations The present study draws strength from its innovation. The study was the first, to our knowledge, to carry out a possible selves intervention in older, retired adults, in contrast to previous inquiries conducted in younger populations. Our research also proved innovative through its focus on physical activity identity in addition to physical activity, which has typically been the outcome of interest of physical activity possible selves interventions. Conducting the intervention entirely online was also novel as this, to the extent of our knowledge, was only the second time a possible selves intervention was delivered online [24]. Finally, in terms of design and analytic strengths, we assessed and controlled for personal characteristics (i.e., imaging ability, consideration for future consequences) and data were analyzed through mixed-effects modeling, which is increasingly advocated over traditional ANOVA-based techniques (e.g., [80]). The strengths, however, must be considered in light of limitations. The demographics of participants (female, Caucasian, and well-educated) limit the generalizability of our findings. Further, our eligibility criteria pertaining to physical activity and identity allowed already fairly active retirees who somewhat identify as physically active to participate in the study, making it difficult for us to find effects. Another notable limitation concerns the use of a self-report physical activity measure which was predicated on practical considerations. Objective measurement was prohibitive given the online nature of the study. Further, we opted for the GLTEQ since it was previously validated against objective measures of physical activity and proven reliable in test-retest methods (for a review of these studies, please see [81]). Finally, the GLTEQ has been used with older adult/retired samples [20;21;36;57]. While our choice to use a validated measure of physical activity addresses some of the measurement limitations noted by Barnett et al. [12], our physical activity measure is limited in that it relies on self-report [82]. Conclusion This study was the first, to our knowledge, to test a possible selves intervention targeting physical activity identity and physical activity in a sample of recent retirees. While we did not find the expected differences between groups, marginal improvements in physical activity identity and behavior were observed over time (regardless of group assignment). We opine that interventions based on self and self-perceptions - like possible selves - may be relevant for new retirees (i.e., exploring/revisiting different roles), and as such, warrant further investigation. We hope our intervention can serve as a launch pad for future intervention efforts. Abbreviations ANOVAAnalysis of variance BMIBody mass index CFCConsideration of future consequences GLTEQGodin leisure time exercise questionnaire IAQImaging ability questionnaire MCARMissing completely at random URLUniform resource locator Not applicable Funding This work was supported by internal funding received from the University of Ottawa (Research Development Program and the Faculty of Health Sciences). The funding sources had no role in any aspects of the work. Authors’ contributions MP, SS and MF conceived the idea of this study and participated in the conception and design of the study. MP collected the data. MP and BD analysed the data. MP drafted the manuscript. SS provided substantial feedback and editing to the manuscript. MF and BD were involved in revising the manuscript critically for its content. All authors read and approved the final manuscript. Authors’ information MP was a Ph.D candidate at the School of Human Kinetics at the University of Ottawa when this research was conducted. SS is an assistant professor at the Faculty of Kinesiology and Recreation Management at the University of Manitoba. MF is a full professor at the School of Human Kinetics at the University of Ottawa. BD is a biostatistician at the George and Fay Yee Centre for Healthcare Innovation at the University of Manitoba. Competing interests The authors declare no potential conflict of interest with respect to the authorship and/or publication of this article. Consent for publication Not applicable Ethics approval and consent to participate This study has received a certificate of REB ethics approval from the University of Ottawa Health Sciences and Science Research Ethics Board. Participants provided informed consent to participate in the study. ==== Refs References 1. Schrack JA Zipunnikov V Goldsmith J Bai J Simonsick EM Crainiceanu C Assessing the physical cliff: detailed quantification of age-related differences in daily patterns of physical activity J Gerontol Ser A Biol Sci Med Sci 2014 2. Paterson DH Warburton DER Physical activity and functional limitations in older adults: a systematic review related to Canada’s physical activity guidelines Int J Behav Nutr Phys Act 2010 20459782 3. Barnett I Van Sluijs EMF Ogilvie D Physical activity and transitioning to retirement: a systematic review Am J Prev Med 2012 4. Rice NE Lang IA Henley W Melzer D Baby boomers nearing retirement: the healthiest generation? Rejuvenation Res 2010 5. Berger U Der G Mutrie N Hannah MK The impact of retirement on physical activity Ageing Soc 2005 6. Chung S, Domino ME, Stearns SC, Popkin BM. Retirement and physical activity. analyses by occupation and wealth. Am J Prev Med. 2009. doi:10.1016/j.amepre.2009.01.026 7. Fonseca AM Paúl C Health and aging: does retirement transition make any difference? Rev Clin Gerontol 2004 8. Slingerland AS Van Lenthe FJ Jukema JW Kamphuis CBM Looman C Giskes K Aging, retirement, and changes in physical activity: prospective cohort findings from the GLOBE study Am J Epidemiol 2007 17420180 9. Zantinge EM Van Den Berg M Smit HA Picavet HSJ Retirement and a healthy lifestyle: opportunity or pitfall? A narrative review of the literature Eur J Public Health 2014 24148557 10. Fox KR Wilson PM Horn TS Self-perceptual systems and physical activity Advances in sport psychology 2008 Champaign Human Kinetics 49 64 11. Stein KF Markus HR The role of the self in behavioral change J Psychother Integr 1996 6 349 384 10.1037/h0101118 12. Barnett I Guell C Ogilvie D The experience of physical activity and the transition to retirement: a systematic review and integrative synthesis of qualitative and quantitative evidence Int J Behav Nutr Phys Act 2012 13. Liechty T, Yarnal C, Kerstetter D. ‘I want to do everything!’: leisure innovation among retirement-age women. Leis Stud. 2012; doi:10.1080/02614367.2011.573571. 14. Teuscher U Change and persistence of personal identities after the transition to retirement Int J Aging Hum Dev 2010 20377167 15. Markus H Nurius P Possible Selves Am Psychol 1986 16. Cross S Markus H Possible selves across the life span Hum Dev 1991 34 230 255 10.1159/000277058 17. Harju BL Reed JM Potential clinical implications of implicit and explicit attitudes within possible exercise selves schemata: a pilot study J Clin Psychol Med Settings 2003 18. Whaley DE Future-oriented self-perceptions and exercise behavior in middle-aged women J Aging Phys Act 2003 11 1 17 10.1123/japa.11.1.1 19. Whaley DE Schrider AF The process of adult exercise adherence: self-perceptions and competence Sport Psychol 2005 19 148 163 10.1123/tsp.19.2.148 20. Perras MGM Strachan SM Fortier MS Back to the future: associations between possible selves, identity, and physical activity among new retirees Act Adapt Aging 2015 21. Perras MGM Strachan SM Fortier MS Possible selves and physical activity in retirees: the mediating role of identity Research on Aging 2015 26408187 22. Ouellette JA Hessling R Gibbons FX Reis-Bergan M Gerrard M Using images to increase exercise behavior: prototypes versus possible selves Pers Soc Psychol Bull 2005 15802656 23. Murru EC Martin Ginis KA Imagining the possibilities: the effects of a possible selves intervention on self-regulatory efficacy and exercise behavior J Sport Exerc Psychol 2010 32 537 554 10.1123/jsep.32.4.537 20733212 24. Strachan SM, Marcotte MME, Giller TMT, Brunet J, Schellenberg BJI. An online intervention to increase physical activity: Self-regulatory possible selves and the moderating role of task self-efficacy. Psychol Sport Exerc [in press]; doi:10.1016/j.psychsport.2016.05.001. 25. Hoyle RH Sherrill MR Future orientation in the self-system: possible selves, self- regulation, and behavior J Pers 2006 26. Norman CC Aron A Aspects of possible self that predict motivation to achieve or avoid it J Exp Soc Psychol 2003 27. Markus H Self-schemata and processing information about the self J Pers Soc Psychol 1977 28. Burke PJ The Self: Measurement Requirements from an Interactionist Perspective Soc Psychol Q 1980 43 18 29 10.2307/3033745 29. Marsh HW Papaioannou A Theodorakis Y Causal ordering of physical self-concept and exercise behavior: reciprocal effects model and the influence of physical education teachers Health Psychol 2006 25 316 328 10.1037/0278-6133.25.3.316 16719603 30. Strachan SM Whaley D Ekkekakis P Identities, schemas, and definitions: how aspects of the self influence exercise behavior Handbook of physical activity and mental health 2013 New York Routledge 212 223 31. Estabrooks P Courneya KS Relationships among self-schema, intention, and exercise behavior J Sport Exerc Psychol 1997 19 156 168 10.1123/jsep.19.2.156 32. Petosa RL Suminski R Hortz B Predicting vigorous physical activity using social cognitive theory Am J Health Behav 2003 27 301 310 10.5993/AJHB.27.4.2 12882424 33. Strachan SM Flora PK Brawley LR Spink KS Varying the cause of a challenge to exercise identity behaviour: reactions of individuals of differing identity strength J Health Psychol 2011 16 572 583 10.1177/1359105310383602 21346019 34. Hays LM Damush TM Clark DO Relationships between exercise self-definitions and exercise participation among urban women in primary care J Cardiovasc Nurs 2005 20 9 17 10.1097/00005082-200501000-00004 15632807 35. Liechty T Dahlstrom L Sveinson K Son JS Rossow-Kimball B Canadian men’s perceptions of leisure time physical activity and the ageing body Qual Res Sport Exerc Health 2014 36. Strachan SM Brawley LR Spink K Glazebrook K Older adults’ physically-active identity: relationships between social cognitions, physical activity and satisfaction with life Psychol Sport Exerc 2010 37. Stephan Y Fouquereau E Fernandez A The relation between self-determination and retirement satisfaction among active retired individuals Int J Aging Hum Dev 2008 18507333 38. U. S. Department of Health and Human Services, National Institutes of Health, National Heart, Lung and Blood Institute. (n.d.). Classification of Overweight and Obesity by BMI, Waist Circumference, and Associated Disease Risks. Retrieved from http://www.nhlbi.nih.gov/health/educational/lose_wt/BMI/bmi_dis.htm 39. Joseph RP Durant NH Benitez TJ Pekmezi DW Internet-based physical activity interventions Am J Lifestyle Med 2014 25045343 40. Marcus BH Nigg CR Riebe D Forsyth LH Interactive communication strategies: implications for population-based physical-activity promotion Am J Prev Med 2000 41. Ammann R Vandelanotte C de Vries H Mummery WK Can a website-delivered computer-tailored physical activity intervention be acceptable, usable, and effective for older people? Health Educ Behav 2013 23077157 42. Irvine AB Gelatt VA Seeley JR Macfarlane P Gau JM Web-based intervention to promote physical activity by sedentary older adults: randomized controlled trial J Med Internet Res 2013 23470322 43. Mouton A Cloes M Efficacy of a web-based, center-based or combined physical activity intervention among older adults Health Educ Res 2015 25772974 44. Reinwand DA Schulz DN Crutzen R Kremers SP De Vries H Who follows ehealth interventions as recommended? A study of participants’ personal characteristics from the experimental arm of a randomized controlled trial J Med Internet Res 2015 45. Wijsman CA Westendorp RGJ Verhagen EALM Catt M Slagboom PE De Craen AJM Effects of a web-based intervention on physical activity and metabolism in older adults: randomized controlled trial J Med Internet Res 2013 24195965 46. Vroege DP, Wijsman CA, Broekhuizen K, De Craen AJM., Van Heemst D, Van Der Ouderaa FJG., et al. Dose–response effects of a web-based physical activity program on body composition and metabolic health in inactive older adults: additional analyses of a randomized controlled trial. J Med Internet Res 2014;16(12):e265. doi:10.2196/jmir.3643. 47. Vandelanotte C Mummery WK Qualitative and quantitative research into the development and feasibility of a video-tailored physical activity intervention Int J Behav Nutr Phys Act 2011 21718546 48. Freedland KE Mohr DC Davidson KW Schwartz JE Usual and unusual care: Existing practice control groups in randomized controlled trials of behavioral interventions Psychosom Med 2011 21536837 49. Boudreau F Godin G Participation in regular leisure-time physical activity among individuals with type 2 diabetes not meeting Canadian guidelines: the influence of intention, perceived behavioral control, and moral norm Int J Behav Med 2014 24442932 50. Gionet NJ Godin G Self-reported exercise behavior of employees: A validity study J Occup Med 1989 31 969 973 10.1097/00043764-198912000-00007 2614536 51. Godin G Jobin J Bouillon J Assessment of leisure time exercise behavior by self-report: a concurrent validity study Can J Public Health 1986 77 359 362 3791117 52. Carraro N Gaudreau P The role of implementation planning in increasing physical activity identification Am J Health Behav 2010 34 298 308 10.5993/AJHB.34.3.5 20001187 53. Aron A Aron EN Smollan d. inclusion of other in the self scale and the structure of interpersonal closeness J Pers Soc Psychol 1992 63 596 612 10.1037/0022-3514.63.4.596 54. Conn VS Older adults and exercise: path analysis of self-efficacy related constructs Nurs Res 1998 47 180 189 10.1097/00006199-199805000-00009 9610652 55. Godin G Shephard RJ A simple method to assess exercise behavior in the community Can J Appl Sport Sci 1985 10 141 146 4053261 56. Jacobs DR Ainsworth BE Hartman TJ Arthur A A simultaneous evaluation of 10 commonly used physical activity questionnaires Med Sci Sports Exerc 1993 25 81 91 10.1249/00005768-199301000-00012 8423759 57. Kim BH Newton RA Sachs ML Giacobbi PR Jr Glutting JJ The effect of guided relaxation and exercise imagery on self-reported leisure-time exercise behaviors in older adults J Aging Phys Act 2011 19 137 146 10.1123/japa.19.2.137 21558568 58. Gagné C Godin G Improving self-report measures of non-adherence to HIV medications Psychol Health 2005 59. Anderson DF Cychosz CM Development of an exercise identity scale Percept Mot Skills 1994 78 747 751 10.2466/pms.1994.78.3.747 8084685 60. Hardcastle S Taylor AH Finding an exercise identity in an older body: “it’s redefining yourself and working out who you are” Psychol Sport Exerc 2005 61. Kwekkeboom KL Measuring imaging ability: psychometric testing of the imaging ability questionnaire Res Nurs Health 2000 23 301 309 10.1002/1098-240X(200008)23:4<301::AID-NUR6>3.0.CO;2-0 10940955 62. Andersson EK Moss TP Imagery and implementation intention: A randomised controlled trial of interventions to increase exercise behaviour in the general population Psychol Sport Exerc 2011 63. Strathman A Gleicher F Boninger DS Edwards CS The consideration of future consequences: weighing immediate and distant outcomes of behavior J Pers Soc Psychol 1994 66 742 752 10.1037/0022-3514.66.4.742 64. Fitzmaurice GM Laird NM Ware JH Applied longitudinal analysis 2004 Hoboken Wiley 65. Faul F Erdfelder E Lang A-G Buchner A G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences Behav Res Methods 2007 39 175 191 10.3758/BF03193146 17695343 66. Giacobbi Jr. PR, Buman MP, Dzierzewski JM, Aiken-Morgan AT, Roberts BL, Marsiske M, et al. Content and perceived utility of mental imagery by older adults in a peer- delivered physical activity intervention. J Appl Sport Psychol. 2014. doi: 10.1080/10413200.2013.803502. 67. Markus H, Kunda Z. Stability and Malleability of the Self-Concept. J Pers Soc Psychol 1986; doi: http://dx.doi.org/10.1037/0022-3514.51.4.858 68. Cardinal BJ Cardinal MK Changes in exercise behavior and exercise identity associated with a 14-week aerobic exercise class J Sport Behav 1997 20 377 386 69. Gunnell KE Crocker PRE Mack DE Zumbo BD Higgins C The effect of a best possible physical activity self intervention on psychological need satisfaction, well-being, and physical activity Theories of psychology: Fundamentals, applications and future directions (pp. 1–13) 2015 New York Nova Publishers 45 62 70. Canadian Society for Exercise Physiology. Canadian physical activity guidelines. Canadian Society for Exercise Physiology. 2015. http://www.csep.ca/CMFiles/Guidelines/CSEP_PAGuidelines_adults_en.pdf. Accessed 26 Aug 2015. 71. French DP Sutton S Reactivity of measurement in health psychology: how much of a problem is it? What can be done about it? Br J Health Psychol 2010 72. Sprott DE Spangenberg ER Block LG Fitzsimons GJ Morwitz VG Williams P The question–behavior effect: what we know and where we go from here Social Influence 2006 73. Godin G Bélanger-Gravel A Amireault S Vohl MC Pérusse L The effect of mere- measurement of cognitions on physical activity behavior: a randomized controlled trial among overweight and obese individuals Int J Behav Nutr Phys Act 2011 21223565 74. Wood C Conner M Sandberg T Godin G Sheeran P Why does asking questions change health behaviours? The mediating role of attitude accessibility Psychol Health 2014 24245778 75. Hooker K Hess TM Blanchard-Fields F Possible selves in adulthood: Incorporating teleonomic relevance into studies of the self Social cognition and aging 1999 San Diego Academic 76. Hobbs N Godfrey A Lara J Errington L Meyer TD Rochester L Are behavioral interventions effective in increasing physical activity at 12 to 36 months in adults aged 55 to 70 years? A systematic review and meta-analysis BMC Med 2013 23506544 77. O’Brien N McDonald S Araújo-Soares V Lara J Errington L Godfrey A The features of interventions associated with long-term effectiveness of physical activity interventions in adults aged 55–70 years: a systematic review and meta-analysis Health Psychol Rev 2015 25689096 78. McDonald S, O’Brien N, White M, Sniehotta FF. Changes in physical activity during the retirement transition: A theory-based, qualitative interview study. International Journal of Behavioral Nutrition and Physical Activity. 2015;12, doi:10.1186/s12966-015-0186-4 79. Kalicinski M Lobinger BH Benefits of motor and exercise imagery for older adults J Imagery Res Sport Phys Act 2013 80. Gueorguieva R Krystal JH Move over ANOVA: Progress in Analyzing Repeated- Measures Data and Its Reflection in Papers Published in the Archives of General Psychiatry Arch Gen Psychiatry 2004 81. Kriska AM Caspersen CJ Introduction to a collection of physical activity questionnaires Med Sci Sports Exerc 1997 29 Suppl 6 5 9 10.1097/00005768-199706001-00003 82. Helmerhorst HJF Brage S Warren J Besson H Ekelund U A systematic review of reliability and objective criterion-related validity of physical activity questionnaires Int J Behav Nutr Phys Act 2012 22938557

PMC005xxxxxx/PMC5002318.txt

==== Front BMC PediatrBMC PediatrBMC Pediatrics1471-2431BioMed Central London 67810.1186/s12887-016-0678-7Study ProtocolPREMM: preterm early massage by the mother: protocol of a randomised controlled trial of massage therapy in very preterm infants Lai Melissa M. melissa.lai@uq.edu.au 123D’Acunto Giulia g.dacunto@inpe.unipi.it 4Guzzetta Andrea a.guzzetta@inpe.unipi.it 4Boyd Roslyn N. r.boyd@uq.edu.au 5Rose Stephen E. stephen.rose@csiro.au 6Fripp Jurgen jurgen.fripp@csiro.au 6Finnigan Simon finnigan@uq.edu.au 12Ngenda Naoni naoni.ngenda@health.qld.gov.au 3Love Penny penny@pennylove.com.au 12Whittingham Koa koawhittingham@uq.edu.au 5Pannek Kerstin kerstin.pannek@gmail.com 6Ware Robert S. r.ware@sph.uq.edu.au 78Colditz Paul B. +61 7 3346 6014p.colditz@uq.edu.au 1231 Perinatal Research Centre, School of Medicine, Royal Brisbane & Women’s Hospital, Brisbane, Qld Australia 2 University of Queensland Centre for Clinical Research, Level 4, Bldg 71/918, Royal Brisbane & Women’s Hospital, Brisbane, Qld Australia 3 Grantley Stable Neonatal Unit, Royal Brisbane & Women’s Hospital, Brisbane, Qld Australia 4 Stella Maris Institute, The University of Pisa, Pisa, Italy 5 Queensland Cerebral Palsy and Rehabilitation Research Centre, School of Medicine, Faculty of Medicine and Biomedical Sciences, The University of Queensland, Brisbane, Qld Australia 6 CSIRO, Australian e-Health Research Centre, Royal Brisbane and Women’s Hospital, Brisbane, Qld Australia 7 UQ Child Health Research Centre, School of Medicine, The University of Queensland, Brisbane, Qld Australia 8 School of Population Health, The University of Queensland, Brisbane, Qld Australia 27 8 2016 27 8 2016 2016 16 1 14626 6 2015 16 8 2016 © Lai et al. 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Preterm infants follow an altered neurodevelopmental trajectory compared to their term born peers as a result of the influence of early birth, and the altered environment. Infant massage in the preterm infant has shown positive effects on weight gain and reduced length of hospital stay. There is however, limited current evidence of improved neurodevelopment or improved attachment, maternal mood or anxiety. The aim of this study is to investigate the effects of infant massage performed by the mother in very preterm (VPT) infants. Effects on the infant will be assessed at the electrophysiological, neuroradiological and clinical levels.  Effects on maternal mood, anxiety and mother-infant attachment will also be measured. Methods/Design A randomised controlled trial to investigate the effect of massage therapy in VPT infants. Sixty VPT infants, born at 28 to 32 weeks and 6 days gestational age, who are stable, off supplemental oxygen therapy and have normal cranial ultrasounds will be recruited and randomised to an intervention (infant massage) group or a control (standard care) group. Ten healthy term born infants will be recruited as a reference comparison group. The intervention group will receive standardised massage therapy administered by the mother from recruitment, until term equivalent age (TEA). The control group will receive care as usual (CAU). Infants and their mothers will be assessed at baseline, TEA, 12 months and 24 months corrected age (CA), with a battery of clinical, neuroimaging and electrophysiological measures, as well as structured questionnaires, psychoanalytic observations and neurodevelopmental assessments. Discussion Optimising preterm infant neurodevelopment is a key aim of neonatal research, which could substantially improve long-term outcomes and reduce the socio-economic impact of VPT birth. This study has the potential to give insights into the mother-baby relationship and any positive effects of infant massage on neurodevelopment. An early intervention such as massage that is relatively easy to administer and could alter the trajectory of preterm infant brain development, holds potential to improve neurodevelopmental outcomes in this vulnerable population. Trial registration Australian New Zealand Clinical Trials Registry: ACTRN12612000335897. Date registered: 22/3/2012. Keywords PretermInfant massageNeurodevelopmentAttachmentissue-copyright-statement© The Author(s) 2016 ==== Body Background Over the past 50 years, there has been a progressive decrease in preterm infant mortality [1]. A lowering of the age limit of viability and an increase in preterm birth numbers has made preterm birth a significant public health issue [1]. Despite advances in technology, the number of preterm infants with neurodevelopmental compromise remains high [2]. In recent years the focus of improving preterm outcomes has shifted from increasing survival to minimising morbidity and improving neurodevelopmental outcomes [1, 3]. Long term neurodevelopmental abnormalities impact up to 50 % of these infants [4] and include motor disability (including cerebral palsy), reduced cognitive performance and behavioural problems [3]. The severity of the deficits is related to the degree of prematurity and the presence of neuroradiological injuries [3], however high rates of more subtle but nonetheless important neurodevelopmental abnormalities also occur in low-risk preterms without obvious brain injury [5, 6]. Even in late preterm infants, the effects of preterm birth on brain development are more significant and long lasting than previously thought [5, 7]. The untimely interruption of the developing fetus’ environment is thought to be a major contributor to this phenomenon [6]. Disruption of optimal fetal environment The altered neurodevelopmental trajectory observed in preterm infants results from interruption of the intrauterine environment. The neuromaturation of the cerebral cortex, initially laid down in cortical layers prior to 20 weeks gestation, is a dynamic process from 30 to 40 weeks gestation. It is during this time, the subplate, a transient population of neurons, guides the development of cortical and thalamocortical connections [8], maturing from its maximal prominence to almost complete regression. These connections are fundamental for cortical processing and cognition [8]. Between 29 and 41 weeks, total brain volume is increased nearly 3-fold, cortical gray matter volume is increased 4-fold and cerebellar volume is increased 4-fold [7]. Although some differences between term infants and preterm infants at term equivalent age are explained by the effect of complications associated with preterm birth, both the experience of a highly stressful ex-utero environment and the lack of the stimulation normally experienced in the womb, exert detrimental effects on the immature brain [6, 9]. The challenges imposed by the extra-uterine environment, disrupt the above described development of a “neurotypical” brain. Results include numerous primary medical problems experienced by preterm infants such as lung disease of prematurity, physiological instability, asphyxia, suboptimal nutrition, infection, medication side effects and hyperbilirubinaemia, each of which may have its own potentially deleterious impact on brain development [6, 10]. Environmental stressors, which include frequent noxious stimulation, excessive sound and constant light, also adversely affect the normal neurodevelopmental trajectory [6, 10]. Repetitive pain universally experienced by premature infants from frequent invasive procedures is hypothesised to cause excessive activation of central afferent pain pathways and excitotoxic damage to the developing brain [11]. While the impact of these factors and negative influence on the brain are recognised, emergent strategies to lessen harm include environmental enrichment and infant massage. Environmental enrichment Environmental enrichment and social stimulation induce experience-dependent neuroplasticity in experimental animal models [10]. Neuroplasticity is the capacity of the mammalian brain to use the activity induced by a given experience, enabling the modification of function in neuronal circuitry [12]. Early interventions based on the manipulation of the extra-uterine environment, among them infant massage therapy, have been used in preterm infants with the aim of optimising the infant’s sensory experience and improving overall functional outcome [13, 14]. Infant massage Infant massage has been investigated as a potentially effective intervention aimed at providing a form of environmental enrichment [15]. It is often combined with other forms of stimulation such as kinaesthetic stimulation (e.g. passive extension/flexion movements of the arms and legs), talking or eye contact [16]. The possibility that infant massage could provide a form of comforting touch with positive effects on growth and neurodevelopment is described in a number of studies [16, 17]. Evidence supporting positive effects of massage in preterm infants include increased weight gain [18–20], improved growth and gastrointestinal function [21, 22], improved body fat deposition [23], improved neurobehavioural outcomes [17, 24–26], pain attenuation [27, 28], reduction of infant stress and stress-related factors [29, 30], reduction of late-onset sepsis [31], improved immune system [32], reduced jaundice [33] and improved heart rate variability [34] as well as a reduction in maternal depression and anxiety [35]. Studies investigating the use of specific oils in massage versus no oil suggest improved weight gain [36–38]. Overall the evidence remains weak, mainly due to small sample sizes, heterogeneity and poor methodology in some studies. The current level of evidence does not support wider use of infant massage without further research [17, 39]. Two key aspects of infant massage as an early intervention have received little attention. The first aspect is the effect of infant massage on direct measures of brain development, such as the maturation of brain electrical activity, brain structure and the relationship to clinical neurobehaviour. A recent meta-analysis reported that the association between massage and neurobehavioural development remained elusive [40]. The second aspect is the potential to enhance the efficacy of the intervention by active involvement of the parents; in particular the mother, and what effect of actively massaging her baby may have on the mother-infant relationship. Clinical neurodevelopmental assessments Traditionally, methods of neurodevelopmental assessment of the preterm infant have included a number of clinical examinations used to evaluate neurological function and neurobehaviour [41]. Those most commonly used in the neonatal period include the Prechtl’s General Movements Assessment (GMA), Hammersmith Neonatal Neurological Examination (HNNE), Amiel-Tison Neurological Assessment at term, Neonatal Behavioural Assessment Scale (NBAS), Neurobehavioural Assessment of the Preterm Infant (NAPI) and the Neonatal intensive care unit Network Neurobehavioural Scale (NNNS) [42]. These assessments vary in the time required for training, as well as in the appropriate age of administration and scoring systems. This affects the ability to use them in clinical practice or research [42]. The Prechtl’s GMA has the greatest predictive accuracy for the diagnosis of cerebral palsy [41, 43] with a sensitivity of 98 % and specificity of 91 % [44]. The assessment is based on a global visual perception system of the quality and complexity of movements [45]. As such, it is an applicable tool to use evaluating the impact of massage. Much like the rest of the brain, considerable development of the visual system occurs in the third trimester, which increases the potential for long-term visual dysfunction in preterm infants. The Neonatal Vision Scale is an assessment that was originally developed to test vision in full term infants. It has subsequently been applied to preterm infant cohorts to reliably measure the integrity of the visual system [46, 47]. Using this scale could assist in evaluating any measureable effects of massage on the visual system. Measuring long-term neurodevelopmental outcomes has been standardised for a number of assessments for use in toddlers and young children. These assessments are particularly useful for assessing high-risk populations such as VPT infants. The Bayley Scales of Infant and Toddler Development, Third Edition (Bayley-III) is a structured instrument to assess cognitive and social-emotional development, language and motor abilities [48] and is the most widely used measure to assess neurodevelopment of VPT and very low birth weight infants in the first three years [49]. In a recent review, Mental Development Index (MDI) scores were strongly predictive of later cognitive functioning, r = 0.61 (95 % CI = 0.57-0.64) and motor scale scores were moderately predictive of later motor function, r = 0.34 (95 % CI = 0.26-0.42) [49]. In another recent study [25], 73 very low birthweight preterm infants who had been randomised to receive massage therapy or not, were followed up at 2 years of age with the Bayley Scales of Infant Development Second edition (BSID-II). Outcomes showed the Mental Development Index (MDI) was higher in the intervention group than the control group indicating better cognitive scores in the group who received massage [25]. In the present study, we will use Bayley-III to review these findings. Magnetic Resonance Imaging to assess structure Magnetic Resonance Imaging (MRI) has become an essential neurodiagnostic tool as it offers high resolution images and can assist prognostication following neonatal brain injury [50]. MRI studies in preterm infants at term-equivalent age have shown that preterm birth alters development of regional brain volume [51], white matter [52–54], cortex [55, 56], deep gray matter [57, 58] and vascular organisation [59]. With diffusion tensor imaging, white matter integrity and white matter maturation can be studied, and white matter pathways can be non-invasively delineated through diffusion tractography [60–62]. Numerous studies have evaluated the ability of MRI at term equivalent age to predict neurodevelopmental outcomes at 1 to 9 years and established it as the best imaging tool available for outcome prediction of children born preterm [63]. To our knowledge, MRI studies of infants who have received massage therapy are yet to be described. Electroencephalography to assess function Electrophysiological studies have reported significant differences in spectral electroencephalography (EEG) measures between healthy term and preterm infants [64–66]. Maturation of the EEG in preterm infants is characterised by decreases in the total amplitude and delta activity power in quiet and active sleep [67]. A preliminary study examining whether preterm infant massage has a beneficial effect on cerebral function as measured by EEG, found a significant difference in the interburst interval duration and also a difference in maturation of visual function in preterm infants who received massage in comparison to preterm infants who received standard care [68]. A subsequent study showed a relative increase in global EEG spectral power in delta and beta frequencies in massaged infants when compared to controls, which was interpreted as suggesting that massage therapy in low-risk preterm infants favors a process of maturation of brain electrical activity similar to that observed in term born infants [24]. Parental stress, infant attachment and related assessments The stress and trauma of VPT birth on the parents is well described [69–72]. Preterm delivery has been identified as a risk factor for stress, postpartum posttraumatic stress disorder, postpartum depression and difficulty with initial bonding and attachment [73]. The increased levels of stress, anxiety and depression may negatively influence the already difficult maternal-infant bonding [73]. Historically, a high dependence on technology for life-support, the institutionalisation of preterm infant care and the fear of infection have often resulted in the separation of mother and baby [74]. In the early 1970s, research demonstrated that mothers who were permitted to enter the nursery showed a greater commitment to their infants, were more confident in their mothering abilities and had increased caretaking skills [74]. In response to this and subsequent evidence, parental involvement was encouraged [74]. An increase in bonding and attachment behaviours and a decrease in parental depression has been reported in studies where mothers attended massage classes 2 months after birth at term [75]. In the same way, actively providing infant massage could help to mitigate the impact of stress and anxiety in the intensive care nursery environment, by allowing mothers to be more proactive in the developmental care of their babies. A variety of maternal factors including frequency of maternal touch and the degree of postpartum depression (PPD) can influence the neurodevelopmental and cognitive skills of the preterm infant [76, 77]. Maternal mood and anxiety have been measured using a number of scales. The most widely researched is the Edinburgh Postnatal Depression Scale [78] which demonstrates moderate to good internal consistency over several studies. Other measures such as the Postpartum Depression Screening Scale (PDSS) and the Beck Depression Inventory (BDI) are also widely used and demonstrate good concurrent validity [78]. The Depression Anxiety Stress Scale (DASS) is another widely studied assessment tool in the postnatal period. It is a 42-item questionnaire, completed by the mother, designed to measure the magnitude of these three negative emotional states [79, 80]. The Depression scale focuses on reports of low mood, motivation and self-esteem, the Anxiety scale assesses physiological arousal, perceived panic and fear, and the Stress scale measures tension and irritability [79, 80]. Internal consistency for each of the subscales is typically high (e.g. Cronbach’s α of 0.96-0.97, 0.84-0.92, 0.90-0.95 for Depression, Anxiety and Stress respectively) [80] and convergent and discriminant validity has been demonstrated [81]. Attachment difficulties can arise from preterm birth disrupting normal physical contact between the mother and infant [82] and withdrawal of the mother from their infant due to distress [83]. This may inhibit the mother’s caregiving behaviour as her desire and ability to provide protection for her preterm infant is interrupted, ultimately impacting the mother’s attachment representations and the child’s attachment patterns [84]. Infant Observation is a method developed by Esther Bick at the Tavistock Clinic in London more than 60 years ago, and has been used to understand the characteristics of the developing relationship between mother and infant [85]. This method has been widely used internationally for training and research in the psychodynamic-psychoanalytic field to capture the rich and complex nature of mother-infant attachment. One of these complexities in particular is that of maternal responsiveness. Maternal responsiveness plays an important role in developing secure attachment and effective bonding patterns between mother and infant. Ineffective maternal responsiveness is directly related to a lack of attachment, low self-esteem and unhealthy growth and development of the child [82]. The Maternal Infant Responsiveness Instrument (MIRI) was designed to measure this concept by appraising the maternal awareness and reflection of the mother’s responsiveness to her infant and her recognition of her infant’s responsiveness to her. It is easy to administer and has an alpha reliability coefficient of 0.86 [82]. The Maternal Postnatal Attachment Scale (MPAS) is designed to evaluate maternal emotions and cognitions relating to attachment. It is specifically designed for use during the first year of life and focuses on the mother’s subjective experience in relation to her infant. Internal consistency as measured by Cronbach’s coefficient alpha is 0.78. In addition to maternal attachment, this research is also aimed to explore infant mental health. Evaluating the infant’s social and emotional functioning can be an indirect measure of the integrity of infant attachment. It has been observed that secure infant attachment status is related to lower risk for later peer interaction and behavioral problems [86]. The Infant Toddler Social and Emotional Assessment (ITSEA) is a robust parent-report questionnaire designed to assess a wide array of social-emotional and behavioural problems and competencies [87]. Its psychometric properties are sound with strong test-retest reliability (0.61-0.91, mean = 0.79), with concurrent and discriminant validities (α = 0.69-0.86, mean = 0.76) [87]. In this project we will investigate several key aspects of the effect of infant massage on preterm neurodevelopment, its neurobiological correlates and the mother-infant relationship. Aim The broad aim is to investigate the potential effects of infant massage performed by the mother, in VPT infants born between 28 and 32 weeks and 6 days gestational age. Effects on the infant will be assessed at the electrophysiological, neuroradiological and clinical levels at term equivalent age and 24 months corrected age. In addition, the impact of infant massage implemented by the mother on maternal mood and anxiety and infant attachment will be measured at term equivalent age, 12 months corrected and 24 months corrected age. Hypothesis The primary hypothesis to be tested is that infant massage by the mother in VPT infants promotes favourable processes in brain development, which can be functionally measured with dense array electroencephalography (dEEG). Secondary hypotheses to be tested include the following:Infant massage by the mother in VPT infants promotes favourable processes in brain development, which can be detected structurally at term equivalent age, and functionally at term equivalent age and 24 months corrected age. Infant massage by the mother in VPT infants reduces maternal depression, anxiety and stress associated with preterm birth. Infant massage by the mother in VPT infants enhances infant attachment. Methods Design A randomised controlled trial will be conducted to investigate the effect of infant massage. Infants will be recruited and randomised to an intervention group (PREMM), which will receive infant massage by the mother or a control group, which will receive care as usual (CAU). A study coordinator, who is not involved in generating the randomisation sequence, will assess infants for eligibility and obtain written informed consent from the parent(s). Participants Preterm participants Preterm infants born between 28 weeks and 32 weeks and 6 days gestational age, admitted to the Grantley Stable Neonatal Unit at the Royal Brisbane & Women’s Hospital will be assessed for eligibility. Parents will be approached for consent if a head ultrasound around day 10 after birth confirms the absence of moderate or severe intraventricular haemorrhage (> grade II) [88], periventricular echogenicity or periventricular cysts. Once written informed consent is obtained, baseline measures will be collected before allocation occurs. Allocation will be performed by opening the next, in sequence, opaque envelope by non-study personnel. In this way infants will be randomised into a massage group or a care as usual group. Inclusion criteria Infants born between 28 and 32 weeks and 6 days gestational age, with a birthweight between the 10th and 90th percentile (for their gestational age and gender), who are clinically stable, off oxygen therapy or respiratory support and have no IVH greater than grade II. The cut-off age for recruitment is 34 weeks and 3 days old, to ensure ample time for infants in the massage group to receive the intervention. Exclusion criteria The presence of abnormalities on brain ultrasound including intraventricular haemorrhage (grade III and IV), periventricular echogenicity or periventricular cysts. Infants with major genetic disorders and malformations will also be excluded. Sample size Preliminary data from a pilot study, which described global spectral EEG power-associated brain maturation in massaged preterm infants, was used to calculate a sample size [24]. To find a 10 % between-group difference in EEG power in the treatment group [24], with type 1 error rate 0.05, and power of 80 %, we require 20 participants in each group to complete the study. Assuming that two thirds of participants will remain in the study at term equivalent age, we will require a total of 60 preterm infants to be recruited (i.e. 30 per group). Randomisation Randomisation will be performed in Statistics Package for the Social Sciences (SPSS) 15.0 (SPSS Inc., Chicago, IL, USA) using random number generation with stratification by gender. The ratio of intervention to care as usual participants will be 1:1. A third party not involved with recruitment will generate the randomisation list and the list will be concealed. After random allocation lists have been generated, the allocation group (intervention or care as usual) of the premature infants will be stored in sequentially numbered, sealed, opaque envelopes. These envelopes will be opened by non-study personnel at each study enrolment, after baseline measures have been obtained. Intervention The intervention will be introduced to the mother following allocation into the massage group. The massage paradigm is modified from the preterm massage protocol of Field et al. [89] that consists of 2 parts, a tactile phase and a kinaesthetic phase. Mothers will be taught the techniques and encouraged to massage their babies for 15 min, twice a day until term equivalent age. Those allocated to the control group will receive care as usual. Mothers will be given the following instructions both in demonstration and in written form for later reference. The tactile phase will begin with asking “baby’s permission”, in the prone position and enclosing the infant with cupped hands in a “resting hands” position (Fig. 1). A small amount of massage oil (cold-pressed fruit or vegetable oil) will be used. Maintaining a “resting hand” to the back, the mother will slowly stroke the baby with the palm of her other hand with a gentle but firm, rhythmic motion, following this sequence: head to the neck (6 times), shoulder to the hand (6 times) left and right, neck to the back to the bottom (6 times), bottom to the feet (6 times) left and right. After returning to “resting hands” position for up to a minute, this sequence will be repeated. During the massage the mother is to breathe deeply and slowly, keeping herself relaxed whilst paying attention to the baby’s stress cues which include crying, skin mottling, hiccups, gagging, apnoea and bradycardia. If signs of stress are evident, the mother will to return to “resting hands” until the baby settles and is ready for further massage. If the baby doesn’t settle, the mother will stop and re-attempt the massage later.Fig. 1 "Resting hands" prone position in massage protocol For the kinaesthetic phase, the baby is placed in the supine position. Again, the mother will place her palms on the baby’s body to embrace the chest in a “resting hands” position. If the baby’s eyes are open, eye contact and verbal interaction will be encouraged. Kinaesthetic stimulation will follow this sequence: 6 flexions and extensions of the right upper limb, 6 flexions and extensions of the left upper limb, 6 flexions and extensions of the right and left upper limbs together, 6 flexions and extensions of the right lower limb, 6 flexions and extensions of the left lower limb, 6 flexions and extensions of the right and left lower limbs together. After returning to “resting hands” position for up to a minute, this concludes the session. The massage sequence will be taught to the mother by the same paediatric physiotherapist trained in the Field method of massage [89]. The mothers will be provided with a structured diary to keep a record of the number and duration of each session. Massage sessions will be preferably performed around 60 min before feeding and at least 2 h after the completion of the previous session. Up to three individual teaching sessions will be given to each mother and support will be continued until the mother is considered to be competent and confident. Regular contact with the mother will ensure that she continues to use the outlined protocol. When the baby is discharged home, the mother will be asked to continue daily massages until TEA. Infants of the CAU group will undergo routine nursery care and there will be no specific interventions. Intervention and CAU mothers will be assigned to different nursery rooms to prevent contamination. In addition, nursery staff will be briefed on the importance of maintaining the study protocol. Very close monitoring of the data collection process will ensure protocol violations will be identified and action will be undertaken to prevent this, in a sensitive manner. Term participants A term born reference group will be recruited from the postnatal wards or via interested parents for comparison to a normal infant population. The term born reference group will be born between 37 and 41 weeks gestation following an uncomplicated pregnancy and delivery, have a birthweight > 10th centile and have not required special care admission. Measures Baseline measures Demographics: Medical risk factors for outcome using standardised Australian and New Zealand Neonatal Network (ANZNN) data definitions [90] (e.g. gestational age, birthweight) will be collected from the baby’s case notes as baseline demographics. Prechtl’s Qualitative Assessment of General Movement [45]: Video of the baby’s general movements will be collected for general movements assessment performed at a later date by a certified assessor who will be blind to group allocation. Hammersmith Neonatal Neurological Examination (HNNE): The HNNE is a discriminative and predictive scale for the neurological examination of the preterm and term newborn [91–93]. It is derived from a combination and adaptation of items taken from Prechtl, Saint Anne Dargassies and Brazelton methods in a simplified and easy to administer protocol [92]. When compared to term born infants, preterm infants at term equivalent age have lower scores when assessed with HNNE [92]. The examination will be performed by a Paediatric Neurologist, Paediatric Physiotherapist or Neonatologist who will score the infant at the same time. Depression Anxiety Stress Scales (DASS): This self-report measure consists of three 14-item scales assessing depression, anxiety and stress over the past week. Each item has 4 response options ranging from 0 (“Did not apply to me at all”) to 3 (“Applied to me most of the time”) [94]. If questionnaire scores indicate significant emotional stress, psychosocial support will be offered. Mothers will be asked to complete this assessment. Edinburgh Postnatal Depression Scale (EPDS): The EPDS is a 10-item self-administered questionnaire developed for screening postpartum depression in women in outpatient or home visiting settings [95]. Psychosocial support will be offered and appropriate referrals made if the EPDS score is greater than 9. Mothers will be asked to complete this assessment. Mother-to-Infant Bonding Scale (MIBS): The MIBS is designed for use from day one postpartum, offering the mother one-word descriptors of possible emotions towards her new child. It is quick and easy to use and has reasonable reliability (α score 0.71) [96]. Mothers will be asked to complete this assessment. Infant Observations: These are conducted on mother-infant pairs, each of 1 h duration by one Psychoanalytic Psychotherapist trained in the Esther Bick Tavistock model of observation [85]. Each observation will be assessed using the following criteria: amount of touching by the mother, amount of visual checking if the mother is away from the infant or if the infant is asleep, amount of auditory checking, amount of talking about the infant by the mother, the infant's desire for contact with the mother, visual/physical reaching, anxiety apparent in the relationship, reports of illness or difficulties, willingness to let the observer see the infant. This will be the first of four such observations to evaluate the evolving mother-infant relationship. Preterm infants timing of assessments All preterm infants will be asked to return to hospital for assessment at term equivalent age (39 to 42 weeks post menstrual age (PMA)), then further assessments at 12 and 24 months corrected age either at home or at the hospital (Table 1).Table 1 Outcome measures for preterm infants Measures for Preterm infants Baseline Term Equivalent Age 12mth corrected 24mth corrected Clinical Demographics HNNE GMA HNNE GMA NVS Bayley-III Infant Observation IO IO IO IO Questionnaires DASS MIBS EPDS DASS MIBS EPDS MIRI MPAS DASS ITSEA MSES Radiological MRI Electrophysiological dEEG HNNE = Hammersmith Neonatal Neurological Examination; GMA = General Movements Assessment; NVS = Neonatal Vision Scale; Bayley-III = Bayley Scales of Infant and Toddler Development, 3rd Ed; IO = Infant Observation; DASS = Depression, Anxiety, Stress Scale; MIBS = Mother to Infant Bonding Scale; EPDS = Edinburgh Postnatal Depression Scale; MIRI = Maternal Infant Responsiveness Instrument; MPAS = Maternal Postnatal Attachment Scale; ITSEA = Infant Toddler Social and Emotional Assessment; MSES = Maternal Self Efficacy Scale; MRI = Magnetic Resonance Imaging; dEEG = dense array electroencephalography Outcome measures Primary outcome The primary outcome measure is beta-frequency global power (EEG parameter) in preterm infants, at term equivalent age, who have received massage therapy or care as usual. High-density electroencephalography (dEEG) will be recorded using a 64-electrode sensor net (HydroCel Geodesic Sensor Net, Electrical Geodesics Inc.). Each electrode is a saline sponge, in a geodesic tension structure comprised of silastic threads. The sensor net is prepared by soaking in a normal saline solution and then applied directly onto the head without need for further scalp preparation. The infant will be fed, wrapped and placed in an open cot, in a darkened room fitted with a Faraday cage. The recording will be conducted during sleep. EEG signals are amplified by a GES 300 series amplifier (Electrical Geodesics Inc.), digitised (at a sampling rate of 256 Hz) and recorded to the hard drive of a Mac desktop computer via NetStation software (Electrical Geodesics Inc.). The EEG data files will be exported from NetStation software then pre-processed and analysed quantitatively using Curry Scan 7 Neuroimaging Suite signal processing software (Compumedics Neuroscan™, Compumedics Limited, Australia). Secondary outcomes i. Other EEG parameters at term equivalent age (TEA) Global power in delta, theta and alpha bands, absolute and relative power for each frequency band, interhemispheric coherence and partial directed coherence will be analysed. ii. MRI measures at TEA A brain MRI will be performed using a 3.0-T (Siemens TIM Trio, Erlangen, Germany) and an MRI-compatible incubator with a dedicated head coil (LMT Lammers Medical Technology, Lubeck, Germany). The infants will be fed, fitted with earmuffs to minimise noise exposure (Natus Mini Muffs, Natus Medical Inc., San Carlos, CA) and monitored by pulse oximetry, then wrapped and placed in an MRI-compatible incubator to keep the infant still, warm and supported in the scanner. Scanning will occur for about 45 to 60 min without sedation. The MRI protocol will include T1 turbo spin echo (TSE), T1w MPRage, T2w HASTE and 3 echo T2 map, 30 direction diffusion weighted imaging (DWI), and 64 direction DWI sequences. A neuroradiologist will review clinical sequences and classify any white and grey matter [58, 97] abnormalities. Diffusion images will be acquired using single-shot echo planar multi-direction diffusion-weighted sequence, employing dual bipolar diffusion gradient and double spin echo. This will include the acquisition of a 30 direction DWI protocol (b = 1000s/mm2) and a 64 direction HARDI protocol (b = 2000s /mm2), in which the encoding gradients are uniformly distributed in space (acquisition time 5 min and 11 min respectively). A field map for diffusion data is acquired using two 2D gradient recalled echo images to assist in correction for residual distortions due to susceptibility in homogeneities (acquisition time 1 min). An extensive pre-processing and quality control procedure will be used to detect and correct image artefacts caused by head movement, cardiac pulsation, and image distortions [60]. Fractional anisotropy (FA) and mean diffusivity (MD) will be estimated from corrected diffusion data using a diffusion tensor model. Constrained spherical deconvolution implemented in MRtrix will be employed to estimate fibre orientation distribution (FOD) [98]. Whole-brain voxel based analysis of FA and MD will be performed using tract-based spatial statistics optimised for neonates [99]. Whole-brain voxel-based analysis of fibre orientation distributions will be conducted using Apparent Fibre Density (AFD) [100]. Probabilistic tractography will be performed using MRtrix. White matter pathways will be delineated using the multi-regions-of-interest approach. A number of pathways, including corticospinal tract, corpus callosum, superior longitudinal fasciculus and thalamic radiations will be extracted. Summary measures of FA, MD, AFD and T2 within pathways will be calculated and related to EEG and clinical findings. iii. Preterm infant clinical measures8. Clinical assessments performed at baseline will also be performed at TEA; Prechtl’s Qualitative General Movements Assessment (GMA) [45] and Hammersmith Neonatal Neurological Examination (HNNE) [92]. All examinations will be performed by a Paediatric Neurologist, Paediatric Physiotherapist or Neonatologist blinded to allocation groups. A visual assessment will be performed at TEA. The Neonatal Vision Scale (NVS) [47] is a clinical assessment consisting of a short test battery that explores various aspects of visual function, ranging from the ability to fix and follow a target, to more complex aspects of visual function, such as reaction to a colour target, discrimination of black and white stripes with increasing spatial frequency and attention at a distance [47]. The Bayley Scales of Infant and Toddler Development, Third Edition (Bayley-III) will be performed [48] at 24 months corrected age. Children will be assessed in the 5 key developmental domains of cognition, language, social-emotional, motor and adaptive behavior [48]. The Infant Toddler Social and Emotional Assessment (ITSEA) will be measured at 24 months corrected age. The ITSEA is a 136-item parent report questionnaire to assess social and emotional problems and competencies in 4 domains of behavior: behavioural dysregulation; externalising behavior problems; internalising behavior problems and competencies [87]. iv. Maternal measuresInfant Observation [85] performed at baseline will be repeated at TEA, 12 months and 24 months corrected age. This will be performed by the Psychoanalytic Psychotherapist who is blinded to randomisation. Questionnaires performed at baseline will also be performed at TEA; Depression Anxiety Stress Scales [79], Edinburgh Postnatal Depression Scale [95] and Mother-to-infant Bonding Assessment [96]. Mothers will be asked to complete these assessments. At 12 months corrected age, maternal infant responsiveness and maternal postnatal attachment will be measured. The MIRI, a 22-item scale [82] will be performed. The MPAS which has 19 items, all scored on a 5-point scale, with 1 and 5 indicating low and high attachment respectively, will also be performed [101]. Mothers will be asked to complete this assessment. At 24 months corrected age, the Maternal Self Efficacy Scale (MSES) will be measured. The MSES is a 20-item measure of the mothers’ perceived self-competence of their maternal practice used at 12 months. The measure has good internal consistency (Cronbach’s α = 0.76-0.89) and is a good predictor of maternal competence, with strong concurrent validity with observation [102]. Mothers will be asked to complete this assessment. Healthy term born infants Once written informed consent has been obtained, an appointment will be arranged for the infants to undertake the same assessments as the preterm cohort, at term equivalent age (39 to 42 weeks postmenstrual age) and infant observations at 12 months of age (Table 2).Table 2 Outcome measures for term born infants Measures for Term born infants 42 weeks post menstrual age 12 months corrected Clinical HNNE GMA NVS Infant Observation IO IO Questionnaires DASS MIBS EPDS Radiological MRI Electrophysiological dEEG HNNE = Hammersmith Neonatal Neurological Examination; GMA = General Movements Assessment; NVS = Neonatal Vision Scale; IO = Infant Observation; DASS = Depression, Anxiety, Stress Scale; MIBS = Mother to Infant Bonding Scale; EPDS = Edinburgh Postnatal Depression Scale; MRI = Magnetic Resonance Imaging; dEEG = dense array electroencephalography Blinding Clinical assessors will be blinded to the intervention allocation. Likewise MRI and EEG data analyses will be performed in a blinded fashion. Potential confounders Limitations include the single blinded nature of the study and the risk of potential contamination between intervention and CAU groups. Parents whose infants are allocated to the control group may be more likely to withdraw from the study after randomisation, if they perceive there is limited benefit to them. Data analysis Summary statistics will be used to describe demographic and clinical data characteristics at baseline by allocated study treatment. Continuous data will be summarised using either mean and standard deviation, or median and inter-quartile range, depending on the distribution of the variable of interest. Categorical data will be presented as frequencies and percentages. Comparisons between the baseline values of the treatment groups will be conducted to assess the degree to which variables are comparable after randomisation. Between group differences will be investigated using linear regression for continuous data and Fisher’s exact test for categorical data. Potential confounding variables for the primary outcome are considered a-priori to be: gestational age at birth, birthweight, parity and maternal education. If any of these variables differ significantly between-groups (at P < 0.005), the identified variable(s) will be controlled for in the main analyses. The primary study outcome is beta frequency global power measured in all VPT infants at term equivalent age. The mean difference between treatment groups will be calculated using linear regression with treatment group entered as the main effect. The corresponding 95 % Wald confidence interval and p-value will be reported. For secondary outcomes, we will present the effect estimate relating to a variable with a continuous outcome as a mean difference, which will be calculated using a linear regression model. If a continuous variable does not meet the assumptions necessary for linear regression, we will compare groups using the Mann–Whitney U Test. We will present the effect estimate relating to a variable with a binary outcome as an odds ratio, which will be calculated using a logistic regression model. We will present the effect estimate relating to a variable with a count outcome as an incident rate ratio, which will be calculated using a Poisson regression model. For all models the corresponding 95 % Wald confidence interval and p-value will be reported. All analyses will primarily be undertaken using the ‘intention-to-treat’ approach, where all evaluable data is analysed in the treatment group according to which the participant was allocated, regardless of treatment received. Additionally, ‘per protocol’ analyses will be conducted where there have been large deviations from the planned intervention; all individuals with evaluable outcome data will be analysed according to the treatment they received (no massage/received < 50 % of total possible massage/received at least 50 % of total possible massage). Any such analyses will be clearly labeled as such and cautiously interpreted as perhaps indicating the maximum theoretical potential of the intervention. Statistical significance will be defined as alpha = 0.05. All tests conducted will be two-tailed. Discussion This protocol paper outlines a randomised controlled trial of infant massage in VPT infants to detect effects on neurodevelopment. To our knowledge this study is the first to directly measure what influences infant massage may have on preterm brain structure and function using EEG, MRI and neurobehavioural assessments, and the mother-infant relationship in a preterm cohort. Abbreviations AFDApparent fibre density Bayley-IIIBayley Scales of Infant and Toddler Development, Third Edition BDIBeck Depression Inventory BSID-IIBayley Scales of Infant Development, Second Edition DASSDepression Anxiety Stress Scale dEEGDense array electroencephalography DWIDiffusion weighted imaging EEGElectroencephalography EPDSEdinburgh Postnatal Depression Scale FAFractional anisotropy FODFibre orientation distribution GMAGeneral Movements Assessment HARDIHigh angular resolution diffusion imaging HNNEHammersmith Neonatal Neurological Examination HRECHuman research ethics committee ITSEAInfant Toddler Social and Emotional Assessment MDMean diffusivity MDIMental Developmental Index MIBSMother-to-Infant Bonding Scale MIRIMaternal Infant Responsiveness Instrument MPASMaternal Postnatal Attachment Scale MRIMagnetic Resonance Imaging MSESMaternal Self-Efficacy Scale NVSNeonatal Vision Scale PDSSPostpartum Depression Screening Scale PMAPostmenstrual age PPDPostpartum depression SPSSStatistical Package for the Social Sciences TEATerm equivalent age TSETurbo spin echo VPTVery preterm infant Acknowledgements We would like to acknowledge allied health clinicians Sonia Sam, Joanne George, Bernadette Shannon & PhD student Annice Kong for their enthusiastic assistance with assessments and acquisition of the clinical data. We would also like to acknowledge Dr Joel Dulhunty who assisted in generation of the randomisation list. Funding QCPRRC funding project grant Merchant Charitable Foundation Postgraduate Research Scholarship (ML) Availability of data and material Data and materials will be made available once the study results have been published but restrictions will apply to prevent compromise of individual privacy. Authors’ contributions Melissa M Lai and Giulia D'Acunto are equal first authors of this paper. MML, JF, RNB, PBC, KW and PL prepared the protocol. GD, AG, RNB, PBC, NN, SER, KP, SF, KW, PL and RSW contributed to protocol design and development. Competing interests The authors declare that they have no competing interests. Consent for publication Consent to publish the images contained in this document was obtained from the parents or legal guardians. Ethics approval and consent to participate The study was approved by the Human Research Ethics Committees (HREC) at the Royal Brisbane & Women’s Hospital (HREC/09/QRBW/296), The Children’s Health Services Queensland (HREC/12/QRCH/40) and The University of Queensland (2014001160). ==== Refs References 1. Allen MC Cristofalo EA Kim C Outcomes of preterm infants: morbidity replaces mortality Clin Perinatol 2011 38 3 441 54 10.1016/j.clp.2011.06.011 21890018 2. Bonifacio SL Glass HC Peloquin S Ferriero DM A new neurological focus in neonatal intensive care Nat Rev Neurol 2011 7 9 485 94 10.1038/nrneurol.2011.119 21808297 3. Saigal S Doyle LW An overview of mortality and sequelae of preterm birth from infancy to adulthood Lancet 2008 371 9608 261 9 10.1016/S0140-6736(08)60136-1 18207020 4. Latal B Prediction of neurodevelopmental outcome after preterm birth Pediatr Neurol 2009 40 6 413 9 10.1016/j.pediatrneurol.2009.01.008 19433273 5. Vohr B Long-term outcomes of moderately preterm, late preterm, and early term infants Clin Perinatol 2013 40 4 739 51 10.1016/j.clp.2013.07.006 24182959 6. Bennet L Van Den Heuij L Dean JM Drury P Wassink G Gunn AJ Neural plasticity and the Kennard principle: Does it work for the preterm brain? Clin Exp Pharmacol Physiol 2013 40 11 774 84 10.1111/1440-1681.12135 23735123 7. Baron IS Litman FR Ahronovich MD Baker R Late preterm birth: a review of medical and neuropsychological childhood outcomes Neuropsychol Rev 2012 22 4 438 50 10.1007/s11065-012-9210-5 22869055 8. Gatti MG Becucci E Fargnoli F Fagioli M Aden U Buonocore G Functional maturation of neocortex: a base of viability J Matern Fetal Neonatal Med 2012 25 Suppl 1 101 3 10.3109/14767058.2012.664351 22393933 9. Field T Massage therapy Med Clin N Am 2002 86 1 163 71 10.1016/S0025-7125(03)00078-6 11795087 10. Perlman JM Cognitive and behavioral deficits in premature graduates of intensive care Clin Perinatol 2002 29 4 779 97 10.1016/S0095-5108(02)00051-9 12516746 11. Anand KJ Scalzo FM Can adverse neonatal experiences alter brain development and subsequent behavior? Biol Neonate 2000 77 2 69 82 10.1159/000014197 10657682 12. Kalia M Brain development: anatomy, connectivity, adaptive plasticity, and toxicity Metabolism 2008 57 Suppl 2 S2 5 10.1016/j.metabol.2008.07.009 18803960 13. Symington A, Pinelli J. Developmental care for promoting development and preventing morbidity in preterm infants. Cochrane Database Syst Rev. 2006 (Issue 2):CD001814. 14. Lubbe W Van der Walt CS Klopper HC Integrative literature review defining evidence-based neurodevelopmental supportive care of the preterm infant J Perinatal Neonatal Nurs 2012 26 3 251 9 10.1097/JPN.0b013e3182650b7e 15. Cioni G D'Acunto G Guzzetta A Perinatal brain damage in children: neuroplasticity, early intervention, and molecular mechanisms of recovery Prog Brain Res 2011 189 139 54 10.1016/B978-0-444-53884-0.00022-1 21489387 16. Vickers A, Ohlsson A, Lacy JB, Horsley A. Massage for promoting growth and development of preterm and/or low birth-weight infants. Cochrane Database Syst Rev. 2004 (Issue 2):CD000390. 17. Smith JR Comforting touch in the very preterm hospitalized infant: an integrative review Adv Neonatal Care 2012 12 6 349 65 10.1097/ANC.0b013e31826093ee 23187642 18. Massaro AN Hammad TA Jazzo B Aly H Massage with kinesthetic stimulation improves weight gain in preterm infants J Perinatol 2009 29 5 352 7 10.1038/jp.2008.230 19148112 19. Field T Diego M Hernandez-Reif M Dieter JN Kumar AM Schanberg S Insulin and insulin-like growth factor-1 increased in preterm neonates following massage therapy J Dev Behav Pediatr 2008 29 6 463 6 10.1097/DBP.0b013e3181856d3b 18714203 20. Diego MA Field T Hernandez-Reif M Preterm infant weight gain is increased by massage therapy and exercise via different underlying mechanisms Early Hum Dev 2014 90 3 137 40 10.1016/j.earlhumdev.2014.01.009 24480603 21. Choi H, Kim SJ, Oh J, Lee MN, Kim S, Kang KA. The effects of massage therapy on physical growth and gastrointestinal function in premature infants: A pilot study. J Child Health Care. 2016;20(3):394–404. 22. Tekgunduz KS Gurol A Apay SE Caner I Effect of abdomen massage for prevention of feeding intolerance in preterm infants Ital J Pediatr 2014 40 89 10.1186/s13052-014-0089-z 25394549 23. Moyer-Mileur LJ Haley S Slater H Beachy J Smith SL Massage improves growth quality by decreasing body fat deposition in male preterm infants J Pediatr 2013 162 3 490 5 10.1016/j.jpeds.2012.08.033 23062248 24. Guzzetta A D'Acunto MG Carotenuto M Berardi N Bancale A Biagioni E The effects of preterm infant massage on brain electrical activity Dev Med Child Neurol 2011 53 Suppl 4 46 51 10.1111/j.1469-8749.2011.04065.x 21950394 25. Procianoy RS Mendes EW Silveira RC Massage therapy improves neurodevelopment outcome at two years corrected age for very low birth weight infants Early Hum Dev 2010 86 1 7 11 10.1016/j.earlhumdev.2009.12.001 20022717 26. Ho YB Lee RS Chow CB Pang MY Impact of massage therapy on motor outcomes in very low-birthweight infants: randomized controlled pilot study Pediatr Int 2010 52 3 378 85 10.1111/j.1442-200X.2009.02964.x 19761514 27. Diego MA Field T Hernandez-Reif M Procedural pain heart rate responses in massaged preterm infants Infant Behav Dev 2009 32 2 226 9 10.1016/j.infbeh.2008.12.001 19185352 28. Jain S Kumar P McMillan DD Prior leg massage decreases pain responses to heel stick in preterm babies J Paediatr Child Health 2006 42 9 505 8 10.1111/j.1440-1754.2006.00912.x 16925535 29. Hernandez-Reif M Diego M Field T Preterm infants show reduced stress behaviors and activity after 5 days of massage therapy Infant Behav Dev 2007 30 4 557 61 10.1016/j.infbeh.2007.04.002 17548111 30. Acolet D Modi N Giannakoulopoulos X Bond C Weg W Clow A Changes in plasma cortisol and catecholamine concentrations in response to massage in preterm infants Arch Dis Child 1993 68 1 Spec No 29 31 10.1136/adc.68.1_Spec_No.29 8439193 31. Mendes EW Procianoy RS Massage therapy reduces hospital stay and occurrence of late-onset sepsis in very preterm neonates J Perinatol 2008 28 12 815 20 10.1038/jp.2008.108 18633421 32. Ang JY Lua JL Mathur A Thomas R Asmar BI Savasan S A randomized placebo-controlled trial of massage therapy on the immune system of preterm infants Pediatrics 2012 130 6 e1549 58 10.1542/peds.2012-0196 23147978 33. Basiri-Moghadam M Basiri-Moghadam K Kianmehr M Jani S The effect of massage on neonatal jaundice in stable preterm newborn infants: a randomized controlled trial J Pak Med Assoc 2015 65 6 602 6 26060154 34. Smith SL Lux R Haley S Slater H Beachy J Moyer-Mileur LJ The effect of massage on heart rate variability in preterm infants J Perinatol 2013 33 1 59 64 10.1038/jp.2012.47 22538325 35. Feijo L Hernandez-Reif M Field T Burns W Valley-Gray S Simco E Mothers' depressed mood and anxiety levels are reduced after massaging their preterm infants Infant Behav Dev 2006 29 3 476 80 10.1016/j.infbeh.2006.02.003 17138300 36. Fallah R Akhavan Karbasi S Golestan M Fromandi M Sunflower oil versus no oil moderate pressure massage leads to greater increases in weight in preterm neonates who are low birth weight Early Hum Dev 2013 89 9 769 72 10.1016/j.earlhumdev.2013.06.002 23830725 37. Saeadi R Ghorbani Z Shapouri MA The effect of massage with medium-chain triglyceride oil on weight gain in premature neonates Acta Medica Iran 2015 53 2 134 8 38. Kumar J Upadhyay A Dwivedi AK Gothwal S Jaiswal V Aggarwal S Effect of oil massage on growth in preterm neonates less than 1800 g: a randomized control trial Indian J Pediatr 2013 80 6 465 9 10.1007/s12098-012-0869-7 23054851 39. Badr LK Abdallah B Kahale L A Meta-Analysis of Preterm Infant Massage: An Ancient Practice With Contemporary Applications MCN Am J Matern Child Nurs 2015 40 6 344 58 10.1097/NMC.0000000000000177 26302088 40. Wang L He JL Zhang XH The efficacy of massage on preterm infants: a meta-analysis Am J Perinatol 2013 30 9 731 8 10.1055/s-0032-1332801 23322391 41. Noble Y Boyd R Neonatal assessments for the preterm infant up to 4 months corrected age: a systematic review Dev Med Child Neurol 2012 54 2 129 39 10.1111/j.1469-8749.2010.03903.x 22142216 42. El-Dib M Massaro AN Glass P Aly H Neurodevelopmental assessment of the newborn: An opportunity for prediction of outcome Brain Dev 2011 33 2 95 105 10.1016/j.braindev.2010.04.004 20494536 43. Spittle AJ Doyle LW Boyd RN A systematic review of the clinimetric properties of neuromotor assessments for preterm infants during the first year of life Dev Med Child Neurol 2008 50 4 254 66 10.1111/j.1469-8749.2008.02025.x 18190538 44. Bosanquet M Copeland L Ware R Boyd R A systematic review of tests to predict cerebral palsy in young children Dev Med Child Neurol 2013 55 5 418 26 10.1111/dmcn.12140 23574478 45. Einspieler C Prechtl HF Ferrari F Cioni G Bos AF The qualitative assessment of general movements in preterm, term and young infants--review of the methodology Early Hum Dev 1997 50 1 47 60 10.1016/S0378-3782(97)00092-3 9467693 46. Mercuri E Ricci D Romeo DM Neurological and visual assessments in very and late low-risk preterm infants Early Hum Dev 2012 88 Suppl 1 S31 3 10.1016/j.earlhumdev.2011.12.025 22285782 47. Ricci D Romeo DM Serrao F Cesarini L Gallini F Cota F Application of a neonatal assessment of visual function in a population of low risk full-term newborn Early Hum Dev 2008 84 4 277 80 10.1016/j.earlhumdev.2007.10.002 17996405 48. Bayley N Bayley Scales of Infant and Toddler Development 2006 3 San Antonio Harcourt Assessment 49. Luttikhuizen dos Santos ES de Kieviet JF Konigs M van Elburg RM Oosterlaan J Predictive value of the Bayley scales of infant development on development of very preterm/very low birth weight children: a meta-analysis Early Hum Dev 2013 89 7 487 96 10.1016/j.earlhumdev.2013.03.008 23597678 50. El-Dib M Massaro AN Bulas D Aly H Neuroimaging and neurodevelopmental outcome of premature infants Am J Perinatol 2010 27 10 803 18 10.1055/s-0030-1254550 20486038 51. Peterson BS Vohr B Staib LH Cannistraci CJ Dolberg A Schneider KC Regional brain volume abnormalities and long-term cognitive outcome in preterm infants J Am Med Assoc 2000 284 15 1939 47 10.1001/jama.284.15.1939 52. Counsell SJ Allsop JM Harrison MC Larkman DJ Kennea NL Kapellou O Diffusion-weighted imaging of the brain in preterm infants with focal and diffuse white matter abnormality Pediatrics 2003 112 1 Pt 1 1 7 10.1542/peds.112.1.1 12837859 53. Rose SE Hatzigeorgiou X Strudwick MW Durbridge G Davies PS Colditz PB Altered white matter diffusion anisotropy in normal and preterm infants at term-equivalent age Magn Reson Med 2008 60 4 761 7 10.1002/mrm.21689 18816850 54. Huppi PS Warfield S Kikinis R Barnes PD Zientara GP Jolesz FA Quantitative magnetic resonance imaging of brain development in premature and mature newborns Ann Neurol 1998 43 2 224 35 10.1002/ana.410430213 9485064 55. Ajayi-Obe M Saeed N Cowan FM Rutherford MA Edwards AD Reduced development of cerebral cortex in extremely preterm infants Lancet 2000 356 9236 1162 3 10.1016/S0140-6736(00)02761-6 11030298 56. Inder TE Huppi PS Warfield S Kikinis R Zientara GP Barnes PD Periventricular white matter injury in the premature infant is followed by reduced cerebral cortical gray matter volume at term Ann Neurol 1999 46 5 755 60 10.1002/1531-8249(199911)46:5<755::AID-ANA11>3.0.CO;2-0 10553993 57. Boardman JP Counsell SJ Rueckert D Kapellou O Bhatia KK Aljabar P Abnormal deep grey matter development following preterm birth detected using deformation-based morphometry NeuroImage 2006 32 1 70 8 10.1016/j.neuroimage.2006.03.029 16675269 58. Inder TE Warfield SK Wang H Huppi PS Volpe JJ Abnormal cerebral structure is present at term in premature infants Pediatrics 2005 115 2 286 94 10.1542/peds.2004-0326 15687434 59. Malamateniou C Adams ME Srinivasan L Allsop JM Counsell SJ Cowan FM The anatomic variations of the circle of Willis in preterm-at-term and term-born infants: an MR angiography study at 3T AJNR Am J Neuroradiol 2009 30 10 1955 62 10.3174/ajnr.A1724 19643922 60. Pannek K Guzzetta A Colditz PB Rose SE Diffusion MRI of the neonate brain: acquisition, processing and analysis techniques Pediatr Radiol 2012 42 10 1169 82 10.1007/s00247-012-2427-x 22903761 61. Feldman HM Yeatman JD Lee ES Barde LH Gaman-Bean S Diffusion tensor imaging: a review for pediatric researchers and clinicians J Dev Behav Pediatr 2010 31 4 346 56 10.1097/DBP.0b013e3181dcaa8b 20453582 62. Pannek K Scheck SM Colditz PB Boyd RN Rose SE Magnetic resonance diffusion tractography of the preterm infant brain: a systematic review Dev Med Child Neurol 2014 56 2 113 24 10.1111/dmcn.12250 24102176 63. Kwon SH Vasung L Ment LR Huppi PS The role of neuroimaging in predicting neurodevelopmental outcomes of preterm neonates Clin Perinatol 2014 41 1 257 83 10.1016/j.clp.2013.10.003 24524459 64. Scher MS Steppe DA Sclabassi RJ Banks DL Regional differences in spectral EEG measures between healthy term and preterm infants Pediatr Neurol 1997 17 3 218 23 10.1016/S0887-8994(97)00101-X 9390697 65. Scher MS Sun M Steppe DA Banks DL Guthrie RD Sclabassi RJ Comparisons of EEG sleep state-specific spectral values between healthy full-term and preterm infants at comparable postconceptional ages Sleep 1994 17 1 47 51 8191202 66. Scher MS Sun M Steppe DA Guthrie RD Sclabassi RJ Comparisons of EEG spectral and correlation measures between healthy term and preterm infants Pediatr Neurol 1994 10 2 104 8 10.1016/0887-8994(94)90041-8 8024658 67. Shany E Berger I Neonatal electroencephalography: review of a practical approach J Child Neurol 2011 26 3 341 55 10.1177/0883073810384866 21383227 68. Guzzetta A Baldini S Bancale A Baroncelli L Ciucci F Ghirri P Massage accelerates brain development and the maturation of visual function J Neurosci 2009 29 18 6042 51 10.1523/JNEUROSCI.5548-08.2009 19420271 69. Gulamani SS Premji SS Kanji Z Azam SI A review of postpartum depression, preterm birth, and culture J Perinat Neonatal Nurs 2013 27 1 52 9 10.1097/JPN.0b013e31827fcf24 23360942 70. Vigod SN Villegas L Dennis CL Ross LE Prevalence and risk factors for postpartum depression among women with preterm and low-birth-weight infants: a systematic review BJOG 2010 117 5 540 50 10.1111/j.1471-0528.2009.02493.x 20121831 71. Korja R Latva R Lehtonen L The effects of preterm birth on mother-infant interaction and attachment during the infant's first two years Acta Obstet Gynecol Scand 2012 91 2 164 73 10.1111/j.1600-0412.2011.01304.x 22007730 72. Kraljevic M Warnock FF Early educational and behavioral RCT interventions to reduce maternal symptoms of psychological trauma following preterm birth: a systematic review J Perinat Neonatal Nurs 2013 27 4 311 27 10.1097/JPN.0b013e3182a8bfe2 24164814 73. Chertok IR McCrone S Parker D Leslie N Review of interventions to reduce stress among mothers of infants in the NICU Adv Neonatal Care 2014 14 1 30 7 10.1097/ANC.0000000000000044 24472886 74. Davis L Mohay H Edwards H Mothers' involvement in caring for their premature infants: an historical overview J Adv Nurs 2003 42 6 578 86 10.1046/j.1365-2648.2003.02661.x 12787231 75. Onozawa K Glover V Adams D Modi N Kumar RC Infant massage improves mother-infant interaction for mothers with postnatal depression J Affect Disord 2001 63 1–3 201 7 10.1016/S0165-0327(00)00198-1 11246096 76. Feldman R Eidelman AI Neonatal state organization, neuromaturation, mother-infant interaction, and cognitive development in small-for-gestational-age premature infants Pediatrics 2006 118 3 e869 78 10.1542/peds.2005-2040 16880249 77. Feldman R Eidelman AI Sirota L Weller A Comparison of skin-to-skin (kangaroo) and traditional care: parenting outcomes and preterm infant development Pediatrics 2002 110 1 Pt 1 16 26 10.1542/peds.110.1.16 12093942 78. Boyd RC Le HN Somberg R Review of screening instruments for postpartum depression Arch Womens Ment Health 2005 8 3 141 53 10.1007/s00737-005-0096-6 16133785 79. Lovibond SL, PF. Manual for the Depression Anxiety Stress Scales. 2nd ed. Sydney: Psychology Foundation; 1995. 80. Parkitny L McAuley J The Depression Anxiety Stress Scale (DASS) J Physiother 2010 56 3 204 10.1016/S1836-9553(10)70030-8 20795931 81. Brown TA Chorpita BF Korotitsch W Barlow DH Psychometric properties of the Depression Anxiety Stress Scales (DASS) in clinical samples Behav Res Ther 1997 35 1 79 89 10.1016/S0005-7967(96)00068-X 9009048 82. Amankwaa L Pickler R Measuring maternal responsiveness ABNF J 2007 18 1 4 15 18402351 83. Nagata M Nagai Y Sobajima H Ando T Honjo S Depression in the mother and maternal attachment--results from a follow-up study at 1 year postpartum Psychopathology 2003 36 3 142 51 10.1159/000071259 12845285 84. George CaS J Shaver JCPR The caregiving behavioural system: A behavioural system approach to parenting The Handbook of attachment: Theory, research and clinical applications 2008 New York Guilford Press 833 56 85. Barnard KE Hammond MA Booth CL Bee HL Mitchell SK Spieker SJ Measurement and meaning of parent–child interaction 1989 New York Academic 86. Easterbrooks MA Goldberg WA Greenberg DC M Cummings EM Security of attachment to mother and to father during toddlerhood: Relation to children's sciopersonality functioning during kindergarten Attachment in the preschool years: Theory, research and intervention 1990 Chicago University of Chicago Press 221 44 87. Carter AS Briggs-Gowan MJ Jones SM Little TD The Infant-Toddler Social and Emotional Assessment (ITSEA): factor structure, reliability, and validity J Abnorm Child Psychol 2003 31 5 495 514 10.1023/A:1025449031360 14561058 88. Papile LA Munsick-Bruno G Schaefer A Relationship of cerebral intraventricular hemorrhage and early childhood neurologic handicaps J Pediatr 1983 103 2 273 7 10.1016/S0022-3476(83)80366-7 6875724 89. Field TM Schanberg SM Scafidi F Bauer CR Vega-Lahr N Garcia R Tactile/kinesthetic stimulation effects on preterm neonates Pediatrics 1986 77 5 654 8 3754633 90. Definitions for audit ANZNN Data Dictionary 2013 165 91. Dubowitz LM Dubowitz V Palmer P Verghote M A new approach to the neurological assessment of the preterm and full-term newborn infant Brain Dev 1980 2 1 3 14 10.1016/S0387-7604(80)80003-9 7416439 92. Dubowitz L, Dubowitz, V., Mercuri E. Neurological Assessment of the Preterm and Fullterm Newborn Infant. Mac Keith Press, London; 1999 93. Dubowitz L Mercuri E Dubowitz V An optimality score for the neurologic examination of the term newborn J Pediatr 1998 133 3 406 16 10.1016/S0022-3476(98)70279-3 9738726 94. Cunningham NK Brown PM Brooks J Page AC The structure of emotional symptoms in the postpartum period: is it unique? J Affect Disord 2013 151 2 686 94 10.1016/j.jad.2013.08.002 23978685 95. Cox JL Holden JM Sagovsky R Detection of postnatal depression. Development of the 10-item Edinburgh Postnatal Depression Scale Br J Psychiatry 1987 150 782 6 10.1192/bjp.150.6.782 3651732 96. Taylor A Atkins R Kumar R Adams D Glover V A new Mother-to-Infant Bonding Scale: links with early maternal mood Arch Womens Mental Health 2005 8 1 45 51 10.1007/s00737-005-0074-z 97. Kidokoro H Neil JJ Inder TE New MR imaging assessment tool to define brain abnormalities in very preterm infants at term AJNR Am J Neuroradiol 2013 34 11 2208 14 10.3174/ajnr.A3521 23620070 98. Tournier JD Calamante F Connelly A Robust determination of the fibre orientation distribution in diffusion MRI: non-negativity constrained super-resolved spherical deconvolution NeuroImage 2007 35 4 1459 72 10.1016/j.neuroimage.2007.02.016 17379540 99. Ball G Counsell SJ Anjari M Merchant N Arichi T Doria V An optimised tract-based spatial statistics protocol for neonates: applications to prematurity and chronic lung disease NeuroImage 2010 53 1 94 102 10.1016/j.neuroimage.2010.05.055 20510375 100. Raffelt D Tournier JD Rose S Ridgway GR Henderson R Crozier S Apparent Fibre Density: a novel measure for the analysis of diffusion-weighted magnetic resonance images NeuroImage 2012 59 4 3976 94 10.1016/j.neuroimage.2011.10.045 22036682 101. Condon JTCC The assessment of parent-to-infant attachment: Development of a self-report questionnaire instrument J Reprod Infant Psychol 1998 16 1 57 76 10.1080/02646839808404558 102. Teti DM Gelfand DM Behavioral competence among mothers of infants in the first year: the mediational role of maternal self-efficacy Child Dev 1991 62 5 918 29 10.2307/1131143 1756667

PMC005xxxxxx/PMC5002319.txt

==== Front BMC Musculoskelet DisordBMC Musculoskelet DisordBMC Musculoskeletal Disorders1471-2474BioMed Central London 120910.1186/s12891-016-1209-2Research ArticleReductions in co-contraction following neuromuscular re-education in people with knee osteoarthritis Preece Stephen J. s.preece@salford.ac.uk 1Jones Richard K. r.k.jones@salford.ac.uk 1Brown Christopher A. Christopher.Brown@manchester.ac.uk 2Cacciatore Timothy W. tim.cacciatore@gmail.com 3Jones Anthony K. P. anthony.jones@manchester.ac.uk 21 Centre for Health Sciences Research, University of Salford, Manchester, M6 6PU UK 2 Human Pain Research Group, University of Manchester, Clinical Sciences Building, Salford Royal NHS Foundation Trust, Salford, M6 8HD UK 3 Institute of Neurology, University College London, Queen Square, London, WC1N 3BG UK 27 8 2016 27 8 2016 2016 17 1 3722 3 2016 10 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Both increased knee muscle co-contraction and alterations in central pain processing have been suggested to play a role in knee osteoarthritis pain. However, current interventions do not target either of these mechanisms. The Alexander Technique provides neuromuscular re-education and may also influence anticipation of pain. This study therefore sought to investigate the potential clinical effectiveness of the AT intervention in the management of knee osteoarthritis and also to identify a possible mechanism of action. Methods A cohort of 21 participants with confirmed knee osteoarthritis were given 20 lessons of instruction in the Alexander Technique. In addition to clinical outcomes EMG data, quantifying knee muscle co-contraction and EEG data, characterising brain activity during anticipation of pain, were collected. All data were compared between baseline and post-intervention time points with a further 15-month clinical follow up. In addition, biomechanical data were collected from a healthy control group and compared with the data from the osteoarthritis subjects. Results Following AT instruction the mean WOMAC pain score reduced by 56 % from 9.6 to 4.2 (P < 0.01) and this reduction was maintained at 15 month follow up. There was a clear decrease in medial co-contraction at the end of the intervention, towards the levels observed in the healthy control group, both during a pre-contact phase of gait (p < 0.05) and during early stance (p < 0.01). However, no changes in pain-anticipatory brain activity were observed. Interestingly, decreases in WOMAC pain were associated with reductions in medial co-contraction during the pre-contact phase of gait. Conclusions This is the first study to investigate the potential effectiveness of an intervention aimed at increasing awareness of muscle behaviour in the clinical management of knee osteoarthritis. These data suggest a complex relationship between muscle contraction, joint loading and pain and support the idea that excessive muscle co-contraction may be a maladaptive response in this patient group. Furthermore, these data provide evidence that, if the activation of certain muscles can be reduced during gait, this may lead to positive long-term clinical outcomes. This finding challenges clinical management models of knee osteoarthritis which focus primarily on muscle strengthening. Trial registration ISRCTN74086288, 4th January 2016, retrospectively registered. Keywords Knee osteoarthritisAlexander TechniqueGaitCo-contractionPainElectroencephalographyBupa Foundation (GB)TBF-PPW10-047Preece Stephen J. issue-copyright-statement© The Author(s) 2016 ==== Body Background Knee osteoarthritis (OA) is a major cause of disability and reduced quality of life across the world [1, 2] and has been estimated to affect over 12.5 % of the UK population [3]. Historically the condition was viewed as a degenerative disease of the articular cartilage and subchondral bone and that OA-related pain with a direct result of this destructive process. However, numerous studies have demonstrated a lack of concordance between radiographic measures of joint degeneration and clinical pain [4] suggesting that a range of different mechanisms may underlie knee OA pain [5] including peripheral and central mechanisms [6]. There is now a large body of evidence demonstrating that patients with knee OA exhibit excessive muscular co-contraction (simultaneous activation of the quadriceps and hamstrings) during walking [7–11] and other functional tasks [12–15]. This co-contraction increases compressive loads at the knee joint surface [16, 17], accelerates structural progression of the disease [18] and increases the likelihood that patients will progress to total knee arthroplasty [19]. Elevated loading may also increase the stress on articular structures, such as the joint, bone, synovium/joint capsule and periarticular structures, resulting in increased pain. Another mechanism that has been suggested to mediate OA-related pain is an alteration in central pain perception, in which supraspinal processes affect nociceptive processing [20]. Support for this idea has been provided by experimental studies which have demonstrated an increase in anticipation-evoked EEG potentials in patients with OA [6]. These findings highlight the possibility that sensitisation of nociceptive pathways may result in patients with OA perceiving relatively low level stimuli as being overtly painful. Interestingly, a recent study demonstrated that it may be possible to reduce anticipatory pain activity with mindfulness training [21]. These findings demonstrate the possibility of using cognitive methods to influence supraspinal processing and potentially nociceptive signals. Current conservative first-phase management of knee OA is primarily focused around physiotherapist-delivered exercise programmes. These programmes typically incorporate different components ranging from simple muscle strengthening/stretching [22–24] and aerobic walking [25, 26] through to balance and coordination training [27]. In a recent review it was concluded that the magnitude of the sustained (2–6 months) benefit from exercise programmes is small, with a typical reduction in pain of 6 points on a 0:100-point scale [28]. However, the mechanism of action of current exercise interventions is not clear. It is possible that these small improvements in pain could be the result of changes in muscular co-contraction or central pain processing. However, exercise interventions do not directly target these factors. Therefore further research is required to understand the potential clinical efficacy of interventions which have the potential to modify muscle activation patterns and/or central pain processing. The Alexander Technique (AT) is a method of neuromuscular re-education which aims to teach individuals how to improve postural support, reduce potentially harmful patterns of muscle tension and improve control of response. AT lessons provide an individualised approach to developing skills that help people recognise, understand, and avoid poor habits adversely affecting postural tone and neuromuscular coordination. For further explanation of the AT, the reader is referred to the description provided by Cacciatore et al. [29]. Randomised controlled studies have shown beneficial effects of one-to-one AT lessons for a range of conditions [30]. For example AT lessons have been shown to improve clinical outcomes in people with chronic low back [31], reduce pain in individuals suffering with neck pain [32] and also to improve self-reported disability and depression in people diagnosed with Parkinson’s disease [33]. Research has also demonstrated that training in the AT can improve neuromuscular coordination and enhance the dynamic regulation of postural tone [34, 35]. In a recent study of the sit-to-stand movement, AT enabled smoother movement at lift-off which could be explained by reduced leg extensor resistance [36]. These findings suggest that improvements in neuromuscular control, which result from AT instruction, may include altered neuromuscular control of knee extensor moments. If this is the case then applying the AT may lead to reduced co-contraction of the knee muscles and therefore may be beneficial for patients with knee OA. As part of AT instruction, individuals are encouraged to become aware of, and consciously inhibit, increases in muscle activity which are triggered in anticipation of certain stimuli, such as those provoking pain. It is possible that this focus on anticipatory muscular behaviour could have an effect on supraspinal processes which influence nociceptive processing and therefore OA-related pain. Given the potential of the AT to influence both muscular co-contraction and central pain processing, we designed a study to investigate potential mechanisms of action of the AT in the clinical management of knee OA. The first aim of the study was to develop an understanding of the magnitude of the clinical change, both in the short and long term, which may result from AT instruction. Secondly, we sought to understand if AT instruction would lead to reduced muscular co-contraction in a knee OA cohort towards the level characteristic of an age-matched healthy population, and also if it would alter anticipatory pain responses. The final aim of the study was to investigate relationships between changes in clinical outcomes and changes in co-contraction or anticipatory pain responses. Methods In order to develop an understanding of the potential effect of the AT in knee OA, an uncontrolled pre-post design was used. With this approach a total of 21 participants with knee OA were assessed for pain, biomechanical function and pain processing before and after AT instruction. Further biomechanical comparisons were then made between the AT group and a cohort of age-matched healthy controls. Participants A total of 22 participants with knee OA were recruited and offered AT lessons. As no previous data for the clinical effectiveness of AT lessons for knee OA were available, the sample size calculation was based on data from a study of a self-management and exercise rehabilitation programme [37]. This study reported a change in WOMAC pain scores of 1.5 SD immediately after the 20-sessson intervention. Based on these data, an a priori, within-sample calculation was performed using the g*power software. This showed that a target sample of n = 20 (assuming 10 % attrition), would be sufficient to detect a change of 0.75 SD in WOMAC pain score (half that reported in [37]) with a statistical power of 0.8 and an α = 0.05. Participants with knee OA were recruited through four general practitioners in the Greater Manchester area. In each of the five practices, electronic patient records were reviewed to identify all participants who satisfied the following criteria:X-ray diagnosis of knee OA. Between 40 and 70 years of age. No diabetes, systematic disorders, such as rheumatoid arthritis. No lower limb arthroplasty. No previous experience of the Alexander Technique. A total of 150 letters of invitation were sent out. Those who replied (n = 38) underwent further telephone screening to ensure that they did not have any problems with balance, satisfied all the above criteria and regularly experienced knee pain during walking. A total of 22 individuals satisfied these criteria and all were invited to participate in the study. A healthy control group (n = 20) was also recruited in order to address the second research question, relating to biomechanical joint loading. This group was recruited via advert around the university and through local community groups, such as U3A and Rotary. Each healthy participant had to satisfy criteria 2–5 above and have no history of musculoskeletal disorders of the lower limb or spine. Healthy participants were selected so that the mean age and BMI of this group was matched to that of the knee OA group. Before testing all subjects provided written informed consent to participate in the study and ethical approval was obtained from the NRES Greater Manchester North ethics committee, reference: 11/NW/0057. The Alexander technique intervention The Alexander Technique (AT) intervention was delivered by an experienced local practitioner who was a certified member of the Society of Teachers of the Alexander Technique. The AT is usually delivered on a one-to-one basis and, for this study, each participant was offered 20 one-to-one AT lessons each lasting 40 min. The lessons were delivered over a 12 week period, twice a week for the first 8 weeks and weekly for the final 4 weeks. This rate of delivery was chosen to ensure similarity with other clinical trials investigating the AT, as well as with routine practice [31, 32]. One of the primary objectives of AT instruction is to improve one’s overall pattern of postural muscle tension. This is achieved by guiding an individual to prioritise attention to maintaining a dynamic coordinated and lengthening central body axis and to improved awareness of appropriate and inappropriate postural and tensional patterns. Typically, these skills are taught using such common movements as sit-to-stand or walking, as well as with the person lying in a semi-supine position. The teacher uses gentle manual guidance and verbal instruction, together with constructive feedback, to enable the individual to lessen habitual interference with and maintain the lengthened axial tensional pattern, thereby reducing inappropriate tensional patterns in general, and typically leading to movement that perceivably requires less effort. In addition, they are taught to be aware of subtle increases in anticipatory muscle activity which are triggered during movement and also in anticipation of pain. Individuals are taught to interpose mental ‘directions’ between their intention to move and their execution of the movement so as to maintain postural support activity in the spine and torso, and to avoid overreacting in anticipation of the usual effort (or pain) involved. Following the guided sessions, participants were encouraged to continue to apply the skills learnt in the lessons as they went about their daily activities, maintaining an improved postural awareness and state and continuing to inhibit inappropriate tensional patterns. Clinical outcomes Clinical effectiveness of the AT intervention was assessed using the WOMAC self-report questionnaire which captures information on pain, stiffness and function [38]. Given our hypothesis that AT instruction may lead to reduced co-contraction and subsequent pain, we defined the WOMAC pain score (5 items from the full WOMAC questionnaire) as the primary outcome measure. The full WOMAC score (20 items) was also analysed as a secondary outcome measure. Clinical outcomes were collected at baseline, immediately post intervention (within 1 week of the final AT lesson) and also at 15 months post baseline. In addition, a record of analgesia use during the week before the baseline assessment and the week before the post intervention assessment was taken as well as a record of any other therapy accessed during the intervention period. Biomechanical assessment and outcomes Biomechanical data, used to characterise joint loading, and EEG data, used to characterise pain processing, were collected at baseline and immediately post intervention. Biomechanical data from a healthy cohort were also collected separately to characterise joint loading in this group. During each of the biomechanical testing sessions, participants walked barefoot at a speed of 1.25 ms−1 along a walkway. Speed was measured with optical timing gates and only speeds within ±10 % considered acceptable. All subjects apart from one participant with knee OA were able to walk at this speed. In this one participant a walking speed of 1 ms−1 was found to be appropriate and used for both the baseline and repeat testing. Although some previous studies of muscle activation, in people with knee OA, have instructed participants to walk at a self-selected speed [8], EMG amplitudes are known to vary with walking speed [39]. Therefore, in order to ensure appropriate comparison across testing sessions and between healthy and OA groups, we opted to control walking speed. EMG data from the hamstrings and quadriceps was collected during walking to quantify co-contraction of the knee muscles. Data were collected using a Noraxon Telemyo system (3000Hz) with electrodes placed on vastus lateralis, vastus medialis, biceps femoris and semimembranosus according to SENIAM guidelines [40]. Following the walking trials, reference EMG data were collected during maximal voluntary isometric contractions (MVIC) for amplitude normalisation of the final EMG signals. To collect these data, the participant was seated in an isometric dynamometer with the knee flexed at 45° and, following a warm-up period, MVIC data collected first from the quadriceps and then the hamstrings. Three MVICs were performed for each muscle group with a 1 min rest between each contraction. During these contractions the net joint torque (measured with the dynamometer) was recorded in order to quantify knee extensor/flexor moments. Following data collection, EMG data was exported to Matlab for processing. After applying a 20Hz high pass FFT filter to remove noise and movement artefact, the signal was rectified and then low pass filtered (6Hz Butterworth) to create a linear envelop [8]. The EMG signals were then time normalised to the stance period of the most affected limb using gait event data captured from the force platforms (see below). The MVIC data was filtered in the same way as the movement data and then a moving window algorithm used to determine the 0.1 s window in which the maximum EMG amplitude occurred [8]. This MVIC value was then used to normalise the EMG data from the walking trials. In order to characterise knee extensor/flexor strength, the peak torque was identified across the three maximal quadriceps/hamstring contractions. We characterised muscular co-contraction during two specific phases of the gait cycle: an early stance phase and a pre-contact phase. Modelling studies have shown that, during early stance, there is a peak in the knee contact force which occurs between 15 and 25 % of stance [17]. It has also been shown that this peak increases with knee muscle co-contraction [17]. We therefore characterised muscular co-contraction during this period. We also characterised co-contraction over a pre-contact phase period (−5 % to 0 % of stance), just before initial contact. It has been shown that patients with knee OA increase background muscular co-contraction in anticipation of a destabilising perturbation [41] and it is possible that such anticipatory muscle activity may occur during walking, in preparation for contact with the ground. Such increased anticipatory contraction of the hamstring and quadriceps may lead to a subsequent increase in knee joint stiffness during the loading period. This may, in turn, affect the rate of increase of the knee contact force which could subsequently affect nociceptor input. As AT training aims to develop awareness and influence anticipatory muscle activity, it is possible that this intervention may influence co-contraction during the pre-contact phase. A range of different algorithms have been proposed to calculate co-contraction of the knee muscles during walking. However, results from a recent modelling study suggested that simply summing the activity of the agonist and antagonist may give the best indication of articular loading [16]. Therefore, separate medial and lateral co-contraction EMG curves were obtained by summing the medial quadriceps and hamstrings and the lateral quadriceps and hamstrings respectively [10]. The final two co-contraction outcomes were then calculated as the respective means of the medial and lateral co-contraction curves over both the pre-contact phase (−5 % to 0 %) and the early stance phase (15 % to 25 %). Each time window was adjusted backwards to account for a 30 ms electromechanical delay. Force and 3D motion plate data were also collected in order to quantify other aspects of joint loading. These data were collected using a 10-camera Qualisys Pro-reflex motion capture system (100 Hz) with two AMIT force plates (1500Hz) embedded in the walkway. Rigid clusters of 4 markers were used to track the motions of the thigh and shank, and a system of 4 markers, placed over anatomical landmarks, used to track motion of the foot [42]. Ankle and knee joint centres were calculated as midpoints between the malleoli and femoral epicondyles respectively and hip joint centres obtained using the regression model of Bell et al. [43]. Following data collection, the Visual 3D software (C-Motion, Rockville, Maryland) was used to derive the sagittal plane knee angle angles and also the sagittal and frontal plane knee moments from the kinematic data using a 6DOF model. These data were then time normalised to the stance phase. Peak sagittal angle, peak sagittal moment and peak frontal moment were then derived to characterise knee loading. Both peak moments were subsequently normalised to the participant’s body mass to define the final outcomes. Pain processing assessment and outcomes We used a 64-channel EEG system (BrainAmp MR, BrainVision UK) to measure brain activity (sampling rate of 500Hz, FCz reference), and a thulium laser stimulator to induce 30 very brief (<150 ms) heat sensations on the right forearm. Prior to EEG recording, a suitable stimulus intensity was determined for each participant that induced moderately painful sensations. This was tailored to each individual patient using a psychophysics procedure, such that they judged a moderately pain sensation (described to the participant as clearly painful but easily tolerable) as a rating of 7 on a numerical rating scale (NRS) from 0 to 10. The resulting laser energy output (mean (standard deviation) across patients was 1.2 (0.3) Joules. During EEG recording, participants provided ratings of pain intensity for each laser pulse on the same 0 – 10 NRS. At 3 s, 2 s and 1 s prior to each stimulus onset, they were cued with an auditory stimulus to anticipate the timing of the pain, allowing for the recording of anticipatory brain responses. After each laser pulse, laser-evoked potentials were also recorded, allowing for measurement of the brain’s reaction to pain sensations. These anticipatory and laser-evoked responses were derived from the average, across all trials, of EEG responses time-locked to the stimulus onset, after pre-processing of the data using standard analysis techniques. These included filtering of the data to retain 0 – 30Hz frequencies, cleaning of eye-movement and other artefacts using Independent Components Analysis, manual rejection of remaining artefactual epochs, baseline correction to -3500 ms to –3000 ms pre-stimulus and re-referencing to the common average of all scalp channels. The analysis of the resulting event-related potentials (ERPs) focussed on two time windows. For anticipation, the mean amplitude of the ERP at electrode Cz and its eight surrounding electrodes during the late anticipatory period (−500 ms to 0 ms pre-stimulus – see Fig. 1) was analysed, as this previously showed enhanced amplitude in a study of OA patients relative to healthy controls [6]. To measure the neural response to pain, the relative difference between the N2 and P2 peaks of the laser-evoked potential was analysed, as these peaks have been previously shown to be modulated by cognitive factors such as attention and expectations of pain intensity [44, 45]. The N2 peak was specified as the largest negative deflection in the ERP in the time period between 200 ms and 350 ms after laser stimulation at electrode Cz, while P2 was the largest positive deflection between 350 ms and 500 ms at electrode Cz. For each peak, activity was averaged across electrode Cz and its eight surrounding electrodes over a 20 ms time window centred on the patients’ peak latencies.Fig. 1 Anticipatory and laser-evoked potentials derived from the EEG signal, averaged across all participants with knee OA and testing sessions. Three auditory tones presented once per second counted down the onset of the laser stimulus Statistical analysis We used a Wilcoxon signed rank test to investigate changes in WOMAC pain and full WOMAC score following the AT intervention. Paired sample t-tests were used to establish whether any of the biomechanical or pain processing outcomes changed significantly following AT instruction. Following this, independent t-tests were used to compare the biomechanical loading variables between the healthy and knee OA groups. Finally, to investigate the link between changes in mechanistic outcomes and changes in clinical outcomes, we used a Pearson’s correlation analysis. For all analyses, α < 0.05 was chosen as the significance level. Although, with the relatively large number of statistical tests, this increases the likelihood of a type 1 error, it was deemed appropriate given the exploratory nature of this study. Results Participant disposition, adherence and baseline characteristics A total of 22 individuals with knee OA were recruited and began the AT intervention. One participant dropped out (after 10 AT lessons) and was removed from the study; however the remaining 21 participants completed all 20 AT lessons. Biomechanical and clinical data was collected from all 21 participants at baseline and immediately post-intervention, however, only 19 participants agreed to undergo the pain processing assessment. Six participants were lost to long-term follow up, which left 15 patients with OA at the 15 month follow up. There were no significant differences in baseline demographics between the healthy and OA groups (Table 1). The main WOMAC pain score of the knee OA participants was relatively high, reflecting our inclusion criteria of knee pain on walking. Kellgren and Lawrence (KL) [46] grades of the OA participants ranged from 2 to 4 with 11 participants being classified as KL = 2, nine participants as KL = 3 and four participants with KL = 4. Only six participants experienced unilateral symptoms with the rest reporting bilateral knee OA pain. Of the 21 participants, 15 reported other musculoskeletal pain, including shoulder, hip, neck and low back pain and 15 reported taking analgesia for their knee OA pain at baseline.Table 1 Baseline demographics and disease characteristics for participants with knee OA and healthy control subjects Knee OA participants (n = 21) Control subjects (n = 20) P-value Male/female 10/11 12/8 0.54 Age, years (SD) 62 (10) 61 (9) 0.68 Weight, kg (SD) 84 (13) 79 (14) 0.22 Height, cm SD) 169 (9) 170 (7) 0.8 BMI, kg/m2 (SD) 29 (4) 27 (4) 0.1 WOMAC pain (SD) 9.6 (3.0) - - WOMAC overall (SD) 45 (13) - - Clinical outcomes At the end of the AT instruction period there was a reduction from baseline of 56 % (p < 0.01) in the WOMAC pain score and 54 % (p < 0.01) in the full WOMAC score (Table 2). These reductions were maintained at 15 months (p < 0.01, Table 2). Of the 15 participants who had taken analgesia for their knee OA pain, 10 participants had reduced or stopped taking the medication and 5 had maintained the same level. Interestingly of the 15 participants reporting other musculoskeletal pain at baseline, 11 reported improvements in this non-knee related musculoskeletal pain immediately after the intervention. None of the participants accessed additional therapy during the intervention period.Table 2 Clinical changes following AT instruction Change from baseline at end of intervention [SD] (N = 21) Change from baseline at 15-month follow-up [SD] (N = 15) WOMAC Pain Score 56 % (9.6 [3.0] - 4.2 [2.7]); p < 0.01 51 % (9.1 [3.2] to 4.4 [2.7]); p < 0.01 WOMAC Overall Score 54 % (45 [13] - 21 [13]); p < 0.01 43 % (43 [14] to 25 [14]); p < 0.01 Biomechanical outcomes Following AT instruction, medial co-contraction was observed to decrease by 13 % (p < 0.05) during the pre-contact phase, however, there was no significant change in lateral co-contraction during this period (n = 19). Note that, due to problems with the instrumentation, it was only possible to obtain EMG for 19 of the 21 OA participants. Comparison with the healthy controls showed medial co-contraction to be 39 % higher (p < 0.02) in the knee OA group at baseline but only 22 % higher following the AT intervention, a difference which was no longer significant (p = 0.12). There was no significant difference between the healthy controls and the OA group in lateral co-contraction either before or after the AT intervention. Similar changes in muscle activation were observed during the early-stance period following the AT intervention. Specifically, there was a reduction in medial hamstring (Fig. 2d) and medial quadriceps (Fig. 2b) activity which led to a 15 % reduction in medial co-contraction (p < 0.01, Fig. 2f). However the reduction in lateral co-contraction was not significant (p = 0.07, Fig. 2e). Comparison with the healthy controls showed medial co-contraction to be 57 % higher (p = 0.001) in the knee OA group at baseline. Although this reduced to 34 % following the intervention, the difference was still significant (p = 0.02). Analysis of the lateral co-contraction data revealed a similar pattern being 81 % higher (P < 0.001) in the OA group at baseline and 66 % (p < 0.001) higher following the AT intervention.Fig. 2 Mean normalised muscle activation during early the stance phase (15-25 %) for (a) vastus lateralis (VL), (b) vastus medialis (VM), (c) biceps femoris and (d) semitendinosus. Healthy participant data is shown in dark grey, baseline knee OA data in white and post-AT knee OA data in light grey. Plots (e) and (f) show lateral/medial co-contraction, calculated as the sum of hamstring and quadriceps activity and horizontal bars denotes significant differences (p < 0.01). Note that statistical testing was only performed on the measures of co-contraction The strength data, measured using the isokinetic dynamometer, showed no differences in either peak knee extensor torque (p = 0.72) or peak flexor torque (p = 0.27) following AT instruction. However, although healthy participants were able to generate peak knee extensor torques which were 47 % greater (p < 0.002) than those produced by the patients with knee OA, there were no group differences in peak knee flexor torque. A similar pattern was observed during walking, with no statistical changes in peak knee joint angle or peak moments (Fig. 3) following AT instruction. However, there was a significantly higher peak sagittal moment in the healthy subjects compared to the baseline OA data (p = 0.03) and a significantly higher (p < 0.01) peak knee adduction moment in the OA participants when compared to the control subjects.Fig. 3 (a) Sagittal plane knee angle, (b) sagittal plane knee moment and (c) frontal plane knee moment for healthysubjects (red), OA patients at baseline (blue) and OA patients after the intervention (green) Pain processing outcomes When the pain processing outcomes were analysed, no statistical changes were observed in either the late-anticipatory potential (p = 0.77) or the N2-P2 (Fig. 1) difference in the laser-evoked potential (p = 0.32). Relationship between biomechanical and clinical outcomes Correlational analyses were carried out to test for a relationship between the change in WOMAC pain score and the change in medial/lateral co-contraction following AT instruction. These analyses were performed separately for the two different phases: the pre-contact phase and the early stance phase. For the pre-contact phase, a moderate correlation (r = 0.45, p < 0.05) was observed between the change in the WOMAC pain score and the change in medial co-contraction (Fig. 4a). This relationship showed a clear outlier (Fig. 4a), which when excluded, increased the correlation to r = 0.63, p < 0.01. There was some evidence of a correlation between the change in lateral co-contraction and change in the WOMAC pain during the pre-contact period, however this failed to reach significance (r = 0.37, p = 0.13, Fig. 4b). Furthermore, there was no evidence of a correlation between either of the co-contraction measures during the early stance period nor were any meaningful correlations observed between the changes in pain processing outcomes and the change in WOMAC pain.Fig. 4 The relationship between the change in WOMAC pain (following the intervention) and the change in a) lateral and b) medial co-contraction during the pre-contact phase. The filled circle shows the outlying data point Discussion We hypothesised that instruction in the AT may influence both muscular co-contraction and central pain processing and that, via these mechanisms, there would be an improvement in clinical pain and function in people with knee OA. Although this was not a randomised controlled trial, the data support the idea that AT may be an effective clinical intervention for people with knee OA and that it may reduce medial co-contraction. However, we did not find evidence that AT lessons influenced central pain processing. Our original hypothesis for investigating anticipatory brain activity was based on a study which showed that mindfulness may alter EEG activity triggered in anticipation of pain [21]. However, although this study did show some evidence that the AT may influence anticipatory muscle activity during walking, there was no evidence that this led to a concomitant reduction in either anticipatory, or pain-evoked, EEG activity resulting from experimentally induced pain at standardised intensities. These finding may indicate that the effects of AT are mainly related to muscle function and less likely to be due to cognitive effects on early anticipatory processing of pain that have been seen with other interventions such as placebo and mindfulness-based CBT. However, it is possible that the AT does influence pain processing via mechanisms which were not captured using the EEG outcomes measured in this study or that the null finding was a result of the relatively low sample size. Nevertheless, the results motivate further study into interventions for people with knee OA which can decrease muscle co-contraction. To our knowledge, this is the first study to investigate the potential effectiveness of an intervention which is aimed at increasing awareness of muscle activation patterns for the clinical management of knee OA. The results are very encouraging and compare favourably with the results of previous trials which have investigated exercise-based management approaches [28]. The present study did not include a control group receiving usual care, however large scale RCTs of knee OA typically show reductions of approximately 1.5 points in WOMAC pain in a usual care arm over a 6–18 month period [37]. Our data showed a reduction of approximately 5 points in our intervention cohort over a 3 month period which appeared to be maintained 12 months after the end of the intervention. This is considerably larger than the 2.5-point reduction in WOMAC pain typical of exercise-based interventions [37]. These results identify the need for further large-scale trials to confirm whether lessons in the AT can bring about long-term improvement in pain and function in people with knee OA. Given the potential implications for disease progression [18], there has been considerable interest in conservative interventions which have the potential to reduce co-contraction. For example, research into knee bracing has consistently shown that this approach can reduce co-contraction, in some cases by up to 35 % [47]. However, compliance is problematic and bracing may not be a viable long-term option for many patients with knee OA. Studies of exercise interventions provide conflicting findings, with some showing no change [48] and others a decrease in co-contraction [49]. Again, long-term compliance to muscle strengthening programmes can be poor [50]. Furthermore, it is not clear whether reductions in co-contraction, associated with strengthening programmes, are the result of an increase in the MVIC value used to normalise the gait EMG measurement. Interestingly, our data did not demonstrate any changes in muscle strength following the AT intervention, and this illustrates that substantial improvements in pain and function are possible without increases in strength. This finding challenges current clinical management models of knee OA which focus primarily on muscle strengthening. Co-contraction has been shown to increase compressive loads at the knee joint surface [16, 17] and increase the likelihood that a patient will opt for a knee replacement at 5 year follow up [19]. In a recent study medial, but not lateral, co-contraction was found to speed up the rate of cartilage loss in people with knee OA [18]. In line with this finding, our data showed that AT lessons led to reduced medial co-contraction during the early stance phase, a period when joint contact forces are near maximal [17]. This finding suggests that AT instruction could reduce the articular loads on the knee joint and this may have a long-term protective effect, reducing the rate of joint destruction. Interestingly, we did not observe any changes in knee kinematics or kinetics following AT instruction which indicates that, although AT lessons led to reduced co-contraction, the net moments generated at the knee joint remained constant. Although this study included participants with a range of KL grades, it was not powered to detect clinical or mechanistic differences in treatment response between different levels of disease severity. It is conceivable that individuals with less severe knee OA (lower KL grade) may derive more clinical benefit from interventions, such as the AT, which have the potential to decrease co-contraction. However, previous research has shown that co-contraction is consistently observed in individuals with knee OA, irrespective of the level of disease severity [51]. Furthermore, there is no clear link between clinical pain and the level of radiographic degeneration. Therefore, it may be the magnitude of the change in medial co-contraction which may dictate the clinical benefit, rather than the severity of the disease. However, appropriately powered controlled studies are required to confirm this idea. The data from this study support the idea of a link between a reduction in medial co-contraction and an improvement in clinical pain (Fig. 4a). However, this link was only observed when muscle activation was characterised during the pre-contact phase and not during the early stance phase of gait. This finding suggests that it may not be the net magnitude of the peak knee loads which dictates pain. Instead, there may be a more complex relationship between preparatory muscle activity, influencing active joint stiffness and rate of knee loading and the subsequent nociceptor response and perceived pain. It is has been suggested that co-contraction may be a coping response to counteract perceived knee joint instability [41], and our data may indicate that this increase in muscle activation occurs immediately before contact is made with the ground. However, there is currently debate as to whether this response should be counteracted [52]. The findings that larger reductions in clinical pain were associated with more reduction in co-contraction (Fig. 4b) supports the idea of co-contraction as a maladaptive response and suggests that medial co-contraction may prove to be an effective treatment target in people with knee OA. There were a number of limitations to this study which should be highlighted. Firstly, as the study was exploratory in nature, we did not include a patient control group. However, biomechanical data were collected data from a healthy control group allowing us to demonstrate that AT lessons led to reductions in medial co-contraction which were more in line with those observed in healthy individuals. Another limitation was that all AT lessons were delivered by a single AT practitioner and so it is not clear whether the large clinical effects observed in this study would be achieved consistently across all AT practitioners. Further large scale trials are needed to address this issue. In addition, we did not explore changes in physical activity patterns which occurred during the intervention period. However, although it is possible that changes in activity patterns influenced the observed clinical benefit, it is also conceivable that reductions in pain may lead to increases in physical activity. Finally, although we quantified muscle co-contraction, we did not use a modelling approach to precisely quantify peak contact forces at the knee. This was deemed to be beyond the scope of this study. However, our results suggest that the relationship between joint loading, muscle contraction and pain may be highly complex. Therefore, despite the omission of complex modelling techniques, we feel this study motivates new enquiry into the mechanism of pain in knee OA populations. Conclusion This study was carried out to understand the potential clinical effectiveness of the AT in the management of knee OA and also to discriminate between different potential mechanisms of therapeutic action. Following AT instruction, there was a significant reduction in knee pain and stiffness and an improvement in function which appeared to be maintained at 15 months post-baseline. The mechanistic data supported the idea that clinical changes in pain were correlated with muscular co-contraction but we did not find evidence that the AT altered central pain processing. These findings suggest reduced medial co-contraction to be a potential mechanism for improvements in pain following 12 weeks of AT. Although further research is required to fully confirm these findings, this study demonstrates the potential efficacy of interventions, such as the AT, which can successfully modify muscle activation patterns in patients with knee OA. Abbreviations OAOsteoarthritis ATAlexander Technique EMGElectromyography EEGElectroencephalogram Acknowledgements We would like to acknowledge the contribution of Peter Bloch who delivered the AT intervention to the participants in this study. We also acknowledge the financial support of the BUPA foundation who provided funding for the study. Funding This project was funded by the BUPA foundation. The funding body played no role in the design of the study, analysis/interpretation of the data or in the writing of the final manuscript. Availability of data and materials The data will not be shared since no informed consent was taken for data sharing other than within the research team. Authors’ contributions SJP conceived the original study idea, collected and processed the experimental data (related to the biomechanical measurements) and drafted the final manuscript. RKJ assisted with collecting and processing the experimental data and made substantial revisions to the final manuscript. CB collected the experimental and processed the experimental data (related to pain processing) and helped to draft the final manuscript. TC developed the experimental protocol, advised on data collection and analysis and made substantial revisions to the final manuscript. AJ conceived the original study idea, advised on data analysis and made substantial revisions to the final manuscript. All authors read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. Consent for publication Not applicable. Ethics approval and consent to participate A statement relating to ethical approval and consent was included in the methods section as follows “Before testing all subjects provided written informed consent to participate in the study and ethical approval was obtained from the NRES Greater Manchester North ethics committee, reference: 11/NW/0057.” ==== Refs References 1. Hootman JM Helmick CG Projections of US prevalence of arthritis and associated activity limitations Arthritis Rheum 2006 54 1 226 229 10.1002/art.21562 16385518 2. Woolf AD Pfleger B Burden of major musculoskeletal conditions Bull World Health Organ 2003 81 9 646 656 14710506 3. Peat G McCarney R Croft P Knee pain and osteoarthritis in older adults: a review of community burden and current use of primary health care Ann Rheum Dis 2001 60 2 91 97 10.1136/ard.60.2.91 11156538 4. Hannan MT Felson DT Pincus T Analysis of the discordance between radiographic changes and knee pain in osteoarthritis of the knee J Rheumatol 2000 27 6 1513 1517 10852280 5. Kidd B: Mechanisms of pain in osteoarthritis. HSS J. 2012;8(1):26-28. 6. Brown CA El-Deredy W Jones AKP When the brain expects pain: common neural responses to pain anticipation are related to clinical pain and distress in fibromyalgia and osteoarthritis Eur J Neurosci 2014 39 4 663 672 10.1111/ejn.12420 24219587 7. Childs JD Sparto PJ Fitzgerald GK Bizzini M Irrgang JJ Alterations in lower extremity movement and muscle activation patterns in individuals with knee osteoarthritis Clin Biomech (Bristol, Avon) 2004 19 1 44 49 10.1016/j.clinbiomech.2003.08.007 8. Hubley-Kozey CL Deluzio KJ Landry SC McNutt JS Stanish WD Neuromuscular alterations during walking in persons with moderate knee osteoarthritis J Electromyogr Kinesiol 2006 16 4 365 378 10.1016/j.jelekin.2005.07.014 16213159 9. Zeni JA Rudolph K Higginson JS Alterations in quadriceps and hamstrings coordination in persons with medial compartment knee osteoarthritis J Electromyogr Kinesiol 2010 20 1 148 154 10.1016/j.jelekin.2008.12.003 19223203 10. Heiden TL Lloyd DG Ackland TR Knee joint kinematics, kinetics and muscle co-contraction in knee osteoarthritis patient gait Clin Biomechanics 2009 24 10 833 841 10.1016/j.clinbiomech.2009.08.005 11. Mills K Hunt MA Leigh R Ferber R A systematic review and meta-analysis of lower limb neuromuscular alterations associated with knee osteoarthritis during level walking Clin Biomechanics 2013 28 7 713 724 10.1016/j.clinbiomech.2013.07.008 12. Brucini M Duranti R Galletti R Pantaleo T Zucchi PL Pain thresholds and electromyographic features of periarticular muscles in patients with osteoarthritis of the knee Pain 1981 10 1 57 66 10.1016/0304-3959(81)90045-2 7232012 13. Lyytinen T Liikavainio T Bragge T Hakkarainen M Karjalainen PA Arokoski JPA Postural control and thigh muscle activity in men with knee osteoarthritis J Electromyogr Kinesiol 2010 20 6 1066 1074 10.1016/j.jelekin.2010.05.005 20541439 14. Bouchouras G Patsika G Hatzitaki V Kellis E Kinematics and knee muscle activation during sit-to-stand movement in women with knee osteoarthritis Clin Biomechanics 2015 30 6 599 607 10.1016/j.clinbiomech.2015.03.025 15. Hortobagyi T Westerkamp L Beam S Moody J Garry J Holbert D DeVita P Altered hamstring-quadriceps muscle balance in patients with knee osteoarthritis Clin Biomechanics 2005 20 1 97 104 10.1016/j.clinbiomech.2004.08.004 16. Winby CR Gerus P Kirk TB Lloyd DG Correlation between EMG-based co-activation measures and medial and lateral compartment loads of the knee during gait Clin Biomechanics 2013 28 9 1014 1019 10.1016/j.clinbiomech.2013.09.006 17. Brandon SCE Miller RH Thelen DG Deluzio KJ Selective lateral muscle activation in moderate medial knee osteoarthritis subjects does not unload medial knee condyle J Biomech 2014 47 6 1409 1415 10.1016/j.jbiomech.2014.01.038 24581816 18. Hodges PW van den Hoorn W Wrigley TV Hinman RS Bowles K-A Cicuttini F Wang Y Bennell K Increased duration of co-contraction of medial knee muscles is associated with greater progression of knee osteoarthritis Man Ther 2015 21 151 158 10.1016/j.math.2015.07.004 26254263 19. Hubley-Kozey CL Hatfield G Stanish WD Muscle activation differences during walking between those with moderate knee osteoarthritis who progress to total knee arthroplasty and those that do not: a follow up study Osteoarthr Cartil 2013 21 S38 10.1016/j.joca.2013.02.099 20. Sofat N Ejindu V Kiely P What makes osteoarthritis painful? The evidence for local and central pain processing Rheumatology 2011 50 12 2157 2165 10.1093/rheumatology/ker283 21954151 21. Brown CA Jones AK Meditation experience predicts less negative appraisal of pain: Electrophysiological evidence for the involvement of anticipatory neural responses Pain 2010 150 3 428 438 10.1016/j.pain.2010.04.017 20494517 22. O'Reilly SC Muir KR Doherty M Effectiveness of home exercise on pain and disability from osteoarthritis of the knee: a randomised controlled trial Ann Rheum Dis 1999 58 1 15 19 10.1136/ard.58.1.15 10343535 23. Maurer BT Stern AG Kinossian B Cook KD Schumacher HR Osteoarthritis of the knee: Isokinetic quadriceps exercise versus an educational intervention Arch Phys Med Rehabil 1999 80 10 1293 1299 10.1016/S0003-9993(99)90032-1 10527090 24. Peloquin L Bravo G Gauthier P Lacombe G Billiard JS Effects of a cross-training exercise program in persons with osteoarthritis of the knee a randomized controlled trial Jcr-J Clin Rheumatol 1999 5 3 126 136 10.1097/00124743-199906000-00004 25. Messier SP Legault C Mihalko S Miller GD Loeser RF DeVita P Lyles M Eckstein F Hunter DJ Williamson JD The Intensive Diet and Exercise for Arthritis (IDEA) trial: design and rationale BMC Musculoskelet Disord 2009 10 93 10.1186/1471-2474-10-93 19638215 26. Ettinger WH Burns R Messier SP Applegate W Rejeski WJ Morgan T Shumaker S Berry MJ Otoole M Monu J A randomized trial comparing aerobic exercise and resistance exercise with a health education program in older adults with knee osteoarthritis - The Fitness Arthritis and Seniors Trial (FAST) Jama-J Am Med Assoc 1997 277 1 25 31 10.1001/jama.1997.03540250033028 27. Rogind H Bibow-Nielsen B Jensen B Moller HC Frimodt-Moller H Bliddal H The effects of a physical training program on patients with osteoarthritis of the knees Arch Phys Med Rehabil 1998 79 11 1421 1427 10.1016/S0003-9993(98)90238-6 9821904 28. Fransen M, McConnell S, Harmer AR, Van der Esch M, Simic M, Bennell KL: Exercise for osteoarthritis of the knee. Cochrane Database of Systematic Reviews. 2015;(1). 29. Cacciatore TW Horak FB Henry SM Improvement in automatic postural coordination following alexander technique lessons in a person with low back pain Phys Ther 2005 85 6 565 578 15921477 30. Woodman JP Moore NR Evidence for the effectiveness of Alexander Technique lessons in medical and health-related conditions: a systematic review Int J Clin Pract 2012 66 1 98 112 10.1111/j.1742-1241.2011.02817.x 22171910 31. Little P Lewith G Webley F Evans M Beattie A Middleton K Barnett J Ballard K Oxford F Smith P Randomised controlled trial of Alexander technique lessons, exercise, and massage (ATEAM) for chronic and recurrent back pain BMJ (Clin Res ed 2008 337 a884 10.1136/bmj.a884 32. MacPherson H Tilbrook H Richmond S Woodman J Ballard K Atkin K Bland M Eldred J Essex H Hewitt C Alexander Technique Lessons or Acupuncture Sessions for Persons With Chronic Neck Pain A Randomized Trial Ann Intern Med 2015 163 9 653 10.7326/M15-0667 26524571 33. Stallibrass C Sissons P Chalmers C Randomized controlled trial of the Alexander technique for idiopathic Parkinson's disease Clin Rehabil 2002 16 7 695 708 10.1191/0269215502cr544oa 12428818 34. Cacciatore TW Gurfinkel VS Horak FB Day BL Prolonged weight-shift and altered spinal coordination during sit-to-stand in practitioners of the Alexander Technique Gait Posture 2011 34 4 496 501 10.1016/j.gaitpost.2011.06.026 21782443 35. Cacciatore TW Gurfinkel VS Horak FB Cordo PJ Ames KE Increased dynamic regulation of postural tone through Alexander Technique training Hum Mov Sci 2011 30 1 74 89 10.1016/j.humov.2010.10.002 21185100 36. Cacciatore TW Mian OS Peters A Day BL Neuromechanical interference of posture on movement: evidence from Alexander technique teachers rising from a chair J Neurophysiol 2014 112 3 719 729 10.1152/jn.00617.2013 25085609 37. Hurley MV Walsh NE Mitchell H Nicholas J Patel A Long-term outcomes and costs of an integrated rehabilitation program for chronic knee pain: A pragmatic, cluster randomized, controlled trial Arthritis Care Res 2012 64 2 238 247 10.1002/acr.20642 38. Bellamy N Buchanan WW Goldsmith CH Campbell J Stitt LW Validation study of WOMAC - A health status instrument for measuring clinically importatnt patient relevant outcomes to antirhuematic drug therapy in patients with osteoarthritis of the hip or knee J Rheumatol 1988 15 12 1833 1840 3068365 39. Hof AL Elzinga H Grimmius W Halbertsma JPK Speed dependence of averaged EMG profiles in walking Gait Posture 2002 16 1 78 86 10.1016/S0966-6362(01)00206-5 12127190 40. Hermens HJ Freriks B Disselhorst-Klug C Rau G Development of recommendations for SEMG sensors and sensor placement procedures J Electromyogr Kinesiol 2000 10 5 361 374 10.1016/S1050-6411(00)00027-4 11018445 41. Lewek MD Ramsey DK Snyder-Mackler L Rudolph KS Knee stabilization in patients with medial compartment knee osteoarthritis Arthritis Rheum 2005 52 9 2845 2853 10.1002/art.21237 16142714 42. Jones RK Chapman GJ Parkes MJ Forsythe L Felson DT The effect of different types of insoles or shoe modifications on medial loading of the knee in persons with medial knee osteoarthritis: a randomised trial J Orthop Res 2015 33 11 1646 1654 10.1002/jor.22947 25991385 43. Bell AL Brand RA Pedersen DR Prediction of hip-joint center location from external landmarks Hum Mov Sci 1989 8 1 3 16 10.1016/0167-9457(89)90020-1 44. Brown CA Seymour B Boyle Y El-Deredy W Jones AK Modulation of pain ratings by expectation and uncertainty: Behavioral characteristics and anticipatory neural correlates Pain 2008 135 3 240 250 10.1016/j.pain.2007.05.022 17614199 45. Legrain V Guerit JM Bruyer R Plaghki L Attentional modulation of the nociceptive processing into the human brain: selective spatial attention, probability of stimulus occurrence, and target detection effects on laser evoked potentials Pain 2002 99 1–2 21 39 10.1016/S0304-3959(02)00051-9 12237181 46. Kellgren JH Lawrence JS Radiological assessment of osteo-arthrosis Ann Rheum Dis 1957 16 4 494 502 10.1136/ard.16.4.494 13498604 47. Ramsey DK Briem K Axe MJ Snydcr-Mackler L A mechanical theory for the effectiveness of bracing for medial compartment osteoarthritis of the knee J Bone Joint Surg-Am Vol 2007 89A 11 2398 2407 10.2106/JBJS.F.01136 48. McQuade KJ de Oliveira AS Effects of progressive resistance strength training on knee biomechanics during single leg step-up in persons with mild knee osteoarthritis Clin Biomechanics 2011 26 7 741 748 10.1016/j.clinbiomech.2011.03.006 49. Al-Khlaifat L Herrington LC Hammond A Tyson SF Jones RK The effectiveness of an exercise programme on knee loading, muscle co-contraction, and pain in patients with medial knee osteoarthritis: A pilot study Knee 2015 23 1 63 69 10.1016/j.knee.2015.03.014 25953672 50. Mazieres B Thevenon A Coudeyre E Chevalier X Revel M Rannou F Adherence to, and results of, physical therapy programs in patients with hip or knee osteoarthritis. Development of French clinical practice guidelines Joint Bone Spine 2008 75 5 589 596 10.1016/j.jbspin.2008.02.016 18805033 51. Hubley-Kozey CL Hill NA Rutherford DJ Dunbar MJ Stanish WD Co-activation differences in lower limb muscles between asymptomatic controls and those with varying degrees of knee osteoarthritis during walking Clin Biomechanics 2009 24 5 407 414 10.1016/j.clinbiomech.2009.02.005 52. Steultjens M Dekker J van der Esch M The pros and cons of muscle co-contraction in osteoarthritis of the knee: comment on the article by Lewek et al Arthritis Rheum 2006 54 4 1354 10.1002/art.21781 16575865

PMC005xxxxxx/PMC5002320.txt

==== Front J Exp Clin Cancer ResJ. Exp. Clin. Cancer ResJournal of Experimental & Clinical Cancer Research : CR0392-90781756-9966BioMed Central London 40410.1186/s13046-016-0404-1ResearchBrain stereotactic biopsy flow cytometry for central nervous system lymphoma characterization: advantages and pitfalls Cordone Iole iole.cordone@ifo.gov.it 1Masi Serena serena.masi@ifo.gov.it 1Carosi Mariantonia mariantonia.carosi@ifo.gov.it 2Vidiri Antonello antonello.vidiri@ifo.gov.it 3Marchesi Francesco marchesi.francesco@tiscali.it 4Marino Mirella mirella.marino@ifo.gov.it 2Telera Stefano stefano.telera@ifo.gov.it 5Pasquale Alessia alessia.pasquale@ifo.gov.it 1Mengarelli Andrea andrea.mengarelli@ifo.gov.it 4Conti Laura laura.conti@ifo.gov.it 1Pescarmona Edoardo edoardo.pascarmona@ifo.gov.it 2Pace Andrea andrea.pace@ifo.gov.it 6Carapella Carmine M. +39.06.52665595carmine.carapella@ifo.gov.it 51 Clinical Pathology, Regina Elena National Cancer Institute, Rome, Italy 2 Pathology, Regina Elena National Cancer Institute, Rome, Italy 3 Radiology, Regina Elena National Cancer Institute, Rome, Italy 4 Hematology and Stem Cell Transplant, Regina Elena National Cancer Institute, Rome, Italy 5 Neurosurgery, Regina Elena National Cancer Institute, Rome, Italy 6 Neuroncology, Regina Elena National Cancer Institute, Rome, Italy 27 8 2016 27 8 2016 2016 35 1 12831 5 2016 9 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Brain stereotactic biopsy (SB) followed by conventional histopathology and immunohistochemistry (IHC) is the gold standard approach for primary central nervous system lymphoma (PCNSL) diagnosis. Flow cytometry (FCM) characterization of fine-needle aspiration cytology and core needle biopsies are increasingly utilized to diagnose lymphomas however, no biological data have been published on FCM characterization of fresh single cell suspension from PCNSL SB. The aim of this study was to establish the feasibility and utility of FCM for the diagnosis and characterization of brain lymphomas from a tissue samples obtained by a single SB disaggregation. Methods Twenty-nine patients with a magnetic resonance suggestive for PCNSL entered the study. A median of 6 SB were performed for each patient. A cell suspension generated from manual tissue disaggregation of a single, unfixed, brain SB, was characterized by FCM. The FCM versus standard approach was prospectively compared. Results FCM and IHC showed an high degree of agreement (89 %) in brain lymphoma identification. By FCM, 16 out of 18 PCNSL were identified within 2 h from biopsy. All were of B cell type, with a heterogeneous CD20 mean fluorescence intensity (MFI), CD10 positive in 3 cases (19 %) with surface Ig light chain restriction documented in 11 cases (69 %). No false positive lymphomas cases were observed. Up to 38 % of the brain leukocyte population consisted of CD8 reactive T cells, in contrast with the CD4 positive lymphocytes of the peripheral blood samples (P < 0.001). By histopathology, 18 B-PCNSL, only one CD10 positive (5 %), 1 primitive neuroectodermal tumor (PNET) and 10 gliomas were diagnosed. A median of 6 days was required for IHC diagnosis. Conclusion Complementary to histopathology FCM can contribute to a better characterization of PCNSL, although necrosis and previous steroid treatment can represent a pitfall of this approach. A single brain SB is a valid source for accurate FCM characterization of both lymphoma and reactive lymphocyte population, routinely applicable for antigen intensity quantification and consistently documenting an active mechanism of reactive CD8 T-lymphocytes migration in brain lymphomas. Moreover, FCM confirmed to be more sensitive than IHC for the identification of selected markers. Keywords PCNSLBrain stereotactic biopsyFlow cytometryTumor side populationThis work was supported by intramural funds provided by the Scientific Directors Office of the Regina Elena National Cancer Institute06/RC/06Cordone Iole issue-copyright-statement© The Author(s) 2016 ==== Body Background Primary central nervous system lymphoma (PCNSL), a rare hematological disease, represents a big challenge for clinicians and researchers [1, 2]. Despite improved therapeutic approaches, the diagnostic - prognostic characterization has documented little improvement and long-term response to treatment is rare [3, 4]. More than 90 % are diffuse large B-cell lymphomas (DLBCL) derived from a late germinal center B cell [5]. PCNSL is usually a widespread - diffusely infiltrating lymphoma. Thereafter, efforts at resection are discouraged, although resection may be associated with prolonged survival in surgically safe, single, PCNSL lesions [6]. Hence, stereotactic biopsy (SB) followed by conventional histopathology and immunohistochemistry (IHC) is the gold standard diagnostic approach useful to perform a differential diagnosis among other brain lesions, such us gliomas [7–11]. Flow cytometry (FCM) is a proven valuable diagnostic tool for the diagnosis and monitoring of hematological malignancies in routine clinical practice [12, 13]. More recently, FCM immunophenotype of fine-needle aspiration and/or core needle biopsy from lymphoid tissues has shown to improve diagnostic accuracy in B-cell non-Hodgkin lymphomas classification [14–18] and FCM significantly increases the sensitivity and specificity of cerebrospinal fluid (CSF) infiltration detection in onco-hematology [19–22]. However, no biological data have been published on FCM immunophenotype of fresh single-cell suspension obtained from SB of brain lymphomas. We performed a feasibility study to evaluate the diagnostic and potential added value of FCM characterization of a single SB taken in vivo from intra-cerebral suspected lymphomas. FCM was compared with conventional IHC results in terms of diagnosis reliability and sensitivity in cell markers identification. The membrane CD20 mean fluorescence intensity (MFI) was evaluated to document, in PCNSL, the intensity of expression of a surface molecule potentially related to clinical response to Rituximab [23]. Finally, the tumor lymphoid side-population was characterized and compared to the peripheral blood (PB) lymphocytes phenotype, to investigate the blood-brain barrier permeability to non-malignant lymphocytes in CNS lymphomas. Methods Patients Twenty-nine patients who underwent diagnostic SB for CNS tumor at Regina Elena National Cancer Institute were evaluated. All cases entered the study because of a magnetic resonance suggestive for PCNSL. About 50 % of enrolled patients underwent a previous steroid treatment to limit the mass-related cerebral edema before SB. The study was approved by our Institutional Ethics Committee and a signed informed consent was obtained from all the patients. Contrast-enhanced magnetic resonance Magnetic Resonance Imaging (MRI) was performed on a 1.5-T system (Optima MR 450w, GE Health-care, Milwaukee, Wisconsin) with dedicated 16-channels receive-only RF coils; slice thickness 3 mm and matrix size of 512 × 512 were used. Before contrast medium infusion Spin-echo (SE) sequences T1, T2 and FLAIR in axial (T1, T2 e FLAIR) e cornal (T2) planes were performed. After contrast medium infusion (0.1 mmol/Kg body weight of gadopentetate dimeglumine contrast agent was administered intravenously, at a rate of 2 ml/s.) a volumetric contrast-enhanced T1-weighted GRE sequence was obtained (TR/TE = 7.8 ms/3.2 ms, matrix size = 512 × 512, voxel size = 0.5 x 0.5x 0.5 mm3) in axial plane with a reconstruction in coronal and sagittal planes to obtain very thin section with fat-suppressed images. Diffusion Imaging (DWIs) were obtained by single-shot SE EPI using two b value (0-1000). The MRI criteria for subject inclusion into the study were: 1) single or multiple lesions with periventricular sites; 2) low signal intensity on T2 weighted sequences; 3) homogeneously enhancement mass; 4) marked oedema; 5) hyperintense on DWI and hypointense on ADC maps; 6) enhancement of cranial nerve with increase nerve diameter [10] (Fig. 1). The MRI findings were evaluated by two neuroradiologists with 20 and 10 years of experience, respectively.Fig. 1 MRI findings strongly suggestive for brain lymphoma: SE T2 sequences in coronal plane (a), diffusion in axial plane (b, c) and fat-suppressed T1 after contrast medium infusion in axial plane (d, e). MRI shows multiple bilateral, periventricula sites with low signal intensity on T2, with oedema, hyperintensity signal on DWI and homogensously enhancement after contrast medium infusion Flow cytometry on brain stereotactic needle biopsy Brain SB were performed using the Cosman-Roberts-Wells frame (Radionics, Burlington, Massachusetts). A median of 6 (range 5-8) tissue biopsies, with a medium size of 1 x 10 mm, were obtained. Immediately after extraction, each specimen was fixed with formalin for pathological evaluation and one biopsy was collected in a tube without any transport medium for FCM analysis. The cell suspension was generated from a single, unfixed, SB fragment by gentle manual disaggregation in a sterile Petri dish containing 2-4 ml of phosphate buffered saline (PBS, Gibco by Life Technologies, UK) using a 10-mL syringe plunger rod. The released cells were collected and washed twice in PBS for 5 min at 1600 revolutions per minute. The supernatant was discarded and the pellet of cells was suspended in PBS and stained, according to the manufacturer’s recommendations, using a 6-colour panel of antibodies (Fitc/PE/PerCP/PE-Cy7/APC/APC-Cy7) and the “Duo-lyse” program of the Becton Dickinson Bioscience (BDB) Lyse-Wash-Assistant according to the following combinations: 1) CD3/CD56/CD45/CD4/CD19/CD8; 2) CD5/CD10/CD45/CD2/CD79b/CD20; 3) surface immunoglobulin (sIg) Ig lambda/sIg kappa/CD45/CD34/CD22/CD14, all from Becton Dickinson Bioscience (BDB). The FCM characterization was performed within 1 to 2 h from biopsy. The whole volume of sample was acquired and analyzed using a BDB FACSCanto flow cytometer and BDB theFACSDiva software. The CD45 positive cells were gated and the percentage of B, T and myeloid population were evaluated. Surface Ig light chain expression was evaluated on the CD22 positive population; CD4 and CD8 expression was evaluated as the percentage of total CD3 positive lymphocytes. The markers expression was reported as percentage of positive cells within the CD45 positive population. The reactive population of small lymphocytes was utilized as positive/negative internal control. The MFI ratio for the CD20 antigen was calculated by comparison with negative control. An immunoperoxidase technique on fixed cytospins was utilized to investigate the presence of intra-cytoplasmic Ig light chains expression in the surface negative cases, as previously described [24]. Immunofenotype was also performed on CSF samples of 5 PCNSL cases. A volume of 3.5 mL (range 2.8-10) of CSF was collected in a tube without any transport medium and processed within a few (1-3) hours from collection to minimize cell loss. Flow cytometry on peripheral blood samples The peripheral blood T/NK lymphocytes subset was evaluated by FCM according to the following combination: CD3/CD56/CD45/CD4/CD19/CD8. The lymphocytes phenotype was evaluated gating on the CD45 bright lymphoid cells. The CD4 and CD8 subset was evaluated on the percentage of total CD3 positive T lymphocytes. Histopathology All tumors were classified according to the World Health Organization (WHO) classification [25] by conventional histology (H&E) and IHC on formalin-fixed, paraffin embedded SB tissue, utilizing the following antibodies: CD20, CD10 (clone 56C6), CD138, CD5, (Novocastra, Menarini, Florence, Italy), CD3, CD45 (CLA), CD79a, Ki-67, CD34 (QBEND clone), CD68 (KP1) p53 and GFAP (Dako Milan, Italy), CD56 (NCAM) (UCS Diagnostic, Rome, Italy) by a streptavidin-biotin enhanced immunoperoxidase technique, according to the manufacturer’s recommendations. The study was conducted in double-blind, with the hematopathologists and the cytometrists not aware of each other’s interpretation. Statistical analysis Concordance analysis was performed to assess the degree of agreement (concordance) between FCM and IHC in PCNSL diagnosis. Fisher’s exact test was conducted to evaluate the different distribution between brain and PB lymphoid T sub-populations. Results Patients Twenty-nine patients entered the study. The median age was 60 years (range 14–81), 16 patients were male and 13 female; all were HIV negative. Histopathology Eighteen PCNSL, 1 primitive neuroectodermal tumor (PNET) and 10 gliomas were diagnosed by histopathology. By IHC, all PCNSL were CD45 CD20 CD79a positive, CD34 negative, one case was CD10 positive with a low intensity of expression, all with >70 % Ki-67 positive cells. A minority (5 to 10 %) of CD3 and CD5 reactive T lymphocytes scattered between the large lymphoma B cells was documented in 17 cases. By contrast, more than 30 % of the leucocytes was represented by a reactive perivascular T-cell infiltration (RPVI) in one case, as confirmed by FCM (Table 1, case n° 2). In 2 cases a diffuse necrosis with a small residual foci of lymphoma was documented by IHC. The PNET was CD56/CD34 positive CD45 negative, with a significant proportion of histiocytes and a minority of T lymphocytes among the tumor cells. The ten cases of gliomas were histologically classified as high-grade and low-grade in 8 and 2 cases, respectively. A median time of 6 days (5–8) was required between biopsy and diagnosis.Table 1 Flow cytometry analysis of the CD45+ population in 16 PCNSL stereotactic biopsy shows a prevalence of large B cells sided by reactive T lymphocytes and monocytes N° % Large B Lymphoma Cells % Small reactive T lymphocytes % Myeloid cells Case n° Histological type CD45+ cells CD19+ CD79b+ CD10+ CD19/Kappa + light chain CD19/Lambda + light chain CD3+ CD4dim or CD14+ 1 DLBCL 42864 75 70 1 2 3 10 8 2 DLBCL 44073 60 65 0 93 2 38 1 3 DLBCL 91746 65 61 0 2 1 18 9 4 DLBCL 78828 80 76 70 85 0 8 3 5 DLBCL 66480 85 0 0 0 0 11 1 6 DLBCL 91726 95 92 0 2 96 5 0 7 DLBCL 113162 96 93 0 90 0 1 3 8 DLBCL 132550 96 91 0 0 83 1 3 9 DLBCL 38657 96 89 0 97 0 3 1 10 DLBCL 55841 70 68 0 96 0 24 1 11 DLBCL 27870 81 78 0 96 1 20 3 12 DLBCL 85000 95 92 0 0 0 5 4 13 DLBCL 61640 77 70 95 95 0 21 4 14 DLBCL 72001 88 12 0 0 0 10 2 15 DLBCL 30660 88 10 83 0 97 10 2 16 DLBCL 49523 72 26 0 0 98 26 2 DLBCL diffuse large B-cell lymphoma Brain stereotactic biopsy characterization by flow cytometry PCNSL markers expression Sixteen out of 18 PCNSL (89 %) were identified by FCM. Leukocyte population was characterized within 2 h from SB, with a median of 66059 (range 27870–132550) CD45 positive cells and a proportion of 60 % to 96 % tumor B cells in all lymphomas (Table 1). Unequivocal identification of large lymphoma CD19 positive B cells (green color), CD3 small reactive T lymphocytes (orange color) and CD45 negative CD56 positive brain/neuro-ectodermal cells (blue color) was obtained in a single tube analysis (Fig. 2). All PCNSL were CD19/CD22 positive, CD20 positive in 15/16 cases (92 %), CD79b positive in 13/16 cases (81 %), CD5-CD34 negative, with a surface Ig light chain restriction detected in 11 cases (69 %), 7 Ig Kappa and 4 Ig Lambda light chain, respectively (Table 1). No intra-cytoplasmic light chain expression was documented by immunoperoxidase technique in the surface negative cases. Three cases were CD10 positive with surface Ig light chain restriction, consistent with the germinal center origin of the lymphoma. The CD20 expression was heterogeneous, with a median MFI ratio of 24.13 (range 7.55–58.55). One case was CD20 negative by FCM and weak positive by IHC. In 2 cases (11 %) FCM failed to identify a CNS lymphoma. Focusing on the 2 brain lymphomas not identified by FCM, a diffuse necrosis with a small residual foci of lymphoma was documented by IHC in both cases. In one case, 4 out of 8 SB were negative for PCNSL by IHC; in the second case, a prominent necrosis (>50 % of the available tissue) was documented in all 8 biopsies evaluated by histopatology. Both of failed cases occurred in patients who underwent a previous steroid treatment.Fig. 2 Flow cytometry analysis of cell suspension from one brain stereotactic needle biopsy of PCNSL. Blue color has been utilized to mark CD56 positive/CD45 negative brain cells (a, b, d), green for CD10 CD19 CD20 sIg-Kappa light chain positive large lymphoma cells (c, e, f), orange for the side population of CD3 positive T lymphocytes (a, b, c) Tumor-side population phenotype From 1 % to 38 % of the leukocyte population was represented by CD2/CD3/CD5 positive reactive T lymphocytes with a prevalence of CD8 positive cells documented in 13/16 PCNSL cases (81 %) and more than 40 % of CD56 positive T cells in two (Table 2, cases n°1 and n°14). A minority (<10 %) of CD4 dim CD14 positive monocytes among the lymphoma cells was also documented.Table 2 Flow cytometry analysis of the lymphoid T reactive population from 16 PCNSL SB and corresponding peripheral blood samples McAb % and CD4/CD8 ratio of reactive lymphocytes in SB and PB samples Case n° 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Brain Streotactic Biopsy CD3 84 86 91 85 89 98 98 98 95 94 95 98 85 97 87 85 CD5 75 79 91 75 89 95 97 98 83 96 95 94 85 95 78 85 CD56 60 2 7 4 4 1 1 5 3 6 1 1 2 40 1 1 CD3/CD4 31 41 45 26 39 6 18 35 25 59 69 18 60 14 34 32 CD3/CD8 87 57 50 59 47 72 60 42 74 28 16 71 33 56 53 45 CD3/CD56 57 4 6 2 6 6 17 5 2 7 1 6 5 42 9 4 T4/T8 ratio 0.35 0.72 0.9 0.44 0.83 0.08 0.3 0.83 0.34 2.1 4.3 0.25 1.8 0.25 0.64 0.71 Peripheral blood samples CD19 nd nd nd 12 16 40 16 59 nd 29 19 15 nd nd 11 9 CD3 nd nd nd 64 63 48 55 16 nd 48 54 72 nd nd 69 89 CD56 nd nd nd 7 21 15 20 17 nd 23 25 20 nd nd 20 11 CD3/CD4 nd nd nd 73 59 50 79 66 nd 37 51 58 nd nd 53 57 CD3/CD8 nd nd nd 23 36 46 19 32 nd 62 47 38 nd nd 43 38 T4/T8 nd nd nd 3.17 1.6 1.08 4.15 2 nd 0.59 1.08 1.5 nd nd 1.23 1.5 nd not done PCNSL negative cases A prevalence of CD45 negative, CD56 positive cells, consistent with a population of neuro-ectodermal origin, was identified in 13 cases, flanked by a minority of CD3 positive reactive T lymphocytes. In the PNET case, a prevalence (70 %) of CD45 negative, CD56/CD34 positive cells was documented flanked by a population of CD45/CD4dim/CD14 positive monocytes and CD4 reactive T lymphocytes. Comparing FCM with IHC by concordance analysis, a statistically significant agreement (P = 0.0034) with a Cohen’s kappa coefficient of 0.859 was found. CSF flow cytometry of PCNSL cases Lumbar puncture yielded adequate material for FCM analysis in all the cases analyzed. The CSF samples had a median cell count of 3 cell/μL (range 1-5). Despite the low CSF absolute cell number, a median of 6229 (range 1436-8285) evaluable cells were analyzed. The cell population was represented by a prevalence of CD4 positive T lymphocytes (76 %; range 65–93) in all the cases, sided by monocytes (CD14 positive = 6 %; range 1–13). In one case, despite the low cell count (1 cell/μL), 30 % of the lymphoid population was represented by clonal B cell, CD19 CD20 CD22 positive, CD5 CD10 negative, with surface Ig lambda light chain restriction (case n°6). Peripheral blood lymphoid asset The peripheral blood T/NK lymphocytes subset was evaluated on 10/16 PCNSL. A lymphocytopenia, 900 cell/μL (range 350–1800), was documented. In contrast with the brain CD8 T lymphocytes prevalence (81 %), a population of CD4 positive T cells was documented in 9/10 (90 %) PB samples (P < 0.001) (Table 2). Discussion This is a feasibility, prospective study that describes for the first time the use of FCM applied on a cohort of brain SB samples, providing evidence that a single SB is a valid source for identification and characterization of brain lymphomas and reactive infiltrating leucocytes, with a significant agreement between the FCM and IHC diagnosis (P = 0.0034). Despite the magnetic resonance strongly suggestive for PCNSL in 29 selected cases, 18 PCNSL, 10 gliomas and 1 PNET were diagnosed by IHC. FCM identified 16 out of 18 (89 %) PCNSL cases. Thereafter, 2 brain lymphomas (11 %) were missed by FCM. The overall experience indicates that FCM has difficulty detecting high-grade lymphomas, particularly DLBCL. Failure must be assigned to necrosis, cell fragility or inadequate sample size [15]. Moreover, necrosis can represent a major problem in PCNSL diagnosis after corticosteroid administration, as shown by a recently published study [26], considering that a large portion of newly diagnosed PCNSL patients often underwent a steroid treatment just before diagnosis, for clinical need. According to the study design, a single biopsy was analyzed by FCM while 6–8 samples were utilized for the IHC characterization. Nevertheless, the manual processing technique for tissue disaggregation allowed a successful FCM identification of brain lymphomas in 89 % of cases, confirming the limits of a correct diagnosis in patients pre-treated with steroids. Our study shows that a single SB is an adequate sample size for a comprehensive PCNSL characterization in absence of tissue necrosis. In fact, in the 2 PCNSL cases not identified by FCM the necrosis was the main factor. A diffuse necrosis with a small residual foci of lymphoma was documented by IHC in both cases, confirming the relevance of necrosis in hampering the diagnosis in large cell lymphomas. It is likely that performing FCM analysis on more than a single bioptic sample the rate of success in FCM brain lymphomas identification would increase. FCM is a proven valuable diagnostic tool in hematological CSF infiltration detection [27, 28], increasing the sensitivity and specificity of leptomeningeal disease identification in CNS lymphomas [22]. We have recently documented that FCM can discriminate between reactive and neoplastic plasma cells in CSF samples with very low cell counts, confirming to be significantly more sensitive than standard approaches [21]. CSF FCM was performed on 5 out of 18 PCNSL cases and, despite the low cell count, a clonal B cell population was identified in one case. However, CSF cytology was negative and, although FCM was indicative of a B-cell clonal disorder, the data was not strong enough to reach the diagnosis of PCNSL. These findings are in agreement with previously published data [22]. A very wide panel of markers could be tested by FCM in a single brain SB, FCM demonstrate to be more sensitive than IHC for the identification of relevant markers. CD10 has been reported to be 10 times lower in PCNSL comparing to systemic diffuse large B-cell lymphomas [29]. In our study, the CD10 expression was documented in 3/16 PCNSL (19 %) by FCM and 1/18 (5 %) by IHC. The reason for this discordance is uncertain. Probably the different clones utilized for FCM (clone HI10a) and IHC (clone 56C6) could be involved in this matter. However, a different sensitivity between FCM and IHC in antigen identification is widely documented. Our report documents a possible underestimation of CD10 frequency of expression in PCNSL by IHC staining, highlighting the added value provided by performing FCM in terms of a more reliable technique for the identification of selected markers. The panel of antigens used was in accordance with the WHO guidelines [25]. However, more recently, a significant differential expression of CD44 between Burkitt lymphoma and CD10 DLBCL has been documented, making CD44 an excellent candidate for rapid immunophenotypic discrimination between these two types of lymphoma [30]. Flanking IHC studies, FCM characterization, enriched by new relevant diagnostic markers, can significantly improve the diagnostic approach to PCNSL. Surface Ig light chain restriction was documented in 11 cases. Immunoglobulin light chains restriction is an important adjunct to the use of surface markers in onco-hematology [13]. However, light chain restriction is not always documented on high grade B-cells NHL. In our series 5 cases were both surface and intra-cytoplasmic negative. Despite the lack of light chain restriction, the high FSC/SSC scatter of the B cell population (large cells) compared to the small T lymphocytes as well as the prevalence of B cells (85 %, range 60-96) compared to the T subset (10 %, range 1-38), strongly supported the diagnosis of a B cell neoplasm, confirmed by histopathology in all the cases. Moreover, since Ig light chain expression in non-Hodgkin lymphomas (NHL) is difficult to identify by IHC, cytometry can easily represent the method of choice for clonality detection in PCNSL. The presence of a reactive immune response and perivascular T-cell infiltrate within the lymphoma has been shown to represent a prognostic biomarker and influence disease outcome in PCNSL [31, 32]. Antigen co-expression, mandatory for the classification of T and NK lymphocyte sub-populations, is easy to identity by FCM while difficult to assess by IHC. In this study, we documented a significant difference in lymphocyte sub-set distribution between brain lymphoma and peripheral blood. A prevalence of CD8 reactive T lymphocytes was documented in the majority (81 %) of PCNSL cases, while a prevalence of CD4 positive lymphocytes was identified in all but one peripheral blood samples (90 %) (P < 0.001). CNS is an immunogical sanctuary with restricted access and an unique microenvironment. The present study consistently shows that the blood-brain barrier actively selects a sub-population of CD8 non-malignant cells in PCNSL, providing a promising rationale for the investigation of cellular immunotherapy in brain tumors. To date, there is an increasing interest in engineered chimeric antigen receptor (CAR) T-cells in hematologic malignancies, also in NHL [33]. In this point of view, exploiting their migration to pathological lesions, reactive T-cells could be used for adoptive immunotherapy strategies based, for example, on targeting brain lymphoma CD19 positive B-cells by bi-specific T-cell engager monoclonal antibodies or engineered CAR T-cells in preclinical models. Tumor associated macrophages, commonly present in NHL, have been related to poor prognosis and are potentially involved in the pathogenesis of brain NHL [34]. We identified a minority of CD4 dim CD14 positive monocytes among the B lymphoma cells and a great amount in the PNET case. Biological studies, aimed at a better understanding of the role of the tumor associated reactive population, would easily benefit from the FCM approach. The identification of new therapeutic strategies is of prominent importance in PCNSL and FCM characterization can be a valuable tool for the detection of potential therapeutic targets and orientate tailored antibody-based therapies. The CD20 B cell-specific antigen is a membrane-bound protein whose expression is very heterogeneous among different B lymphoma subtypes [35] and significantly associated with disease prognosis [36, 37]. This difference may correlate with clinical response to Rituximab, a monoclonal antibody directed against the CD20 antigen that, in combination with conventional chemotherapy, has dramatically improved the survival of patients with diffuse large B-cell lymphoma [3]. By contrast, reduced CD20 expression has been associated with an inferior survival, suggesting that the evaluation of cell surface antigen intensity of expression can improve treatment tailoring of antibody-based therapy [38]. Anti-CD20 antibody therapy has shown to improve disease control and outcome in patients with primary CNS lymphoma [23, 39]. No biological data, however, is available regarding the intensity of surface antigen expression in intra-cranial B lymphomas. An objective quantification of antigen intensity of expression can be easily performed by FCM by the difference between antigen MFI and negative control. This technique is practical for use in routine setting. Utilizing this approach, our study documents for the first time a marked variability between surface CD20 expression level in brain lymphomas, ranging from very “bright” to “dim” in MFI with one CD20 negative case by FCM. CD20 (L26) staining is not usually graded by pathologists, and hematopathologists report all lymphomas that react with L26 antibody as “CD20 positive” regardless of expression level [40]. Therefore, the determination of CD20 MFI could represent a novel biomarker for a better stratification of PCNSL patients more likely to benefit from Rituximab treatment, for a more effective tailoring of the first line therapy. Actually, this hypothesis should be confirmed in further prospective studies. A prevalence of CD45 negative CD56 positive cells, consistent with a neuro-ectodermal population [41], was documented in 13 cases, with a strong CD34 expression identified in the PNET case, suggesting a potential role of FCM in non hematological tumors. Conclusions This paper provides evidence that a single brain stereotactic core biopsy is a valid source for rapid identification and accurate FCM characterization of CNS lymphomas. FCM has shown to be more sensitive than IHC for selected markers, easily applicable for antigen intensity quantification and, documenting a prevalence of CD8 reactive T lymphocytes in brain lymphomas, suggests an active mechanism of lymphoid migration through the blood-brain barrier, potentially providing relevant information for further therapeutic decision making. In the contest of the gold standard approach, the advantages of FCM over IHC include an excellent sensitivity, a short execution time, an objective quantification of antigen intensity of expression, a reliable approach for the evaluation of antigens co-expression on both tumor and tumor-side populations. The presence of necrosis remains a major disadvantage, as well as the requirement of fresh tissue for analysis. Finally, establishing a provisional diagnosis of brain lymphoma, FCM could positively affect the cost-benefit ratio, contributing in a shortening of time-to-treat interval with reduction of hospitalization indirect costs. Prospective studied will establish the synergic and complementary role of these two techniques for a better diagnostic approach and clinical management of brain lesions. Abbreviations PCNSLPrimary central nervous system lymphoma DLBCLDiffuse Large B-Cell Lymphoma SBStereotactic biopsy IHCImmunohistochemistry CNSCentral Nervous System FCMFlow Cytometry CSFCerebrospinal Fluid MFIMean Fluorescence Intensity PBPeripheral Blood MRIMagnetic Resonance Imaging SESpin-echo DWISDiffusion Imaging PBSPhosphate Buffered Saline BDBBecton Dickinson Bioscience WHOWorld Health Organization PNETPrimitive Neuroectodermal Tumor RPVIReactive Perivascular T-cell Infiltration NHLNon-Hodgkin Lymphoma CARChimeric Antigen Receptor We thank Claudia Abbruzzese for the statistical analysis and Prof. Barbara Bain for careful review of the manuscript. Dedication This article is dedicated to the memory of Emanuele Occhipinti, past-Director of Neurosurgery in our Institute, who confidently encouraged and enthusiastically supported this project. Funding This work was supported by intramural funds (06/RC/06) provided by the Scientific Director’s Office of the Regina Elena National Cancer Institute in Rome, Italy. Disclosure All authors report no disclosures. Authors’ contributions IC conceived and designed the study, developed, analyzed and interpreted the flow cytometry studies, drafted the manuscript; SM developed and performed the flow cytometry studies, participate in coordination and data acquisition and interpretation; MC, MM and EP performed the histopathology diagnosis and immunohistochemistry evaluation, critically reviewed the manuscript; AV preformed and interpreted the magnetic resonance and participate in data acquisition and analysis; ST performed the stereotactic biopsies and participate in data acquisition; AP developed and performed the flow cytometry studies, participate in coordination; FM and AM provided clinical information and reviewed the manuscript; AnP participate in study coordination, data acquisition and analysis, helped in drafting the manuscript; LC participate in revising the manuscript; CMC performed the stereotactic biopsies, participated in the study design, coordination and data acquisition, critically revised the manuscript. All authors read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. ==== Refs References 1. Gerstner E Batchelor T Primary CNS lymphoma Expert Rev Anticancer Ther 2007 7 689 700 10.1586/14737140.7.5.689 17492932 2. Korfel A Schlegel U Diagnosis and treatment of primary CNS lymphoma Nat Rev Neurol 2013 9 317 327 10.1038/nrneurol.2013.83 23670107 3. Dupire S Coiffier B Targeted treatment and new agents in diffuse large B cell lymphoma Int J Hematol 2010 92 12 24 10.1007/s12185-010-0609-6 20559762 4. Jin Z Qing K Ouyang Y Liu Z Wang W Li X Low dose of lenalidomide and PI3K/mTOR inhibitor trigger synergistic cytoxicity in activated B cell-like subtype of diffuse large B cell lymphoma J Exp Clin Cancer Res 2016 35 52 10.1186/s13046-016-0327-x 27009084 5. Montesinos-Rongen M Brunn A Bentink S Basso K Lim WK Klapper W Gene expression profiling suggests primary central nervous system lymphomas to be derived from a late germinal center B cell Leukemia 2008 22 400 405 10.1038/sj.leu.2405019 17989719 6. Weller M Martus P Roth P Thiel E Korfel A German PCNSL Study Group Surgery for primary CNS lymphoma? Challenging a paradigm Neuro Oncol 2012 14 1481 1484 10.1093/neuonc/nos159 22984018 7. Krieger MD Chandrasoma PT Zee CS Apuzzo ML Role of stereotactic biopsy in the diagnosis and management of brain tumours Semin Surg Oncol 1998 14 13 25 10.1002/(SICI)1098-2388(199801/02)14:1<13::AID-SSU3>3.0.CO;2-5 9407627 8. Batchelor T Loeffler JS Primary CNS lymphoma J Clin Oncol 2006 24 1281 1288 10.1200/JCO.2005.04.8819 16525183 9. Schlegel U Primary CNS, lymphoma Therap Advances in Neurol Disord 2009 2 93 104 10.1177/1756285608101222 21180644 10. Yap KK Sutherland T Liew E Tartaglia CJ Pamg M Trost N Magnetic resonance features of primary central nervous system lymphoma in the immunocompetent patient: a pictorial essay J Med Imaging Radiat Oncol 2012 56 179 186 10.1111/j.1754-9485.2012.02345.x 22498191 11. Qang ZL Zhang CB Cai JQ Li QB Wang Z Jiang T Integrated analysis of genome-wide DNA methylation, gene expression and protein expression profiles in molecular subtypes of WHO II-IV gliomas J Exp Clin Cancer Res 2015 34 127 10.1186/s13046-015-0249-z 26502731 12. Craig FE Foon KA Flow cytometric immunophenotyping for hematologic neoplasms Blood 2008 111 3941 3967 10.1182/blood-2007-11-120535 18198345 13. Cordone I Marchesi F Masi S Summa V Pisani F Merola R Flow cytometry remission by Ig light chains ratio is a powerful marker of outcome in multiple myeloma after tandem autologous transplant: a real-life study J Exp Clin Cancer Res 2016 35 49 10.1186/s13046-016-0324-0 26992692 14. Zeppa P Vigliar E Cozzolino I Troncone G Picardi M De Renzo A Fine needle aspiration cytology and flow cytometry immunophenotyping of non-Hodgkin lymphoma: can we do better? Cytopathology 2010 21 300 310 10.1111/j.1365-2303.2009.00725.x 20132132 15. Bertram HC Check IJ Milano MA Immunophenotyping large B-cell lymphomas. Flow cytometric pitfalls and pathologic correlation Am J Clin Pathol 2001 116 191 203 10.1309/BA3U-RMTU-D7UJ-M8DR 11488065 16. Frederiksen JK Sharma M Casulo C Burack WR Systematic review of the effectiveness of fine-needle aspiration and/or core needle biopsy for subclassifying lymphoma Arch Pathol Lab Med 2015 139 245 251 10.5858/arpa.2013-0674-RA 25611108 17. Demurtas A Accinelli G Pacchioni D Godio L Novero D Bussolati G Utility of flow cytometry immunophenotyping in fine-needle aspirate cytologic diagnosis of non-Hodgkin lymphoma: A series of 252 cases and review of the literature Appl Immunohistochem Mol Morphol 2010 18 311 322 10.1097/PAI.0b013e3181827da8 19247181 18. Vallangeon BD, Tyer C, Williams B, Lagoo AS. Improved detection of diffuse large B-cell lymphoma by flow cytometric immunophenotyping-Effect of tissue disaggregation method. Cytometry B Clin Cytom. 2015. [Epub ahead of print]. 19. Bromberg JE Breems DA Kraan J Bikker G van der Holt B Smitt PS CSF flow cytometry greatly improves diagnostic accuracy in CNS hematologic malignancies Neurology 2007 15 68 1674 1679 10.1212/01.wnl.0000261909.28915.83 17502548 20. Alvarez R Dupuis J Plonquet A Christov C Copie-Bergman C Hemery F Clinical relevance of flow cytometric immunophenotyping of the cerebrospinal fluid in patients with diffuse large B-cell lymphoma Ann Oncol 2012 23 1274 1279 10.1093/annonc/mdr436 21965472 21. Marchesi F Masi S Summa V Gumenyuk S Merola R Orlandi G Flow cytometry characterization in central nervous system and pleural effusion multiple myeloma infiltration: an Italian National Cancer Institute experience Br J Haematol 2016 172 980 982 10.1111/bjh.13549 26081466 22. Schroers R Baraniskin A Heute C Vorgerd M Brunn A Kuhnhenn J Diagnosis of leptomeningeal disease in diffuse large B-cell lymphomas of the central nervous system by flow cytometry and cytopathology Eur J Haematol 2010 85 520 528 10.1111/j.1600-0609.2010.01516.x 20727005 23. Birnbaum T Stadler EA von Baumgarten L Straube A Rituximab significantly improves complete response rate in patients with primary CNS lymphoma J Neurooncol 2012 109 285 291 10.1007/s11060-012-0891-7 22570142 24. Cordone I Masi S Mauro FR Soddu S Morsilli O Valentini T p53 expression in B-cell chronic lymphocytic leukemia: a marker of disease progression and poor prognosis Blood 1998 91 4342 4349 9596683 25. Swerdlow SH Campo E Harris NL Jaffe ES Pileri SA Stein H WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues 2008 Lyon IARC 26. Önder E Arıkök AT Önder S Han Ü Sorar M Kertmen H Corticosteroid pre-treated primary CNS lymphoma: a detailed analysis of stereotactic biopsy findings and consideration of interobserver variability Int J Clin Exp Pathol 2015 8 7798 7808 26339344 27. Canovi S, Campioli D. Accuracy of flow cytometry and cytomorphology for the diagnosis of meningeal involvement in lymphoid neoplasms: A systematic review. Diagn Cytopathol. 2016. doi: 10.1002/dc.23539. 28. Galati D Di Noto R Del Vecchio L Diagnostic strategies to investigate cerebrospinal fluid involvement in haematological malignancies Leuk Res 2013 37 231 237 10.1016/j.leukres.2012.11.016 23287431 29. Camilleri-Broët S Crinière E Broët P Delwail V Mokhtari K Moreau A A uniform activated B-cell-like immunophenotype might explain the poor prognosis of primary central nervous system lymphomas: analysis of 83 cases Blood 2006 107 190 196 10.1182/blood-2005-03-1024 16150948 30. Schniederjan SD Li S Saxe DF Lechowicz MJ Lee KL Terry PD A novel flow cytometric antibody panel for distinguishing Burkitt lymphoma from CD10+ diffuse large B-cell lymphoma Am J Pathol 2010 133 718 726 31. Ponzoni M Berger F Chassagne-Clement C Tinguely M Jouvet A Ferreri AJ International Extranodal Lymphoma Study Group Reactive perivascular T-cell infiltrate predicts survival in primary central nervous system B-cell lymphomas Br J Haematol 2007 138 316 323 10.1111/j.1365-2141.2007.06661.x 17555470 32. He M Zuo C Wang J Liu J Jiao B Zheng J Prognostic significance of the aggregative perivascular growth pattern of tumor cells in primary central nervous system diffuse large B-cell lymphoma Neuro Oncol 2013 15 727 734 10.1093/neuonc/not012 23482670 33. Wang X, Popplewell LL, Wagner JR, Naranjo A, Blanchard MS, Mott MR, et al. Phase I studies of central-memory-derived CD19 CAR T cell therapy following autologous HSCT in patients with B-cell NHL. Blood. 2016. [Epub ahead of print]. 34. Kitai R Ishisaka K Sato K Sakuma T Yamauchi T Imamura Y Primary central nervous system lymphoma secretes monocyte chemoattractant protein 1 Med Mol Morphol 2007 40 18 22 10.1007/s00795-006-0350-7 17384985 35. Olejniczak SH Stewart CC Donohue K Czuczman MS A quantitative exploration of surface antigen expression in common B-cell malignancies using flow cytometry Immunol Invest 2006 35 93 114 10.1080/08820130500496878 16531332 36. Kusenda J Quantitative identification of blood cell markers in human hematopoietic malignancies with diagnostic and prognostic significance Neoplasma 2008 55 381 386 18665747 37. Fang C Zhuang Y Wang L Fan L Wu YJ Zhang R High levels of CD20 expression predict good prognosis in chronic lymphocytic leukemia Cancer Sci 2013 104 996 1001 10.1111/cas.12192 23659384 38. Johnson NA Boyle M Bashashati A Leach S Brooks-Wilson A Sehn LH Diffuse large B-cell lymphoma: reduced CD20 expression is associated with an inferior survival Blood 2009 113 3773 3780 10.1182/blood-2008-09-177469 19029441 39. Morris PG Correa DD Yahalom J Raizer JJ Schiff D Grant B Rituximab, methotrexate, procarbazine, and vincristine followed by consolidation reduced-dose whole-brain radiotherapy and cytarabine in newly diagnosed primary CNS lymphoma: final results and long-term outcome J Clin Oncol 2013 31 3971 3979 10.1200/JCO.2013.50.4910 24101038 40. Mason DY Comans-Bitter WM Cordell JL Verhoeven MA van Dongen JJ Antibody L26 recognizes an intracellular epitope on the B-cell-associated CD20 antigen Am J Pathol 1990 136 1215 1222 1694056 41. Bryson GJ Lear D Williamson R Wong RC Detection of the CD56+/CD45- immunophenotype by flow cytometry in neuroendocrine malignancies J Clin Pathol 2002 55 535 537 10.1136/jcp.55.7.535 12101203

PMC005xxxxxx/PMC5002321.txt

==== Front Reprod HealthReprod HealthReproductive Health1742-4755BioMed Central London 20610.1186/s12978-016-0206-0ResearchMaternal and fetal outcomes in pregnancies complicated by overweight and obesity Vernini Joice Monaliza 1Moreli Jusciele Brogin 1Magalhães Claudia Garcia 2Costa Roberto Antônio Araújo 2Rudge Marilza Vieira Cunha 12Calderon Iracema Mattos Paranhos +55 14 38801383calderon@fmb.unesp.br 121 Graduate Program in Gynecology, Obstetrics and Mastology, Botucatu Medical School, São Paulo State University - Unesp, São Paulo, Brazil 2 Department of Obstetrics and Gynecology, Botucatu Medical School, São Paulo State University – Unesp, São Paulo, Brazil 27 8 2016 27 8 2016 2016 13 1 10027 5 2015 29 7 2016 © Vernini et al. 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Overweight and obesity are associated with pregnancy complications and adverse perinatal outcomes, posing short and long-term risks for maternal and child health. This study evaluated maternal, delivery and neonatal outcomes in pregnancies complicated by overweight and obesity. Methods This prospective cross-sectional study included 258 pregnant women. According to prepregnancy body mass index (BMI), participants were classified as normal weight, overweight, or obese. Data were analyzed using the chi-square test and analysis of variance followed by the Tukey test. Logistic regression was performed to calculate odds ratios and 95 % confidence intervals (p < 0.05). Results Most women ≥ 35 years old were overweight (22.7 %) and obese (27.6 %). Prepregnancy diabetes was significantly associated with obesity (15.7 %, p < 0.000). Obese women showed the lowest weight gain (9.6 ± 7.5Kg). Overweight and obese women practiced physical exercise more frequently (p = 0.010) than normal weight women. A greater proportion of obese mothers (13.4 %) had large for gestational age babies (p = 0.021), with higher thoracic circumference (33.6 ± 2.0 cm) and abdominal circumference (31.6 ± 2.3 cm). Obesity increased the risk of developing hypertension (OR = 7.0; 3.1-15.9), hyperglycemic disturbances (OR = 5.5; 2.9-10.6) and HbA1c ≥ 6.5 % (OR = 3.7; 1.2-11.1). The infants born to obese mothers had longer hospital stay (3.9 ± 3.9 days) (p = 0.005). Conclusion Our results confirm that obesity in pregnancy can lead to adverse outcomes, and underscore the importance of identifying and treating inadequate weight status during pregnancy. Keywords ObesityOverweightPregnancyMaternal outcomesPerinatal outcomesissue-copyright-statement© The Author(s) 2016 ==== Body Background Overweight and obesity are defined as abnormal or excess accumulation of adipose tissue in the body. These conditions are caused by a combination of genetic, metabolic, behavioral, environmental, cultural, and socioeconomic factors. Overweight and obesity are public health problems that contribute to preventable deaths each year, and the number of overweight and obese women at reproductive age is increasing in many countries [1, 2]. Since pregnancy can serve as a triggering or aggravating factor for obesity [3, 4], diagnosing and monitoring the weight status of pregnant women should be a routine prenatal care procedure [5, 6]. A number of factors, such as water retention, uterine growth, formation of fetal tissues and placenta, and increasing amniotic fluid volume, can limit the evaluation of maternal body mass index (BMI) during pregnancy [7, 8]. Therefore, since 2004, the Brazilian Ministry of Health has issued guidelines for categorizing ideal gestational weight gain based on prepregnancy BMI. These guidelines, recommend assessing a woman’s nutritional status during pregnancy using a specific tool [6] which was developed based on Atalah’s curve [9] and IOM recommendations [8]. Overweight pregnant women have higher risks of hypertension, gestational diabetes mellitus (GDM), Cesarean section, shoulder dystocia and early neonatal death [10–14]. Obesity in pregnancy compromises fetal metabolic programming, increasing the risk of obesity, diabetes and cardiovascular disorders in the offspring [15, 16]. Thus, the present study aimed at evaluating maternal, delivery and neonatal outcomes in pregnancies complicated by overweight and obesity. Methods Study design and subjects This is a prospective cross-sectional study including 258 pregnant women and their infants. Based on prepregnancy weight as self-reported during the first prenatal visit, pregnant women were classified according to BMI as normal weight (18.5–24.9 Kg/m2 BMI; n = 65); overweight (25.0 - 29.9 Kg/m2 BMI; n = 66); or obese (BMI ≥ 30.0 Kg/m2; n = 127) [17]. Ethics approval was granted by the Human Research Ethics Committee of Botucatu Medical School-Unesp, and written informed consent was obtained from all study participants. Inclusion and exclusion criteria Inclusion criteria were prenatal care commencing up to 20 weeks of gestationand singleton pregnancy. Underweight pregnant women were excluded. Data collection During the first prenatal visit (before 20th gestational week), the following maternal characteristics were evaluated: age (full years); number of gestations including current pregnancy; previous Cesarean section; smoking habits, history of hypertension and diabetes mellitus,and family history of diabetes, obesity, hypertension, cardiovascular disease and hypercholerestolemia. At the end of gestation (37th to 38th gestational week), maternal pregnancy outcomes were evaluated based on final pregnancy weight; gestational weight gain; gestational BMI; practice of light to moderate-intensity exercises during pregnancy (yes or no); and intercurrent diseases (such as urinary infection, genital infection, gestational hypertension, preelampsia, hyperglycemic disorders (mild gestational hyperglycemia – MGH, and gestational diabetes mellitus – GDM) [18, 19], and glycated hemoglobin (HbA1c) levels in the third trimester of pregnancy. Data collected at birth included mode of delivery (vaginal or C-section); gestational age; New Ballard scores; birth weight; birth weight/gestational age category; ponderal index [(birth weight/lenght3) × 100]; length; circumference of the head, thorax, and abdomenl; 1, 5 and 10 min Apgar scores; placenta weight; and placental index (PI = placenta weight/birth weight). In the cord blood, the levels of glucose, hematocrit, hemoglobin, bilirubin, and white and red blood cells were evaluated.During the neonatal period, information on hypoglycemia episodes; need for phototherapy; malformations and infant hospital stay length was obtained.All data were recorded on forms specifically developed for the study. Follow-up All women received follow up care at the Diabetes and Pregnancy Referral Center of Botucatu Medical School-Unesp. Reasons for referral were previous diabetes mellitus (type 1-DM or type 2-DM) or risk of developing hyperglycemic disorders [20], i.e. gestational diabetes mellitus (GDM) and mild gestational hyperglycemia (MGH) [17, 19]. According to our center’s routine protocol [18, 19], diabetic pregnant women (type 1-DM or type 2-DM) were immediately managed with glycemic control, individualized nutritional intervention, and a light to moderate-intensity exercise program (most frequently walking for 30 minutes five times a week). Insulin therapy was introduced when necessary [21]. Pregnant women who were nondiabetic, but were overweight or obese, were advised as to the importance of lifestyle changes to prevent GDM and MGH, and were promptly assigned to individualized nutritional guidance and home walking for 30 minutes five times a week for weight control during pregnancy. Regardless of these preventive measures, all nondiabetic pregnant women underwent glucose tolerance (75 g-OGTT) and gycemic profile (GP) testing between 24 and 28 weeks of pregnancy for confirmation or ruling out of GDM and MGH [18–20]. Pregnant women with confirmed GDM or MGH were treated according to the same protocol to achieve glycemic control. Insulin therapy was introduced when necessary. Glycemic control and management of diabetes were evaluated by 24-h GP (fasting, pre- and post- prandial glycemic levels) performed at 2-week intervals until week 32, and weekly until delivery [18–20]. Statistical analyses Data were systematically transferred from medical records to Microsoft Excel 2010. Statistical analyses were performed using SAS 9.2 for Windows. Data frequencies (in percentage) were compared using the chi-square test; means were evaluated by analysis of variance (ANOVA), followed by the Tukey-Kramer test for multiple comparisons. Data on maternal, neonatal and delivery outcomes were assessed using logistic regression to calculate odds ratios with confidence intervals (CI 95 %). Significance level was set at p < 0.05. Results Overweight and obesity predominated among pregnant women ≥ 35 years, whereas normal weight was most frequently observed among pregnant adolescents (≤19 years old) (p < 0.001). Prepregnancy diabetes was more frequent in obese (15.7 %) and overweight (13.6 %) women than in women with normal weight (6.2 %). The prevalence of of family history of obesity was greater in obese and overweight women, although not statistically significant (p = 0.051) (Table 1).Table 1 Characteristics of the participating pregnant women Normal weight N = 65 Overweight N = 66 Obese N = 127 p-value* N (%) N (%) N (%) Age (years)* <0 .00 1  ≤19 12 (18.5) 2 (3.0) a 1 (0.8) a  20 – 35 46 (70.8) 49 (74.2) 91(71.6)  >35 7 (10.8) 15 (22.7) a 35 (27.6) a Pregnancies (including current pregnancy) 0.187  1 18 (27.7) 11 (16.7) 19 (15.0)  2 19 (29.2) 18 (27.3) 33 (26.0)  ≥3 28 (43.1) 37 (56.1) 75 (59.6) Previous C-section 0.366  Zero 44 (67.7) 35 (53.0) 68 (53.5)  1 14 (21.5) 20 (30.3) 36 (28.4)  ≥2 7(10.8) 11 (16.7) 23 (18.1) Smoking habits 9 (13.8) 17 (25.8) 20 (15.8) 0.141 Cronic hypertension 4 (50.0) 17 (81.0) 47 (74.6) 0.231 Prepregnancy diabetes 4 (6.2) 9 (13.6) a 20 (15.7) a <0.001 Family history  Diabetes 54 (83.1) 48 (72.7) 93 (73.2) 0.266  Obesity 17 (26.2) 24 (36.4) 56 (44.1) a,b 0.051  Hypertension 44 (67.7) 49 (74.2) 101 (79.5) 0.195  Cardiovascular disease 14 (21.5) 19 (28.8) 43 (33.9) 0.206  Hypercholesterolemia 15 (23.1) 20 (30.3) 41 (32.3) 0.410 * Chi-square test a statistically different from the control group (p < 0.05) b statistically different from the overweight group (p < 0.05) Compared to women with normal weight, overweight and obese women showed higher mean baseline weight and final weight, prepregnancy BMI and gestational BMI (at 37–38 weeks of pregnancy) (p < 0.001). Nonetheless, weight gain (WG) was lowest in the obese group (Table 2).Table 2 Maternal outcomes Normal weight N = 65 Overweight N = 66 Obese N = 127 *p-value Mean ± sd Mean ± sd Mean ± sd Initial weight (Kg) 57.1 ± 5.7 69.9 ± 6.7 a 91.2 ± 13.7 a,b <0.001 Final weight (Kg) 71.5 ± 7.1 82.9 ± 7.6 a 100.7 ± 15.4 a,b <0.001 Weight gain (Kg) 14.4 ± 4.6 12.8 ± 5.7 9.6 ± 7.5 a,b <0.001 Prepregnancy BMI (Kg/m2) 22.4 ± 1.7 27.5 ± 1.3 a 36.2 ± 4.9 a,b <0.001 Gestational BMI (Kg/m2) 28.1 ± 2.4 32.6 ± 2.7 a 39.9 ± 5.5 a,b <0.001 N (%) N (%) N (%) †p-value Weight gain (Kg) <0.001  <8 5 (7.8) 12 (18.2) 58 (45.7) a,b  8 – 16 36 (55.4) 36 (54.6) 42 (33.1) a,b  >16 24 (36.9) 18 (27.3) 27 (21.3) Gestational BMI (Kg/m2) <0.001  Underweight 7 (10.8) 0 (0.0) a 0 (0.0) a  Adequate 33 (50.8) 5 (7.6) a 0 (0.0) a,b  Overweight 23 (35.4) 35 (53.0) a 8 (6.3) a,b  Obese 2 (3.1) 26 (39.4) a 119 (93.7) a,b Physical activity (yes) 17 (26.2) 34 (51.5) a 55 (43.3) a 0.010 Intercurrent diseases 30 (46.2) 26 (39.4) 62 (48.8) 0.458  Urinary infection 8 (12.5) 8 (11.9) 30 (23.6)  Genital infection 12 (18.8) 12 (17.9) 15 (11.8)  GH or PE 4 (50.0) 4 (19.0) 16 (25.4)  GDM 8 (12.3) 22 (33.3) a 50 (39.4) a <0.001  MGH 12 (18.4) 15 (22.7) a 27 (21.2) a 0.005 HbA1c (%) [3rd trimester] 0.075  <6.5 61 (93.8) 53 (80.3) 102 (80.3)  6.5 – 8.0 3 (4.6) 13 (19.7) 23 (18.1)  ≥8.0 1 (1.6) 0 (0.00) 2 (1.57) * ANOVA; Tukey test † Chi-square test a statistically different from the control group (p < 0.05) b statistically different from the overweight group (p < 0.05) WG < 8 Kg was more frequent among the obese, while excess WG (WG > 16 kg) was more frequently seen in women with normal weight (p < 0.001). By the end of pregnancy, most normal weight women were either overweight (35.4 %) or obese (3.1 %) (p < 0.001). Most overweight and obese women engaged in physical activity during pregnancy (p = 0.010), and developed either GDM (p < 0.001) or MGH (p = 0.005). Irrespective of group, over 80.0 % of the pregnant women with hyperglycemia exhibited normal HbA1c levels (<6.5 %) in the third trimester of pregnancy, with no differences among groups (p = 0.075) (Table 2). Obese women had the highest rate of LGA (13.4 %) and the lowest rate of SGA infants (4.7 %) (p = 0.021); with 40.9 % of these infants showing disproportionate growth (IP ≥ 2.98) (p = 0.037). Mean thoracic circumference (p = 0.006), abdominal circumference (p = 0.032), placental weight (p < 0.001) and placental index (p = 0.037) were higher in the obese group (Table 3). The infants born to obese mothers remained in hospital for a longer period (p = 0.005), but maternal weight status did not affect neonatal outcomes (Table 4).Table 3 Delivery and neonatal outcomes Normal weight Overweight Obese *p-value N = 65 N = 66 N = 127 N (%) N (%) N (%) Mode of delivery 0.304  Vaginal 25 (39.5) 21 (31.8) 35 (27.6)  Cesarean 40 (61.5) 45 (68.2) 92 (72.4) Gestational age (weeks) 0.494  <37 3 (4.6) 5 (7.6) 12 (9.4)  ≥37 62 (95.4) 61 (92.4) 115 (90.6) New Ballard (weeks)± 0.115  <37 3 (5.8) 6 (10.9) 19 (17.1)  ≥37 49 (94.2) 49 (89.1) 92 (82.9) Birth weight (g) 0.304  <2500 4 (6.2) 2 (3,0) 2 (1.6)  2500 − 4000 57 (87.7) 61 (92.4) 113 (89.0)  ≥4000 4 (6.2) 3 (4.6) 12 (9.4) Weight class 0.021  SGA 8 (12.3) 10 (15,2) 6 (4.7) a,b  AGA 54 (83.1) 53 (80.3) 104 (81.9)  LGA 3 (4.6) 3 (4.6) 17 (13.4) Ponderal index (PI) 0.037  <2.98 (proportional) 50 (76.9) 46 (69,7) 75 (59.1)  ≥2.98 (disproportionate) 15 (23.1) 20 (30.3) 52 (40.9) a,b Mean ± sd Mean ± sd Mean ± sd #p-value Weight (g) 3210.2 ± 469.2 3208.9 ± 430.4 3365.9 ± 483.4 0.199 Length (cm) 48.6 ± 2.2 48.4 ± 2,0 48.6 ± 2.1 0.759 Head Circumf (cm) 34.4 ± 1.5 34.4 ± 1,3 34.8 ± 1.5 0.094 Thoracic Circumf (cm) 32.9 ± 1.9 32.8 ± 1,7 33.6 ± 2.0 a,b 0.006 Abdom Circumf (cm) 31.1 ± 2.3 30.7 ± 1,8 31.6 ± 2.3 a,b 0.032 1-min Apgar 8.3 ± 1.4 8.2 ± 1,8 7.8 ± 1.5 0.076 5-min Apgar 9.3 ± 0.8 9.3 ± 0.9 9.1 ± 0.8 0.143 10-min Apgar 9.7 ± 0.5 9.6 ± 0.6 9.5 ± 0.6 0.071 Ponderal index 2.8 ± 0.3 2.8 ± 0.3 2.9 ± 0.3 a,b 0.005 Placental weight (g) 560.1 ± 138,9 574,6 ± 119,8 635,6 ± 156,6 a,b 0.001 Placental index 0.2 ± 0.0 0.2 ± 0.0 0.2 ± 0.0 a,b 0.037 ± N = 217 [41 newborns with incomplete data] * Chi-square test # ANOVA; Tukey test a statistically different from the control group p < 0.05) b statistically different from the overweight group (p < 0.05) Table 4 Neonatal outcomes± Normal weight N = 65 Overweight N = 66 Obese N = 127 *p-value N (%) N (%) N (%) Hypoglycemia 0 0 1 (1.0) 0.620 Malformation 2 (4.4) 2 (4.0) 6 q 0.833 Phototherapy 14 (23.0) 10 (16.4) 37 (37.0) 0.091 Mean ± SD Mean ± SD Mean ± SD #p-value Hematocrit (%) 48.3 ± 5.3 51.0 ± 9.3 49.6 ± 5.4 0.163 Hemoglobin (g/dL) 16.2 ± 1.8 16.4 ± 1.9 16.3 ± 2.4 0.832 Ind. Bilirubin (mg/dL) 1.9 ± 0.5 2.0 ± 0.7 2.0 ± 0.6 0.575 White cells (x103/mm3) 14755 ± 3095.8 14786 ± 6627.7 14543 ± 4332.8 0.975 Red cells (/mm3) 4.8 ± 0.6 4.4 ± 0. 5 4.6 ± 0.5 0.082 Glucose levels (mg/dL) 63.4 ± 16.3 65.8 ± 25.5 66.0 ± 20.5 0.091 Hospitalization (days) 3.0 ± 1.5 3.0 ± 1.4 3.9 ± 3.9 a,b 0.005 ± N = 204 [54 infants with incomplete records] * Chi-square test # ANOVA; Tukey test a statistically different from the control group (p < 0.05) b statistically different from the overweight group (p < 0.05) Compared to normal weight women, obese and overweight mothers had lower weight gain (WG), with odds ratio (OR) of 0.2 (0.1–0.4) and 0.4 (0.2–0.7), respectively. This indicates that overweight and obesity were protective against excess maternal weight gain. Nevertheless, obese women showed higher risk of developing hypertension (OR = 7.0; 3.1–15.9), hyperglycemic disturbances (OR = 5.5; 2.9–10.6) and HbA1c > 6.5 % (OR = 3.7; 1.2–11.1) during pregnancy. Maternal weight status did not affect neonatal outcomes (Table 5).Table 5 Odds ratio (OR) and confidence interval (CI 95 %) of prepregnancy BMI vs maternal and neonatal outcomes Obese vs normal weight Overweight vs normal weight OR CI 95 % OR CI 95 % Mother  Weight gain ≥ 16 Kg 0.2 0.1 – 0.4 0.4 0.2 − 0.7  Hypertension 7.0 3.1 − 15.9 2.1 1.1 − 3.9  Hyperglycemia 5.5 2.9 – 10.6 1.3 0.7 – 2.6  HbA1c ≥ 6.5 % 3.7 1.2 – 11.1 1.0 0.5 – 2.2 Newborn  Weight 2.3 0.8 – 6.6 1.9 0.7 – 5.4  Weight classes 1.0 0.4 – 2.1 0.8 0.4 – 1.7  1-min Apgar score 0.4 0.1 – 1.1 0.6 0.2 – 1.5  5-min Apgar score 2.0 0.1 – 32.0 1.9 0.1 – 31.5 Discussion In the present study, overweight and obesity were associated with increased maternal weight, and gestational BMI, as well as higher prevalence of GDM and MGH. In the obese group, the proportion of LGA babies with disproportionate ponderal index (≥2.98), placenta weight and placental index (PI) were higher, and hospitalstay length was longer. Obesity was associated with higher risk of developing hypertension, hyperglycemia and HbA1c ≥ 6.5 % in late pregnancy. Unexpectedly, obesity and overweight were protective factors against maternal weight gain, without affecting neonatal outcomes. Obese and overweight women were heavier and showed higher prepregnancy BMI, which was proportional to their nutritional status. These features persisted throughout pregnancy and were associated with hypertension, GDM and MGH. Similar results have also been observed in previous studies showing association of increased BMI with development of hypertension and hyperglycemia [3, 7, 22–24]. In this study, logistic regression analysis indicated obesity as a risk factor for gestational hypertension and hyperglycemia, increasing the likelihood of their occurrence seven and five-fold, respectively. This maternal characteristic was also a determining factor for the four-fold increase in the occurrence of HbA1c ≥ 6.5 % at the end of pregnancy. Hypertension and hyperglycemia associated with central obesity (in women, BMI ≥ 30 Kg/m2 or waist circumference ≥ 80 cm) are clinical criteria for the diagnosis of metabolic syndrome (MS), whose physiopathological basis is the association between obesity and insulin resistance [25]. Insulin resistance, classically considered as a characteristic of healthy pregnancy, is more pronounced in pregnancies complicated by diabetes [26]. The relationship between maternal MS and hyperglycemia during pregnancy was first reported by Bo et al. [27] and has been recently corroborated by our research group [28, 29]. Thus, in addition to the 13–15 % rates of pregestational DM observed in the present study, the overweight and obese groups were expected to show elevated rates of MGH or GDM, and a high proportion of HbA1c ≥ 6.5 %. However, although elevated rates of MGH (about 22 %) and GDM (approximately 30 to 40 %) were indeed observed, only 20.0 % of the overweight and obese participants showed HbA1c ≥ 6.5 %. All overweight and obese women participating in this study received individualized dietary advice and were encouraged to engage in physical activity regardless of diagnosis of diabetes or hyperglycemia. Indeed, 51.5 and 43.3 % of overweight and obese pregnant women, respectively, adhered to physical activity recommendations, but these measures did not prevent the occurrence of MGH or GDM, probably because the period between admission and diagnostic testing for MGH/GDM was short (about four weeks). Nonetheless, there was a 60–70 % reduction in the risk of excessive weight gain, which may have contributed to the lower weight gain observed in the obese group, as well as to the significant rates of overweight and obese women (35 to 40 %) moving to lower BMI classes in late pregnancy, and the achievement of maternal hyperglycemia control (HbA1c levels < 6.5 %) in 80 % of overweight and obese women. The relationship of diet/exercise with maternal weight remains controversial. Several studies have evaluated the effects of dietary advice alone or associated with exercise, on maternal weight gain, but no consistent results have been obtained [30–33]. A recent meta-analysis has suggested that compared to unsupervised physical activity and diet interventions, supervised exercise plus diet programs were most effective in managing weight among overweight or obese pregnant and postpartum women [34]. In this study, LGA infants, with PI ≥ 2.98 and higher thoracic and abdominal circumference measures, as well as increased placental weight and placental índex were more frequent in the obese group. According to previous studies, gestational BMI and DM or hyperglycemia are independent risk factors for excessive fetal growth [7, 12, 35, 36]. Moreover, macrosomia, disproportionate growth, placentomegaly, high placental index, and increased fetal measures (especially abdominal circumference) have been considered common features in babies born to diabetic mothers [19, 20, 35]. In Brazil, a recent review has shown that a large proportion of women are overweight or obese at the beginning of pregnancy, and that this is associated with higher risk of excessive weight gain, cesarean delivery, and fetal macrosomia [37]. In our study, overweight and obesity were identified as preventive factors against excessive weight gain without risk for the baby. Thus, the adverse neonatal outcomes here observed were probably caused by maternal hyperglycemia. Previous results obtained by our research group show that fetal macrosomia is significantly associated with prepregnancy BMI, hyperglycemic levels, and personal history of macrosomia in pregnancies complicated by diabetes or mild hyperglycemia [38]. A cohort study of 6,125 deliveries, using logistic multivariate analysis to exclude potential confounding variables, has demonstrated that only neonatal macrosomia and meconium aspiration syndrome remain significantly associated with maternal overweight and obesity [39]. Macrosomia and excessive fetal growth are frequently associated with BMI ≥ 25 kg/m2 and maternal diabetes or hyperglycemia, which are also interdependent. Therefore, it is practically impossible to determine the intensity of such effects. Although indirectly, our results corroborate the usefulness of dietary advice plus exercise as a preventive intervention against excessive weight gain in pregnancies complicated by overweight or obesity. However, this study has some limitations that include the adoption of a cross-sectional rather than a randomized controlled design, the potential bias of confounding factors relative to missing personal information (pre-gestational weight, compliance to treatment protocol); and sample size, which might not have been large enough for the evaluation of neonatal outcomes and a more elaborate statistical analysis (adjustment for age, parity and maternal weight or multiple regression analysis). Conclusions In the present study, adverse pregnancy outcomes were associated with obesity and overweight, and indirectly related to maternal hyperglycemia. Obesity was a determinant risk factor for hypertension, hyperglycemia, and increased HbA1c levels at the end of pregnancy. Overweight elevated the risk of hypertension, and both obesity and overweight were protective factors against excessive maternal weight gain without affecting neonatal outcomes. These findings indicate that adverse outcomes resulted from inadequate nutritional status during pregnancy, and indirectly underscore the importance of identifying and treating inadequate weight status during pregnancy. Acknowledgements The authors are thankful to the São Paulo Research Foundation (Fundação de Amparo à Pesquisa do Estado de São Paulo—FAPESP, grant # 2012/51257-9/PP-SUS 2012), and the Research Support Group (Escritório de Apoio à Pesquisa of Botucatu Medical School/UNESP) for their technical support in statistical analysis. Authors’ contributions IMPC substantially contributed to study conception and design, as well as data analysis and interpretation. JMV was responsible for data collection. JBM, CGM and RAAC supervised data collection, and contributed to data analysis and interpretation.All authors have read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. ==== Refs References 1. World Health Organization Obesity: preventing and managing the global epidemic. Report of a WHO consultation 2000 Geneva World Health Organization technical report series 2. World Health Organization Global strategy on diet, physical activity and health: obesity and overweight 2003 Geneva World Health Organization 3. World Health Organization (WHO). Obesity and overweight. Fact sheet No.311. 2006. [www.who.int/mediacentre/factsheets/fs311/en/index.html]. Last accessed 21 April 2016. 4. Mattar R Torloni MR Betrán AP Merialdi M Obesity and pregnancy Rev Bras Ginecol Obstet 2009 31 107 110 10.1590/S0100-72032009000300001 19547884 5. World Health Organization. WHO Expert Committee on Physical status: the Use and Interpretation of Anthropometry. WHO Technical report series: 854. Geneva: World Health Organization; 1995. 6. Ministério da Saúde (BR). Secretaria de Atenção à Saúde. Departamento de Ações Programáticas Estratégicas. Área Técnica de Saúde da Mulher. Pré-natal e puerpério: atenção qualificada e humanizada: manual técnico. 3a ed. Brasília DF: Ministério da Saúde; 2006. 7. Nomura RMY Paiva LV Costa VN Liao AW Zugaib M Influência do estado nutricional materno, ganho de peso e consumo energético sobre o crescimento fetal, em gestações de alto risco Rev Bras Ginecol Obstet 2012 34 107 112 22488493 8. Institute of Medicine (IOM) Subcommitee on nutritional status and weight gain during pregnancy. Institute of Medicine. Nutrition during pregnancy 2009 Washington, DC Nat Acad Press 9. Atalah SE Castilho LC Castro SR Aldea PA Propuesta de um nuevo estándar de evalución nutricional en embarazadas Rev Med Chil 1997 125 1429 1436 9609018 10. Assunção PL Melo ASO Gondim SSR Benício MHA Amorim MMR Cardoso MAA Ganho ponderal e desfechos gestacionais em mulheres atendidas pelo Programa de Saúde da Família em Campina Grande, PB (Brasil) Rev Bras Epidemiol 2007 10 352 360 10.1590/S1415-790X2007000300006 11. Torloni MR Betrán AP Horta BL Nakamura MU Atallah AN Moron AF Prepregnancy BMI and the risk of gestational diabetes: a systematic review of the literature with meta-analysis Obes Rev 2009 10 2 194 203 10.1111/j.1467-789X.2008.00541.x 19055539 12. Yu Z, Han S, Zhu J, Sun X, Ji C, Guo X. Pre-Pregnancy Body Mass Index in Relation to Infant Birth Weight and Offspring Overweight/Obesity: A Systematic Review and Meta-Analysis. PLoS ONE 8(4): e61627. doi:10.1371/journal.pone.0061627. 13. Guelinckx I Devlieger R Beckers K Vansant G Maternal obesity: pregnancy complications, gestational weight gain and nutrition Obes Rev 2007 9 14 50 14. Sebastian Manzanares G Angel Santalla H Irene Vico Z Lopez Criado MS Alicia Pineda L Abnormal maternal body mass index and obstetric and neonatal outcome J Matern Fetal Neonatal Med 2012 25 308 312 10.3109/14767058.2011.575905 21615231 15. Simmons D Diabetes and obesity in pregnancy Best Practice & Research Clinical Obstetrics & Gynaecol 2011 25 25 36 10.1016/j.bpobgyn.2010.10.006 16. Adamo KB Ferraro ZM Brett KE Pregnancy is a critical period for prevention of obesity and cardiometabolic Risk Can J Diabetes 2012 36 133 141 10.1016/j.jcjd.2012.05.004 17. World Health Organization Diet, nutrition and the prevention of chronic diseases Tech Rep Ser 2003 916 1 149 18. American Diabetes Association Gestational Diabetes Mellitus: position statements Diabetes Care 2012 5 S51 S71 19. Rudge MVC Calderon IMP Ramos MD Brasil MAM Rugolo LMSS Bossolan G Od Land JO Hiperglicemia materna diária diagnosticada pelo perfil glicêmico: um problema de saúde pública materno e perinatal Rev Bras Ginecol Obstet 2005 27 691 697 20. Rudge MV Calderon IM Ramos MD Abbade JF Rugolo LM Perinatal outcome of pregnancies complicated by diabetes and by maternal daily hyperglycemia not related to diabetes. A retrospective 10-year analysis Gynecol Obstet Invest 2000 50 108 112 10.1159/000010293 10965194 21. American College of Obstetricians and Gy- necologists ACOG Committee opinion. Number 267, January 2002: Exercise during pregnancy and the postpartum period Ob- stet Gynecol 2002 99 171 173 22. Abenhaim HA Kinch RA Morin J Benjamin A Usher R Effect of prepregnancy body mass index categories on obstetrical and neonatal utcomes Arch Gynecol Obstet 2007 275 39 43 10.1007/s00404-006-0219-y 16967276 23. Fortner RT Pekow P Solomon CG Markenson G Chasan-Taber L Prepregnancy body mass index, gestational weight gain, and risk of hypertensive pregnancy among Latina women Am J Obstet Gynecol 2009 200 e1 e7 10.1016/j.ajog.2008.08.021 19070831 24. Herring SJ Oken E Rifas-Shiman SL Rich-Edwards JW Stuebe AM Kleinman KP Gillman NW Weight gain in pregnancy and risk of maternal hyperglycemia Am J Obstet Gynecol 2009 201 e1 e7 10.1016/j.ajog.2009.01.039 25. International Diabetes Association (IDF) The IDF consensus worldwide definition of the metabolic syndrome 2007 Brussels IDF 26. Hollingsworth DR Maternal metabolism in normal pregnancy and pregnancy complicated by diabetes mellitus Clin Obstet Gynecol 1985 28 457 472 10.1097/00003081-198528030-00002 4053447 27. Bo S Menato G Gallo ML Bardelli C Lezo A Signorile A Gambino R Cassader M Massobrio M Pagano G Mild gestational hyperglycemia, the metabolic syndrome and adverse neonatal outcomes Acta Obstet Gynecol Scand 2004 83 335 340 10.1111/j.0001-6349.2004.00314.x 15005779 28. Negrato CA Javanovic L Tambascia MA Calderon IMP Geloneze B DIAS A Rudge MVC Mild Gestacional hyperglycemia as a risk factor for metabolic syndrome in pregnancy and adverse perinatal outcomes Diabetes Metab Res Rev 2008 24 324 330 10.1002/dmrr.815 18254163 29. Negrato CA Javanovic L Tambascia MA Calderon IMP Geloneze B Dias A Rudge MVC Association between insulin resistance, glucose intolerance, and hypertension in pregnancy Metab Syndr Relat Disord 2009 7 53 59 10.1089/met.2008.0043 18847384 30. Wolff S Legarth J Vangsgaard K Toubro S Astrup A A randomized trial of the effects of dietary counseling on gestational weight gain and glucose metabolism in obese pregnant women Int J Obes 2008 32 495 501 10.1038/sj.ijo.0803710 31. Thornton YS Smarkola C Kopacz SM Ishoof SB Perinatal outcomes in nutritionally monitored obese pregnant women: a randomized clinical trial J Natl Med Assoc 2009 01 569 577 10.1016/S0027-9684(15)30942-1 19585925 32. Guelinckx I Devlieger R Mullie P Vansant G Effect of lifestyle intervention on dietary habits, physical activity, and gestational weight gain in obese pregnant women: a randomized controlled trial Am J Clin Nutr 2010 91 373 380 10.3945/ajcn.2009.28166 19955397 33. Nascimento S Surita F Parpinelli M Siani S Pinto e Silva J The effect of an antenatal physical exercise programme on maternal/perinatal outcomes and quality of life in overweight and obese pregnant women: a randomised clinical trial BJOG 2011 118 1455 1463 10.1111/j.1471-0528.2011.03084.x 21895947 34. Choi J Fukuoka Y Lee JH The Effects of Physical Activity and Physical Activity plus Diet Interventions on Body Weight in Overweight or Obese Women who are Pregnant or in Postpartum: A Systematic Review and Meta Analysis of Randomized Controlled Trials Prev Med 2013 56 6 351 364 10.1016/j.ypmed.2013.02.021 23480971 35. Najafian M, Cheraghi M. Occurrence of Fetal Macrosomia Rate and Its Maternal and Neonatal Complications: A 5-Year Cohort Study. ISRN Obstetrics and Gynecology. 2012; Article ID 353791 (5 pages), doi:10.5402/2012/353791. 36. HAPO Study Cooperative Research Group Hyperglycaemia and adverse pregnancy outcome (HAPO) study: associations with maternal body mass index Int J Obstet Gynecol 2010 117 575 584 37. Godoy AC Nascimento SL Surita FG A systematic review and meta-analysis of gestational weight gain recommendations and related outcomes in Brazil Clinics 2015 70 11 758 764 10.6061/clinics/2015(11)08 26602524 38. Kerche LT Abbade JF Costa RA Rudge MV Calderon IM Fetal macrosomia risk factors in pregnancies complicated by diabetes or daily hyperglycemia Rev Bras Ginecol Obstet 2005 27 10 580 587 10.1590/S0100-72032005001000003 39. Narchi H Skinner A Overweight and obesity in pregnancy do not adversely affect neonatal outcomes: new evidence J Obstet Gynaecol 2010 30 7 679 686 10.3109/01443615.2010.509824 20925609

PMC005xxxxxx/PMC5002322.txt

==== Front BMC Infect DisBMC Infect. DisBMC Infectious Diseases1471-2334BioMed Central London 174910.1186/s12879-016-1749-yResearch ArticleFactors associated with optimal pharmacy refill adherence for antiretroviral medications and plasma HIV RNA non-detectability among HIV-positive crack cocaine users: a prospective cohort study Hayashi Kanna uhri-kh@cfenet.ubc.ca 12Wood Evan uhri-ew@cfenet.ubc.ca 12Kerr Thomas uhri-tk@cfenet.ubc.ca 12Dong Huiru hdong@cfenet.ubc.ca 1Nguyen Paul pnguyen@cfenet.ubc.ca 1Puskas Cathy M. cpuskas@cfenet.ubc.ca 13Guillemi Silvia sguillemi@cfenet.ubc.ca 1Montaner Julio S. G. jmontaner@cfenet.ubc.ca 12Milloy Michael-John +1 604 558 6676uhri-mjsm@cfenet.ubc.ca 121 British Columbia Centre for Excellence in HIV/AIDS, St. Paul’s Hospital, 608 - 1081 Burrard Street, Vancouver, BC V6Z 1Y6 Canada 2 Division of AIDS, Department of Medicine, University of British Columbia, 608 - 1081 Burrard Street, Vancouver, BC V6Z 1Y6 Canada 3 Faculty of Health Sciences, Simon Fraser University, 8888 University Drive, Burnaby, BC V5A 1S6 Canada 27 8 2016 27 8 2016 2016 16 1 45524 9 2015 3 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Crack cocaine use is known to contribute to poor adherence to antiretroviral medications; however, little is known about facilitators of or barriers to effective HIV treatment use among HIV-infected crack cocaine users. We sought to identify correlates of optimal pharmacy refill adherence for antiretroviral medications and plasma HIV RNA viral load (pVL) suppression among this population. Methods Data from a prospective cohort of HIV-positive people who use illicit drugs in Vancouver, Canada, were linked to comprehensive HIV clinical monitoring and pharmacy dispensation records. We used multivariable generalized linear mixed-effects modelling to longitudinally identify factors associated with ≥95 % adherence to pharmacy refills for antiretroviral medications and pVL <50 copies/mL among crack cocaine users exposed to highly-active antiretroviral therapy (HAART). Results Among 438 HAART-exposed crack cocaine users between 2005 and 2013, 240 (54.8 %) had ≥95 % pharmacy refill adherence in the previous 6 months at baseline. In multivariable analyses, homelessness (adjusted odds ratio [AOR]: 0.58), ≥daily crack cocaine smoking (AOR: 0.64), and ≥ daily heroin use (AOR: 0.43) were independently associated with optimal pharmacy refill adherence (all p < 0.05). The results for pVL non-detectability were consistent with those of medication adherence, except that longer history of HAART (AOR: 1.06), receiving a single tablet-per-day regimen (AOR: 3.02) and participation in opioid substitution therapies was independently associated with pVL non-detectability (AOR: 1.55) (all p < 0.05). Conclusions Homelessness, and daily crack cocaine and/or heroin use were independently and negatively associated with optimal HAART-related outcomes. With the exception of opioid substitution therapies, no addiction treatment modalities assessed appeared to facilitate medication adherence or viral suppression. Evidence-based treatment options for crack cocaine use that also confer benefits to HAART need to be developed. Keywords Crack cocaineART adherencePlasma HIV-1 RNA viral load suppressionAddiction treatmenthttp://dx.doi.org/10.13039/100000002National Institutes of HealthR01DA021525NIDA R01DA036307http://dx.doi.org/10.13039/501100000024Canadian Institutes of Health ResearchMSH-141971Hayashi Kanna Tier 1 Canada Research Chair in Inner City Medicineissue-copyright-statement© The Author(s) 2016 ==== Body Background Highly-active antiretroviral therapy (HAART) has substantially improved the life expectancy of people living with HIV [1] and has also been shown to prevent HIV transmission by reducing individuals’ viral load to undetectable levels [2, 3]. Despite the remarkable impact of HAART on the HIV/AIDS pandemic, its benefits on disease progression and the incidence of new infections have yet to be fully realized among all HIV-seropositive groups, such as people who use illicit drugs. Crack cocaine use has been shown to be an independent risk factor of HIV seroconversion in some settings, presumably through high-risk sexual activities associated with crack use [4–7]. While evidence-based strategies to reduce HIV transmission risk through injection drug use exist, such as needle and syringe programs [8], crack cocaine is often smoked, and interventions to reduce sexual risk behaviour among crack cocaine smokers have proven to be of limited effectiveness [5, 9, 10]. In 2014, the Treatment-as-Prevention (TasP) approach, which aims to expand access and adherence to HAART among people living with HIV and thereby prevent HIV transmission, was incorporated into the global HIV/AIDS response strategies by the Joint United Nations Programme on HIV/AIDS [11]. This is likely a potentially effective HIV prevention approach to address HIV transmission associated with persistent high-risk behaviour among HIV-infected crack cocaine users [12]. However, there remains concern regarding the viability of the intervention due to previous studies describing sub-optimal adherence to HAART in crack cocaine using populations [13–15]. For example, a 2002 study of HIV-infected people who use drugs reported that the strongest predictor of poor adherence to antiretroviral medications was cocaine/crack cocaine use [13]. A more recent study found that stimulant users, particularly those who used cocaine/crack cocaine, had seven times higher odds of HAART adherence failure (i.e., <90 % of doses taken) compared to patients who did not use drugs [15]. Despite many reports of poor adherence to HAART among HIV-infected crack cocaine users [13–15], to date, there is a paucity of research identifying factors associated with HAART adherence and plasma HIV RNA viral load (pVL) suppression in this population, which may be amenable to intervention. This is a concern as the success of TasP-based approaches relies on ensuring access and adherence to HAART. In this study, we sought to identify correlates of optimal adherence to antiretroviral medications and pVL non-detectability among HAART-exposed crack cocaine users in Vancouver, Canada, where crack cocaine use is common and has been identified as an independent predictor of HIV seroconversion [7]. Methods Study design Data for this analysis were derived from the AIDS Care Cohort to evaluate Exposure to Survival Services (ACCESS), a prospective cohort study of HIV-positive people who use illicit drugs in Vancouver, Canada. Recruitment and data collection procedures have been described in detail elsewhere [16]. In brief, beginning in 1996, participants were recruited through self-referral and street-based outreach from Vancouver’s Downtown Eastside neighbourhood, a post-industrial area with a large open drug market and high levels of illicit drug use, poverty, and HIV infection. Individuals were eligible to participate in ACCESS if they were aged 18 years or older, were HIV-seropositive, had used illicit drugs other than cannabis in the month prior to enrolment, and provided written informed consent. Participants were compensated $20 CAD at each study visit. At baseline and semi-annually thereafter, participants completed an interviewer-administered questionnaire soliciting demographic data, information on drug use patterns, as well as other characteristics and exposures. At each visit, individuals also underwent an examination by a study nurse and provided blood samples for serologic analyses. Through a confidential linkage with the provincial Drug Treatment Program (DTP), a complete clinical profile of all CD4 T-cell counts, pVL observations, exposure to specific antiretroviral agents, and treatment outcomes for each participant were obtained. In British Columbia, all provision of HIV treatment is centralized through a province-wide dispensation program, where treatment and related care are provided free of charge. Data collection by the DTP has been described previously [17, 18]. The ACCESS study has been approved by the University of British Columbia/Providence Health Care Research Ethics Board. Participants and measures Participants were eligible for the present analysis if they were recruited by ACCESS between December 1, 2005 and November 30, 2013, reported having smoked crack cocaine in the six months prior to the baseline assessment, and had initiated HAART prior to enrolment in ACCESS. As well, at least one observation of CD4 cell count and pVL had to be completed within ±180 days of the day they entered the study. For a sub-analysis focusing on dual crack cocaine and opioid users, the sample was further restricted to those who had a history of any opioid use or participation in methadone maintenance therapy at baseline. The primary outcomes of interest were optimal adherence to pharmacy refills for antiretroviral medications and pVL non-detectability in the previous six months. We measured pharmacy refill adherence in each 6-month period as the number of days for which medications were dispensed over the number of days since the initiation of HAART, which was capped at 180 days. We defined optimal pharmacy refill adherence as equal to or greater than 95 %. We have previously demonstrated the clinical validity of this pharmacy refill data and shown that it reliably predicts virologic suppression [19, 20] and survival [21]. Plasma HIV RNA viral load non-detectability was defined as having achieved a HIV RNA viral load of <50 copies/mL of plasma. In the event that more than one pVL observation was made within a 6-month follow-up period, we used the median of all the observations, which was then categorized into either <50 or ≥50 copies/mL of plasma. We considered a range of explanatory variables that might be associated with optimal pharmacy refill adherence and/or pVL non-detectability. Demographic characteristics included: age (per 10-year older), gender (male vs. female), ethnicity (Caucasian vs. other), and education level (<secondary education vs. ≥secondary education). Substance-using behaviours included: crack cocaine smoking (≥daily vs. <daily), heroin use (≥daily vs. <daily), crystal methamphetamine use (≥daily vs. <daily), and heavy alcohol use (≥5 drinks per day vs. <5 drinks per day). Other behavioural variables included: homelessness, engagement in sex work or drug dealing, incarceration, participation in inpatient addiction treatment (i.e., recovery houses, detoxification facilities, and treatment centres), participation in outpatient detoxification services, seeing drug counsellors or participation in addiction-related peer support meetings (e.g., Narcotics Anonymous, Cocaine Anonymous, etc.), and participation in opioid substitution therapies (i.e., methadone maintenance therapy or Suboxone treatment) (all: yes vs. no). The use of opioid substitution therapies was considered in the sub-analysis focusing on dual crack cocaine and opioid users only. Clinical variables included: CD4 cell count (per 100-cells/mm3 increase); antiretroviral tablet regimen (single tablet vs. 2-3 tablets vs. >3 tablets per day); years since HAART initiation (per year increase); physician experience on HIV care at the time of antiretroviral therapy initiation (having ever treated <6 patients vs. ≥6 patients). All variables except for the demographic characteristics referred to the previous six months unless otherwise indicated. Statistical analyses First, we examined baseline characteristics of the sample, stratified by medication adherence (≥95 % vs. <95 %) in the six months prior to the baseline assessment. Categorical variables were analysed using Pearson’s χ2 test (or Fisher’s exact test in the presence of small sell counts) and continuous variables were analysed using the Wilcoxon rank-sum test. Next, we used generalized linear mixed-effects modelling (GLMM) with the logit link to longitudinally identify factors associated with optimal pharmacy refill adherence and pVL non-detectability, respectively. This form of regression modelling accounts for the correlation between repeated observations gathered over time from the same individual, and estimates the effect of the explanatory variables on the likelihood of optimal pharmacy refill adherence and pVL non-detectability in each individual. Since our study aimed to identify the set of variables that best explains a higher odds of optimal pharmacy refill adherence and pVL non-detectability, respectively, we used a priori-defined statistical protocol based on the examination of the Akaike information criterion (AIC) and Type III p-values to construct multivariable GLMM logistic regression models. This procedure balances model selection on finding the best explanatory model with best model fit, as described previously [22]. In brief, we first included in the full multivariable models all explanatory variables that were significantly associated with optimal pharmacy refill adherence and pVL non-detectability, respectively, at the p < 0.10 level in the univariable analyses. After examining the AIC value of the models, we removed the variable with the largest p-value and built a reduced model. We continued this iterative process and selected the multivariable models with the lowest AIC value. Through this process, some variables, even though being significantly associated with an outcome at the level of p < 0.05 in bivariable analyses, were dropped from the final multivariable model. For the sub-analysis focusing on dual crack cocaine and opioid users, we added participation in opioid substitution therapies to the final multivariable models as a covariate. All p-values were two-sided. All statistical analyses were performed using the SAS software version 9.3 (SAS, Cary, NC). Results In total, 438 HAART-exposed crack cocaine users were included in this study, of whom 293 (66.9 %) were male and 239 (54.6 %) self-reported Caucasian ancestry. Of the 438 participants, 406 (92.7 %) had returned for at least one follow-up visit, contributing a median follow-up duration of 48.0 months (interquartile range [IQR]: 35.0 – 75.9). At baseline, 240 (54.8 %) exhibited optimal pharmacy refill adherence in the previous six months, and a total of 390 (89.0 %) individuals attained optimal adherence at some point during the study period. Table 1 presents sample characteristics at baseline, stratified by medication adherence levels. A substantial proportion of the sample was homeless (29.5 %), smoked crack cocaine at least on a daily basis (40.0 %), and engaged in sex work or drug dealing (37.4 %) in the previous six months. Very few individuals accessed any types of addiction treatment except for opioid substitution therapies (1.1–9.8 %), while the majority (63.5 %) of 293 dual crack cocaine and opioid users participated in opioid substitution therapies in the previous six months.Table 1 Baseline characteristics of HAART-exposed crack cocaine users in Vancouver, Canada, stratified by optimal pharmacy refill adherence (n = 438) Characteristic Total (n = 438) n (%) ≥95 % pharmacy refill adherence for antiretroviralsa p-value Yes 240 (54.8 %) No 198 (45.2 %) Age (median, IQR) 44 (38–49) 45 (40–51) 42 (37–46) <0.001 Gender  Male 293 (66.9 %) 171 (71.3 %) 122 (61.6 %) 0.033  Female 145 (33.1 %) 69 (28.7 %) 76 (38.4 %) reference Ancestry  Caucasian 239 (54.6 %) 147 (61.3 %) 92 (46.5 %) 0.002  Non-Caucasian 199 (45.4 %) 93 (38.7 %) 106 (53.5 %) reference Education   < Secondary education 228 (52.1 %) 116 (48.3 %) 112 (56.6 %) 0.069   ≥ Secondary education 196 (44.7 %) 117 (48.8 %) 79 (39.9 %) reference Homelessa  Yes 129 (29.5 %) 57 (23.8 %) 72 (36.4 %) 0.003  No 306 (69.9 %) 182 (75.8 %) 124 (62.6 %) reference Crack cocaine smokinga   ≥ Daily 175 (40.0 %) 78 (32.5 %) 97 (49.0 %) 0.001   < Daily 263 (60.0 %) 162 (67.5 %) 101 (51.0 %) reference Heroin usea   ≥ Daily 48 (11.0 %) 19 (7.9 %) 29 (14.7 %) 0.024   < Daily 389 (88.8 %) 221 (92.1 %) 168 (84.9 %) reference Crystal methamphetamine usea,c   ≥ Daily 9 (2.1 %) 5 (2.1 %) 4 (2.0 %) 1.000   < Daily 428 (97.7 %) 235 (97.9 %) 193 (97.5 %) reference Heavy alcohol use (≥5 drinks per day)a,c  Yes 13 (3.0 %) 5 (2.1 %) 8 (4.0 %) 0.266  No 425 (97.0 %) 235 (97.9 %) 190 (96.0 %) reference Engagement in sex work or drug dealinga  Yes 164 (37.4 %) 78 (32.5 %) 86 (43.4 %) 0.018  No 270 (61.6 %) 160 (66.7 %) 110 (55.6 %) reference Incarcerationa  Yes 58 (13.2 %) 32 (13.3 %) 26 (13.1 %) 0.951  No 380 (86.8 %) 208 (86.7 %) 172 (86.9 %) reference HIV physician experienceb  Previously treated <6 patients 48 (11.0 %) 25 (10.4 %) 23 (11.6 %) 0.397  Previously treated ≥6 patients 340 (77.6 %) 199 (82.9 %) 141 (71.2 %) reference CD4 cell counta (median, IQR) 300 (180–450) 340 (223.5–495) 235 (120–360) <0.001 Years since HAART initiationa (median, IQR) 5.7 (2.2–8.5) 4.9 (1.6–8.0) 6.9 (3.2–8.8) <0.001 Regimenb  Single tablet per day 34 (7.8 %) 24 (10.0 %) 10 (5.1 %) 0.278  2-3 tablets per day 121 (27.6 %) 76 (31.7 %) 45 (22.7 %) 0.723   > 3 tablets per day 230 (52.5 %) 140 (58.3 %) 90 (45.5 %) reference Participation in inpatient addiction treatmenta  Yes 35 (8.0 %) 19 (7.9 %) 16 (8.1 %) 0.950  No 403 (92.0 %) 221 (92.1 %) 182 (91.9 %) reference Participation in outpatient detoxification or other outpatient addiction treatmenta,c  Yes 5 (1.1 %) 4 (1.7 %) 1 (0.5 %) 0.384  No 433 (98.9 %) 236 (98.3 %) 197 (99.5 %) reference Seeing drug counsellors or participation in addiction-related peer support meetingsa  Yes 43 (9.8 %) 31 (12.9 %) 12 (6.1 %) 0.016  No 395 (90.2 %) 209 (87.1 %) 186 (93.9 %) reference Restricted to dual crack cocaine and opioid users (n = 293) Participation in opioid substitution therapiesa  Yes 186 (63.5 %) 115 (70.6 %) 71 (54.6 %) 0.005  No 107 (36.5 %) 48 (29.4 %) 59 (45.4 %) reference HAART highly active antiretroviral therapy, CI confidence interval, IQR interquartile range a denotes activities/events in the past 6 months b denotes at the time of antiretroviral therapy initiation c p-values were derived from Fisher’s exact test Not all cells added up to 100 % due to some missing values Compared to those who had suboptimal pharmacy refill adherence in the previous six months, individuals who had optimal pharmacy refill adherence were older, more likely to be male, self-report Caucasian ancestry, possess a high school diploma, have higher CD4 cell count and a shorter length of exposure to HAART, and have seen drug counsellors or participated in addiction-related peer support groups in the previous six months, while they were less likely to be homeless, smoke crack cocaine or use heroin at least daily, engage in sex work or drug dealing in the previous six months (all p < 0.05). Additionally, among a sub-sample of dual crack cocaine and opioid users, those who accessed opioid substitution therapies were more likely to have exhibited optimal pharmacy refill adherence (p = 0.005). Table 2 presents baseline characteristics stratified by pVL non-detectability. The observed significant associations were generally similar to those found with optimal pharmacy refill adherence, except that those who were on a 2-3 tablet-per-day regimen were more likely to have pVL non-detectability compared to those who were on a >3 tablet-per-day regimen (p = 0.008).Table 2 Baseline characteristics of HAART-exposed crack cocaine users in Vancouver, Canada, stratified by pVL non-detectability (n = 438) Characteristic Total (n = 438) n (%) pVL <50 copies/mLa p-value Yes 212 (48.4 %) No 226 (51.6 %) Age (median, IQR) 44 (38–49) 45 (41–51) 42 (36–47) <0.001 Gender  Male 293 (66.9 %) 149 (70.3 %) 144 (63.7 %) 0.145  Female 145 (33.1 %) 63 (29.7 %) 82 (36.3 %) reference Ancestry  Caucasian 239 (54.6 %) 122 (57.5 %) 117 (51.8 %) 0.225  Non-Caucasian 199 (45.4 %) 90 (42.5 %) 109 (48.2 %) reference Education   < Secondary education 228 (52.1 %) 107 (50.5 %) 121 (53.5 %) 0.462   ≥ Secondary education 196 (44.7 %) 99 (46.7 %) 97 (42.9 %) reference Homelessa  Yes 129 (29.5 %) 36 (17.0 %) 93 (41.2 %) <0.001  No 306 (69.9 %) 174 (82.1 %) 132 (58.4 %) reference Crack cocaine smokinga   ≥ Daily 175 (40.0 %) 72 (34.0 %) 103 (45.6 %) 0.013   < Daily 263 (60.0 %) 140 (66.0 %) 123 (54.4 %) reference Heroin usea   ≥ Daily 48 (11.0 %) 16 (7.5 %) 32 (14.2 %) 0.026   < Daily 389 (88.8 %) 196 (92.5 %) 193 (85.4 %) reference Crystal methamphetamine usea,c   ≥ Daily 9 (2.1 %) 6 (2.8 %) 3 (1.3 %) 0.326   < Daily 428 (97.7 %) 206 (97.2 %) 222 (98.2 %) reference Heavy alcohol use (≥5 drinks per day)a,c  Yes 13 (3.0 %) 4 (1.9 %) 9 (4.0 %) 0.263  No 425 (97.0 %) 208 (98.1 %) 217 (96.0 %) reference Engagement in sex work or drug dealinga  Yes 164 (37.4 %) 72 (34.0 %) 92 (40.7 %) 0.145  No 270 (61.6 %) 138 (65.1 %) 132 (58.4 %) reference Incarcerationa  Yes 58 (13.2 %) 25 (11.8 %) 33 (14.6 %) 0.386  No 380 (86.8 %) 187 (88.2 %) 193 (85.4 %) reference HIV physician experienceb  Previously treated <6 patients 48 (11.0 %) 21 (9.9 %) 27 (11.9 %) 0.418  Previously treated ≥6 patients 340 (77.6 %) 170 (80.2 %) 170 (75.2 %) reference CD4 cell counta (median, IQR) 300 (180–450) 360 (260–500) 230 (110–360) <0.001 Years since HAART initiationa (median, IQR) 5.7 (2.2–8.5) 5.9 (2.3–8.7) 5.5 (2.1–8.4) 0.243 Regimenb  Single tablet per day 34 (7.8 %) 18 (8.5 %) 16 (7.1 %) 0.713  2-3 tablets per day 121 (27.6 %) 78 (36.8 %) 43 (19.0 %) 0.008   > 3 tablets per day 230 (52.5 %) 114 (53.8 %) 116 (51.3 %) reference Participation in inpatient addiction treatmenta  Yes 35 (8.0 %) 14 (6.6 %) 21 (9.3 %) 0.300  No 403 (92.0 %) 198 (93.4 %) 205 (90.7 %) reference Participation in outpatient detoxification or other outpatient addiction treatmenta,c  Yes 5 (1.1 %) 4 (1.9 %) 1 (0.4 %) 0.203  No 433 (98.9 %) 208 (98.1 %) 225 (99.6 %) reference Seeing drug counsellors or participation in addiction-related peer support meetingsa  Yes 43 (9.8 %) 30 (14.2 %) 13 (5.8 %) 0.003  No 395 (90.2 %) 182 (85.8 %) 213 (94.2 %) reference Restricted to dual crack cocaine and opioid users (n = 293) Participation in opioid substitution therapiesa  Yes 186 (63.5 %) 98 (72.6 %) 88 (55.7 %) 0.003  No 107 (36.5 %) 37 (29.4 %) 70 (45.4 %) reference HAART highly active antiretroviral therapy, CI confidence interval, IQR interquartile range, pVL plasma HIV RNA viral load a denotes activities/events in the past 6 months b denotes at the time of antiretroviral therapy initiation c p-values were derived from Fisher’s exact test Not all cells added up to 100 % due to some missing values Table 3 shows the results of univariable and multivariable GLMM logistic regression analyses of factors associated with optimal pharmacy refill adherence. As shown, in univariable analyses, older age (odds ratio [OR]: 2.29; 95 % confidence interval [CI]: 1.84 – 2.86), male gender (OR: 1.50; 95 % CI: 1.02 – 2.20), Caucasian ethnicity (OR: 1.61; 95 % CI: 1.12 – 2.32), higher CD4 count (OR: 1.39; 95 % CI: 1.30 – 1.48) and longer history of HAART (OR: 1.08; 95 % CI: 1.04 – 1.12) were significantly and positively associated with optimal pharmacy refill adherence. Homelessness (OR: 0.44; 95 % CI: 0.34 – 0.58), at least daily crack cocaine smoking (OR: 0.49; 95 % CI: 0.39 – 0.62), at least daily heroin use (OR: 0.32; 95 % CI: 0.22 – 0.47), and engagement in sex work or drug dealing (OR: 0.59; 95 % CI: 0.47 – 0.76) were significantly and negatively associated with optimal pharmacy refill adherence.Table 3 Univariable and multivariable GLMM logistic regression analyses of factors associated with optimal pharmacy refill adherence for antiretroviral medications among HAART-exposed crack cocaine users in Vancouver, Canada, 2005–2013 Characteristic Total sample of crack cocaine users (n = 438) Dual crack cocaine and opioid users (n = 293) Unadjusted OR (95 % CI) p-value Adjusted OR (95 % CI) p-value Unadjusted OR (95 % CI) p-value Adjusted OR (95 % CI) p-value Ageb  (per 10-year older) 2.29 (1.84 – 2.86) <0.001 1.65 (1.33 – 2.04) <0.001 2.75 (2.06 – 3.68) <0.001 1.72 (1.30 – 2.26) <0.001 Gender  (male vs. female) 1.50 (1.02 – 2.20) 0.041 1.10 (0.68 – 1.79) 0.696 Ethnicity  (Caucasian vs. other) 1.61 (1.12 – 2.32) 0.011 1.40 (0.87 – 2.24) 0.167 Education  (<secondary vs. ≥secondary) 0.83 (0.57 – 1.20) 0.327 0.98 (0.61 – 1.58) 0.943 Homelessa,b  (yes vs. no) 0.44 (0.34 – 0.58) <0.001 0.58 (0.44 – 0.77) <0.001 0.39 (0.28 – 0.54) <0.001 0.56 (0.40 – 0.79) <0.001 Crack cocaine smokinga,b  (≥daily vs. <daily) 0.49 (0.39 – 0.62) <0.001 0.64 (0.50 – 0.81) <0.001 0.41 (0.31 – 0.55) <0.001 0.57 (0.43 – 0.75) <0.001 Heroin usea,b  (≥daily vs. <daily) 0.32 (0.22 – 0.47) <0.001 0.43 (0.29 – 0.65) <0.001 0.30 (0.20 – 0.46) <0.001 0.45 (0.29 – 0.70) <0.001 Crystal methamphetamine usea,b  (≥daily vs. <daily) 1.07 (0.50 – 2.30) 0.858 0.79 (0.32 – 1.94) 0.609 Heavy alcohol usea,b  (≥5 drinks per day vs. <5 drinks per day) 1.17 (0.65 – 2.10) 0.596 1.37 (0.60 – 3.10) 0.452 Engagement in sex work or drug dealinga,b  (yes vs. no) 0.59 (0.47 – 0.76) <0.001 0.54 (0.41 – 0.71) <0.001 Incarcerationa,b  (yes vs. no) 0.71 (0.49 – 1.02) 0.063 0.63 (0.41 – 0.96) 0.031 HIV physician experiencec  (<6 patients vs. ≥6 patients) 0.74 (0.40 – 1.35) 0.322 0.60 (0.29 – 1.24) 0.169 CD4 cell counta,b  (per 100-cell increase) 1.39 (1.30 – 1.48) <0.001 1.31 (1.23 – 1.41) <0.001 1.43 (1.32 – 1.55) <0.001 1.34 (1.23 – 1.45) <0.001 Years since HAART initiationa,b  (per year increase) 1.08 (1.04 – 1.12) <0.001 1.11 (1.06 – 1.16) <0.001 Regimena,b  (1 pill per day vs. >3 pills per day) 1.23 (0.80 – 1.90) 0.339 1.29 (0.76 – 2.19) 0.347  (2-3 pills per day vs. >3 pills per day) 1.01 (0.79 – 1.30) 0.920 1.07 (0.78 – 1.48) 0.659 Participation in inpatient addiction treatmenta,b  (yes vs. no) 0.82 (0.57 – 1.18) 0.294 0.84 (0.54 – 1.32) 0.454 Participation in outpatient detoxification or other outpatient addiction treatmenta,b  (yes vs. no) 0.65 (0.20 – 2.09) 0.472 0.43 (0.09 – 2.02) 0.284 Seeing drug counsellors or participation in addiction-related peer support meetingsa,b  (yes vs. no) 1.06 (0.76 – 1.48) 0.730 0.84 (0.55 – 1.28) 0.424 Participation in opioid substitution therapiesa,b  (yes vs. no) ––– ––– 1.71 (1.23 – 2.38) 0.001 1.39 (0.99 – 1.96) 0.056 GLMM generalized linear mixed-effect modelling, HAART highly active antiretroviral therapy, OR odds ratio, CI confidence interval a denotes activities/events in the past 6 months b denotes time-varying variables c denotes at the time of antiretroviral therapy initiation In the final multivariable model, older age (adjusted odds ratio [AOR]: 1.65; 95 % CI: 1.33 – 2.04) and higher CD4 count (AOR: 1.31; 95 % CI: 1.23 – 1.41) remained independently and positively associated with optimal pharmacy refill adherence, while homelessness (AOR: 0.58; 95 % CI: 0.44 – 0.77), at least daily crack cocaine smoking (AOR: 0.64; 95 % CI: 0.50 – 0.81), and at least daily heroin use (AOR: 0.43; 95 % CI: 0.29 – 0.65) remained independently and negatively associated with optimal pharmacy refill adherence. When the sample was restricted to dual crack cocaine and opioid users, the same set of variables remained significantly associated with optimal pharmacy refill adherence. Participation in opioid substitution therapies was not independently associated with optimal pharmacy refill adherence (AOR: 1.39; 95 % CI: 0.99 – 1.96). Table 4 presents the results univariable and multivariable GLMM logistic regression analyses of factors associated with pVL non-detectability. As shown, in addition to the covariates that were included for medication adherence, the final multivariable models included years since HAART initiation and antiretroviral tablet regimen. Longer history of HAART (AOR: 1.06; 95 % CI: 1.01 – 1.12) and receiving a single tablet-per-day regimen (AOR: 3.02; 95 % CI: 1.67 – 5.46) were independently and positively associated with pVL non-detectability. The results for the other covariates were consistent with those found in the multivariable model of medication adherence except that among dual crack cocaine and opioid users, participation in opioid substitution therapies was independently and positively associated with pVL non-detectability (AOR: 1.55; 95 % CI: 1.02 – 2.34).Table 4 Univariable and multivariable GLMM logistic regression analyses of factors associated with pVL non-detectability among HAART-exposed crack cocaine users in Vancouver, Canada, 2005–2013 Characteristic Total sample of crack cocaine users (n = 438) Dual crack cocaine and opioid users (n = 293) Unadjusted OR (95 % CI) p-value Adjusted OR (95 % CI) p-value Unadjusted OR (95 % CI) p-value Adjusted OR (95 % CI) p-value Ageb  (per 10-year older) 5.24 (3.72 – 7.37) <0.001 1.86 (1.37 – 2.52) <0.001 5.22 (3.45 – 7.89) <0.001 1.59 (1.11 – 2.26) 0.011 Gender  (male vs. female) 1.86 (1.10 – 3.15) 0.020 1.08 (0.59 – 1.97) 0.804 Ethnicity  (Caucasian vs. other) 1.31 (0.79 – 2.17) 0.290 1.38 (0.77 – 2.49) 0.282 Education  (<secondary vs. ≥secondary) 1.03 (0.62 – 1.72) 0.909 1.39 (0.77 – 2.49) 0.272 Homelessa,b  (yes vs. no) 0.26 (0.19 – 0.36) <0.001 0.39 (0.28 – 0.54) <0.001 0.24 (0.17 – 0.35) <0.001 0.37 (0.25 – 0.56) <0.001 Crack cocaine smokinga,b  (≥daily vs. <daily) 0.47 (0.36 – 0.61) <0.001 0.39 (0.29 – 0.53) <0.001 0.69 (0.49 – 0.96) 0.028 Heroin usea,b  (≥daily vs. <daily) 0.35 (0.23 – 0.54) <0.001 0.60 (0.36 – 0.98) 0.042 0.32 (0.20 – 0.51) <0.001 Crystal methamphetamine usea,b  (≥daily vs. <daily) 1.09 (0.45 – 2.62) 0.853 1.15 (0.43 – 3.10) 0.779 Heavy alcohol usea,b  (≥5 drinks per day vs. <5 drinks per day) 0.70 (0.36 – 1.37) 0.302 0.57 (0.24 – 1.37) 0.209 Engagement in sex work or drug dealinga,b  (yes vs. no) 0.58 (0.44 – 0.75) <0.001 0.58 (0.43 – 0.78) <0.001 Incarcerationa,b  (yes vs. no) 0.57 (0.38 – 0.86) 0.008 0.49 (0.31 – 0.78) 0.002 HIV physician experiencec (<6 patients vs. ≥6 patients) 1.15 (0.51 – 2.62) 0.738 0.84 (0.35 – 2.00) 0.688 CD4 cell counta,b  (per 100-cell increase) 1.93 (1.76 – 2.13) <0.001 1.67 (1.52 – 1.83) <0.001 1.98 (1.77 – 2.21) <0.001 1.72 (1.54 – 1.91) <0.001 Years since HAART initiationa,b  (per year increase) 1.27 (1.21 – 1.34) <0.001 1.06 (1.01 – 1.12) 0.012 1.29 (1.22 – 1.37) <0.001 1.08 (1.02 – 1.14) 0.011 Regimena,b  (1 pill per day vs. >3 pills per day) 3.11 (1.76 – 5.52) <0.001 3.02 (1.67 – 5.46) <0.001 2.84 (1.45 – 5.57) 0.003 2.60 (1.30 – 5.21) 0.007  (2-3 pills per day vs. >3 pills per day) 1.70 (1.27 – 2.30) <0.001 1.24 (0.91 – 1.69) 0.179 1.60 (1.13 – 2.29) 0.009 1.14 (0.79 – 1.65) 0.493 Participation in inpatient addiction treatmenta,b  (yes vs. no) 0.67 (0.45 – 1.02) 0.061 0.58 (0.35 – 0.97) 0.037 Participation in outpatient detoxification or other outpatient addiction treatmenta,b  (yes vs. no) 1.10 (0.33 – 3.71) 0.872 1.63 (0.34 – 7.74) 0.539 Seeing drug counsellors or participation in addiction-related peer support meetingsa,b  (yes vs. no) 1.32 (0.90 – 1.94) 0.158 1.16 (0.73 – 1.84) 0.532 Participation in opioid substitution therapiesa,b  (yes vs. no) ––– ––– 2.28 (1.58 – 3.30) <0.001 1.55 (1.02 – 2.34) 0.039 GLMM generalized linear mixed-effect modelling, pVL plasma HIV RNA viral load, HAART highly active antiretroviral therapy, OR odds ratio, CI confidence interval a denotes activities/events in the past 6 months b denotes time-varying variables c denotes at the time of antiretroviral therapy initiation Discussion Among our sample of HAART-exposed crack cocaine users, we found that homelessness, and high-intensity crack cocaine and heroin use were independently and negatively associated with optimal pharmacy refill adherence and pVL non-detectability. Except for opioid substitution therapies, none of the addiction treatment modalities assessed, including inpatient treatment, outpatient detoxification services, drug counselling, and participation in peer support groups such as Narcotics Anonymous, were significantly associated with optimal pharmacy refill adherence or pVL non-detectability. Specifically among a sub-sample of dual crack cocaine and opioid users, enrolment in opioid substitution therapies was independently associated with pVL non-detectability but not with optimal pharmacy refill adherence. We found an independent association between at least daily use of crack cocaine and poor medication adherence and having a detectable viral load. This is consistent with previous reports that crack cocaine users are less likely to enjoy gains from HAART [13–15]. In addition, the finding that at least daily use of heroin was also independently associated with suboptimal pharmacy refill adherence and outcome indicates a need to address poly-substance use among HIV-infected crack cocaine users. Previous research suggested that poly-substance use is quite common among people who use drugs, although those who primarily use cocaine may continue to use it consistently over an extended period of time while keeping the use of other non-primary drugs at relatively lower levels [23]. While current literature suggests no clinically significant pharmacokinetic interactions between crack cocaine and antiretroviral agents, methamphetamines may have possible lethal interactions with antiretroviral agents that inhibit the CYP450 system [24]. In our sample of HIV-positive crack cocaine users, high-intensity use of crystal methamphetamine was not common or associated with optimal pharmacy refill adherence or pVL non-detectability; however, in a setting where co-use of crack cocaine and crystal methamphetamine is common, a greater caution is required when prescribing HAART. Collectively, these findings emphasize the need to ensure that a variety of addiction treatment options be integrated with HIV treatment and care for individuals who use crack cocaine. The finding that none of the addiction treatment modalities except for opioid substitution therapies were associated with optimal pharmacy refill adherence or viral suppression serves to underscore previous calls for identifying more evidence-based addiction treatment approaches that also confer benefits on HIV disease management among HIV-positive crack cocaine users [5, 14, 25]. While a previous pilot study suggested that a video-based behavioural intervention employing the Information-Motivation-Behavioural Skill model might be efficacious in reducing crack cocaine use and improving medication adherence among HIV-positive crack cocaine users [26], the efficacy of the intervention needs to be determined with a larger sample. Also, the long-term post-intervention effect on crack cocaine use and medication adherence is an important outcome to be examined, as a Cochrane Collaboration systematic review has reported that many behavioural interventions for substance use suffer from a lack of the sustained effect following the intervention [27]. Further, unlike the case of opiate addiction, there is currently no approved medication to treat crack cocaine addiction. While many candidate medications have been evaluated for the efficacy in the treatment of crack cocaine/cocaine addiction during the past decade, overall results have been inconclusive [28–30]. Further research is needed in this area. In contrast, we found that dual crack cocaine and opioid users who were enrolled in opioid substitution therapies were more likely to achieve virological success, and the positive effect of opioid substitution therapies remained significant even after the adjustment for the high-intensity use of crack cocaine and heroin. This adds important evidence to a body of research indicating the effectiveness of these therapies for the improvement of HIV treatment adherence and outcomes [31]. Previous studies reported that ongoing cocaine use undermined the effectiveness of opioid substitution therapies for HIV treatment-related outcomes among HIV-infected patients accessing these therapies [32, 33]. However, our findings suggest that opioid substitution therapies still help HIV-infected dual crack cocaine and opioid users achieve optimal HAART outcomes. Nonetheless, as crack cocaine/cocaine users have been shown to have significant barriers to adherence and continuance of opioid substitution therapies [34, 35], adjunct interventions to opioid substitution therapies that help retain crack cocaine/cocaine-using patients in treatment need to be developed. Similarly to our previous investigation that indicated that homelessness posed a barrier to antiretroviral therapy among HIV-infected people who use illicit drugs generally [36], we found that homelessness was independently associated with suboptimal pharmacy refill adherence and viral suppression failures among HIV-infected individuals who use crack cocaine in a setting where HAART is provided free of charge. Collectively, these findings indicate that, in addition to identification of evidence-based addiction treatment strategies for crack cocaine use, structural interventions to address homelessness in general (e.g., housing assistance) are needed. Expanding and promoting support services for HIV-positive homeless individuals focusing on both access and adherence to HAART would maximise the benefits of HAART on disease progression and onward viral transmission in this population. Our study has several limitations. First, as the study sample was not randomly selected, our findings may not be generalizable to other populations of crack cocaine users. Second, the self-reported data may be affected by reporting biases, including recall bias and socially desirable responding. However, we note that this type of data has been commonly utilized in observational studies involving people who use drugs and found to be valid [37]. Also, as our primary outcomes were ascertained through a confidential record linkage to the provincial HIV-related clinical database, we believe that it is unlikely that potential reporting biases have impacted the data differentially by the medication adherence or pVL levels. Third, as with all observational studies, the relationships between the explanatory variables and outcome assessed may be under the influence of unobserved confounding, although we sought to address this bias with multivariable adjustment of the key demographic, behavioural, and clinical predictors of medication adherence or viral suppression. Conclusions In sum, among our sample of HAART-exposed crack cocaine users, those who were homeless and those who engaged in high-intensity crack cocaine or heroin use were more likely to have suboptimal adherence to antiretroviral medications and fail to achieve viral suppression. Further, none of the addiction treatment modalities assessed appeared to facilitate medication adherence or viral suppression, except for opioid substitution therapies. These findings suggest an urgent need to develop evidence-based treatment options for crack cocaine use that also confer benefits to HAART-related outcomes. Abbreviations ACCESS, the AIDS Care Cohort to evaluate Exposure to Survival Services; AIC, Akaike information criterion; AOR, adjusted odds ratio; GLMM, generalized linear mixed-effects modelling; HAART, Highly-active antiretroviral therapy; IQR, interquartile range; OR, odds ratio; pVL, plasma HIV RNA viral load; TasP, Treatment as Prevention Acknowledgements The authors thank the study participants for their contribution to the research, as well as current and past researchers and staff. Funding The study was supported by the US National Institutes of Health (R01DA021525). This research was undertaken, in part, thanks to funding from the Canada Research Chairs program through a Tier 1 Canada Research Chair in Inner City Medicine which supports Dr. Wood. Dr. Hayashi is supported by the Canadian Institutes of Health Research New Investigator Award (MSH-141971). Dr. Milloy is supported in part by the US National Institutes of Health (R01DA021525). Dr. Montaner’s TasP research has received support from the BC Ministry of Health, US National Institutes of Health (NIDA R01DA036307), International AIDS Society, UNAIDS, WHO, ANRS, IAPAC, UNICEF, MAC AIDS Fund and Open Society Foundations. Institutional grants have been provided by Abbvie, BI, BMS, Gilead Sciences, J&J, Merck and ViiV. He has served on Advisory Boards for Teva, Gilead Sciences and InnaVirVax. The funding bodies had no role in the study design, data collection, analysis, interpretation, or manuscript writing. Availability of data and materials The data used for this study is not publicly available. For further information on the data and materials used in this study, please contact the corresponding author. Authors’ contributions KH and MJM designed the study. PN and HD conducted the statistical analyses. KH drafted the manuscript, and incorporated suggestions from all co-authors. All authors made significant contributions to the conception of the analyses, interpretation of the data, and drafting of the manuscript. All authors approved the final version of the manuscript. Competing interests The authors declare that they have no competing interests. Consent for publication Not applicable. Ethics approval and consent to participate All participants provided written informed consent for study participation. The ACCESS study received annual ethics approval from the University of British Columbia and Providence Health Care Research Ethics Board (H05-50233). The PI of ACCESS (MJM) granted permission to use the data for the present study, which was part of the larger ACCESSS study activities approved by the research ethics board. The dataset used for the present study was de-identified. The data were located at the British Columbia Centre for Excellence in HIV/AIDS. ==== Refs References 1. Nakagawa F May M Phillips A Life expectancy living with HIV: recent estimates and future implications Curr Opin Infect Dis 2013 26 1 17 25 10.1097/QCO.0b013e32835ba6b1 23221765 2. Montaner JS Lima VD Barrios R Yip B Wood E Kerr T Association of highly active antiretroviral therapy coverage, population viral load, and yearly new HIV diagnoses in British Columbia, Canada: a population-based study Lancet 2010 376 9740 532 9 10.1016/S0140-6736(10)60936-1 20638713 3. Cohen MS Chen YQ McCauley M Gamble T Hosseinipour MC Kumarasamy N Prevention of HIV-1 infection with early antiretroviral therapy N Engl J Med 2011 365 6 493 505 10.1056/NEJMoa1105243 21767103 4. Des Jarlais DC Mcknight C Arasteh K Feelemyer J Perlman DC Hagan H A perfect storm: crack cocaine, HSV-2, and HIV among non-injecting drug users in New York City Subst Use Misuse 2014 49 7 783 92 10.3109/10826084.2014.880176 24502371 5. Wechsberg WM Golin C El-Bassel N Hopkins J Zule W Current interventions to reduce sexual risk behaviors and crack cocaine use among HIV-infected individuals Curr HIV/AIDS Rep 2012 9 4 385 93 10.1007/s11904-012-0131-y 22872433 6. Edlin BR Irwin KL Faruque S McCoy CB Word C Serrano Y Intersecting epidemics--crack cocaine use and HIV infection among inner-city young adults. Multicenter Crack Cocaine and HIV Infection Study Team N Engl J Med 1994 331 21 1422 7 10.1056/NEJM199411243312106 7969281 7. DeBeck K Kerr T Li K Fischer B Buxton J Montaner J Smoking of crack cocaine as a risk factor for HIV infection among people who use injection drugs CMAJ 2009 181 9 585 9 10.1503/cmaj.082054 19841052 8. World Health Organization, United Nations Office on Drugs and Crime, Joint United Nations Programme on HIV/AIDS WHO, UNODC, UNAIDS technical guide for countries to set targets for universal access to HIV prevention, treatment and care for injecting drug users: 2012 revision 2012 Geneva World Health Organization 9. Shoptaw S Montgomery B Williams CT El-Bassel N Aramrattana A Metsch L Not just the needle: the state of HIV-prevention science among substance users and future directions J Acquir Immune Defic Syndr. 2013 63 Suppl 2 S174 8 10.1097/QAI.0b013e3182987028 23764632 10. Meader N Li R Des Jarlais DC Pilling S Psychosocial interventions for reducing injection and sexual risk behaviour for preventing HIV in drug users Cochrane Database Syst Rev. 2010 1 CD007192 20091623 11. Joint United Nations Programme on HIV/AIDS Fast-track: ending the AIDS epidemic by 2030 2014 Geneva Joint United Nations Programme on HIV/AIDS 12. Montaner JS Wood E Kerr T Lima V Barrios R Shannon K Expanded highly active antiretroviral therapy coverage among HIV-positive drug users to improve individual and public health outcomes J Acquir Immune Defic Syndr. 2010 55 Suppl 1 S5 9 10.1097/QAI.0b013e3181f9c1f0 21045601 13. Arnsten JH Demas PA Grant RW Gourevitch MN Farzadegan H Howard AA Impact of active drug use on antiretroviral therapy adherence and viral suppression in HIV-infected drug users J Gen Intern Med 2002 17 5 377 81 10.1007/s11606-002-0044-3 12047736 14. Gonzalez A Barinas J O’Cleirigh C Substance use: impact on adherence and HIV medical treatment Curr HIV/AIDS Rep 2011 8 4 223 34 10.1007/s11904-011-0093-5 21858414 15. Hinkin CH Barclay TR Castellon SA Levine AJ Durvasula RS Marion SD Drug use and medication adherence among HIV-1 infected individuals AIDS Behav 2007 11 2 185 94 10.1007/s10461-006-9152-0 16897351 16. Strathdee SA Palepu A Cornelisse PGA Yip B O'Shaughnessy MV Montaner JSG Barriers to use of free antiretroviral therapy in injection drug users JAMA 1998 280 6 547 9 10.1001/jama.280.6.547 9707146 17. Hogg RS Heath KV Yip B Craib KJ O'Shaughnessy MV Schechter MT Improved survival among HIV-infected individuals following initiation of antiretroviral therapy JAMA 1998 279 6 450 4 10.1001/jama.279.6.450 9466638 18. Hogg RS Yip B Chan KJ Wood E Craib KJP O'Shaughnessy MV Rates of disease progression by baseline CD4 cell count and viral load after initiating triple-drug therapy JAMA 2001 286 20 2568 77 10.1001/jama.286.20.2568 11722271 19. Low-Beer S Yip B O'Shaughnessy MV Hogg RS Montaner JS Adherence to triple therapy and viral load response J Acquir Immune Defic Syndr 2000 23 4 360 1 10.1097/00126334-200004010-00016 10836763 20. Wood E Montaner JSG Yip B Tyndall MW Schechter MT O'Shaughnessy MV Adherence and plasma HIV RNA responses to highly active antiretroviral therapy among HIV-1 infected injection drug users CMAJ 2003 169 7 656 61 14517122 21. Wood E Hogg RS Yip B Harrigan PR O'Shaughnessy MV Montaner JSG Effect of medication adherence on survival of HIV-infected adults who start highly active antiretroviral therapy when the CD4+ cell count is 0.200 to 0.350 × 10(9) cells/L Ann Intern Med 2003 139 10 810 6 10.7326/0003-4819-139-10-200311180-00008 14623618 22. Lima VD Harrigan R Murray M Moore DM Wood E Hogg RS Differential impact of adherence on long-term treatment response among naive HIV-infected individuals AIDS 2008 22 17 2371 80 10.1097/QAD.0b013e328315cdd3 18981777 23. Brecht M-L Huang D Evans E Hser Y-I Polydrug use and implications for longitudinal research: ten-year trajectories for heroin, cocaine, and methamphetamine users Drug Alcohol Depend 2008 96 3 193 201 10.1016/j.drugalcdep.2008.01.021 18329825 24. Bruce RD Altice FL Friedland GH Pharmacokinetic drug interactions between drugs of abuse and antiretroviral medications: implications and management for clinical practice Expert Rev Clin Pharmacol 2008 1 1 115 27 10.1586/17512433.1.1.115 24410515 25. Metzger DS Woody GE O'Brien CP Drug treatment as HIV prevention: a research update J Acquir Immune Defic Syndr. 2010 55 S32 6 10.1097/QAI.0b013e3181f9c10b 21045597 26. Ingersoll KS Farrell-Carnahan L Cohen-Filipic J Heckman CJ Ceperich SD Hettema J A pilot randomized clinical trial of two medication adherence and drug use interventions for HIV+ crack cocaine users Drug Alcohol Depend 2011 116 1-3 177 87 10.1016/j.drugalcdep.2010.12.016 21306837 27. Knapp WP Soares BGO Farrel M Lima MS Psychosocial interventions for cocaine and psychostimulant amphetamines related disorders Cochrane Database Syst Rev. 2007 3 CD003023 17636713 28. Castells X Casas M Perez-Mana C Roncero C Vidal X Capella D Efficacy of psychostimulant drugs for cocaine dependence Cochrane Database Syst Rev. 2010 2 CD007380 20166094 29. Minozzi S Amato L Davoli M Farrell M Lima Reisser AARL Pani PP Anticonvulsants for cocaine dependence Cochrane Database Syst Rev. 2008 2 CD006754 18425968 30. Kim JH Lawrence AJ Drugs currently in Phase II clinical trials for cocaine addiction Expert Opin Investig Drugs 2014 23 8 1105 22 10.1517/13543784.2014.915312 24773297 31. Spire B Lucas GM Carrieri MP Adherence to HIV treatment among IDUs and the role of opioid substitution treatment (OST) Int J Drug Policy 2007 18 4 262 70 10.1016/j.drugpo.2006.12.014 17689374 32. Roux P Carrieri MP Villes V Dellamonica P Poizot-Martin I Ravaux I The impact of methadone or buprenorphine treatment and ongoing injection on highly active antiretroviral therapy (HAART) adherence: evidence from the MANIF2000 cohort study Addiction 2008 103 11 1828 36 10.1111/j.1360-0443.2008.02323.x 18778390 33. Gonzalez A Mimiaga MJ Israel J Andres Bedoya C Safren SA Substance use predictors of poor medication adherence: the role of substance use coping among HIV-infected patients in opioid dependence treatment AIDS Behav 2013 17 1 168 73 10.1007/s10461-012-0319-6 23008124 34. Roux P Lions C Michel L Cohen J Mora M Marcellin F Predictors of non-adherence to methadone maintenance treatment in opioid-dependent individuals: implications for clinicians Curr Pharm Des 2014 20 25 4097 105 10.2174/13816128113199990623 24001291 35. Marsden J Eastwood B Bradbury C Dale-Perera A Farrell M Hammond P Effectiveness of community treatments for heroin and crack cocaine addiction in England: a prospective, in-treatment cohort study Lancet 2009 374 9697 1262 70 10.1016/S0140-6736(09)61420-3 19800681 36. Milloy MJ Kerr T Bangsberg DR Buxton J Parashar S Guillemi S Homelessness as a structural barrier to effective antiretroviral therapy among HIV-seropositive illicit drug users in a Canadian setting AIDS Patient Care STDS 2012 26 1 60 7 10.1089/apc.2011.0169 22107040 37. Darke S Self-report among injecting drug users: a review Drug Alcohol Depend 1998 51 3 253 63 10.1016/S0376-8716(98)00028-3 9787998

PMC005xxxxxx/PMC5002323.txt

==== Front BMC Public HealthBMC Public HealthBMC Public Health1471-2458BioMed Central London 354110.1186/s12889-016-3541-8Research ArticleRational use of antibiotics by community health workers and caregivers for children with suspected pneumonia in Zambia: a cross-sectional mixed methods study Graham Kirstie kirstiejanegraham@gmail.com 1Sinyangwe Chomba csinyangwe@yahoo.com 2Nicholas Sarala saralanicholas@gmail.com 1King Rebecca r.king@leeds.ac.uk 3Mukupa Samuel smukupa@yahoo.com 2Källander Karin k.kallander@malariaconsortium.org 1Counihan Helen h.counihan@malariaconsortium.org 1Montague Mark mark.montague@rescue-uk.org 4Tibenderana James j.tibenderana@malariaconsortium.org 4Hamade Prudence p.hamade@malariaconsortium.org 11 Malaria Consortium, Development House, 56-64 Leonard Street, London, EC2A 4LT UK 2 Malaria Consortium, Mansa, Luapula Province Zambia 3 Nuffield Centre for International Health and Development, University of Leeds, Leeds, UK 4 Malaria Consortium, Kampala, Uganda 27 8 2016 27 8 2016 2016 16 1 89718 2 2016 18 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Antibiotic resistance is an issue of growing global concern. One key strategy to minimise further development of resistance is the rational use of antibiotics, by providers and patients alike. Through integrated community case management (iCCM), children diagnosed with suspected pneumonia are treated with antibiotics; one component of an essential package to reduce child mortality and increase access to health care for remote populations. Through the use of clinical algorithms, supportive supervision and training, iCCM also offers the opportunity to improve the rational use of antibiotics and limit the spread of resistance in resource-poor contexts. This study provides evidence on antibiotic use by community health workers (CHWs) and caregivers to inform iCCM programmes, safeguarding current treatments whilst maximising access to care. Methods 1497 CHW consultations were directly observed by non-clinical researchers, with measurement of respiratory rate by CHWs recorded by video. Videos were used to conduct a retrospective reference standard assessment of respiratory rate by experts. Fifty-five caregivers whose children were prescribed a 5-day course of antibiotics for suspected pneumonia were followed up on day six to assess adherence through structured interviews and pill counts. Six focus group discussions and nine in depth interviews were conducted with CHWs and caregivers to supplement quantitative findings. Results The findings indicate that CHWs adhered to treatment guidelines for 92 % of children seen, prescribing treatment corresponding to their assessment. However, only 65 % of antibiotics prescribed were given for children with experts’ confirmed fast breathing pneumonia. Qualitative data indicates that CHWs have a good understanding of pneumonia diagnosis, and although caregivers sometimes applied pressure to receive drugs, CHWs stated that treatment decisions were not influenced. 46 % of caregivers were fully adherent and gave their child the full 5-day course of dispersible amoxicillin. If caregivers who gave treatment for 3 to 5 days were considered, adherence increased to 76 %. Conclusions CHWs are capable of prescribing treatment corresponding to their assessment of respiratory rate. However, rational use of antibiotics could be strengthened through improved respiratory rate assessment, and better diagnostic tools. Furthermore, a shorter course of dispersible amoxicillin could potentially improve caregiver adherence, reducing risk of resistance and cost. Keywords PneumoniaCommunity health workersIntegrated community case managementRational useAntibioticsAdherenceCaregiversCOMDIS-HSDMalaria ConsortiumDFID UKaidCOMDIS-HSD RPCissue-copyright-statement© The Author(s) 2016 ==== Body Background In recent years, there has been increasing attention given to the issue of antibiotic resistance, as demonstrated by the World Health Organization (WHO) report on antimicrobial resistance and surveillance [1], and the global action plan on antimicrobial resistance [2], which was endorsed at the World Health Assembly in May 2015. The WHO report called for ‘concerted cross-sectional action by governments and society as a whole’ and described the issue as a ‘global health security threat’ [1]. With limited numbers of antibiotics available, and resistance to first-line treatments becoming increasingly common, resistance results in higher health care costs, as patients require more expensive and longer treatments, and a greater risk of mortality from diseases that were previously curable. In 2011, WHO highlighted a six-point policy package to combat the problem, including regulating and promoting the rational use of medicines [3]. The rational use of antibiotics not only concerns the actions of providers, in ensuring patients receive appropriate treatment for their condition, at the right dose and duration, but also those of patients, in adhering to the treatment regimens prescribed, completing the full course and not sharing or storing medicines for future use. In 2001, the WHO Global Strategy for containment of antimicrobial resistance [4] highlighted the need for the ‘development and use of guidelines and treatment algorithms to foster appropriate use of antimicrobials’, as well as the importance of ‘supervision and support of clinical practices, especially diagnostic and treatment strategies’. Furthermore, it highlighted that health care providers have an important role to play in educating patients on the importance of treatment adherence [4]. In many resource-poor contexts, with limited surveillance and weaker regulatory systems [5, 6], containment of resistance may be more challenging than in higher-income settings. However, in these settings, there remains a high burden of childhood infectious diseases, such as pneumonia, that often require antibiotic treatment. Pneumonia is responsible for the second largest number of child deaths globally, with a disproportionate burden and the highest number of deaths occurring in sub-Saharan Africa [7, 8]. Frequently, community members obtain antibiotics of unknown quality, without a prescription from local pharmacies, informal drug vendors, or private facilities, which are typically difficult to regulate and where prescription of medication is inherently linked to payment received [5, 9]. Integrated community case management (iCCM), a package of essential care delivered by community health workers (CHWs) for treatment of three key childhood diseases; pneumonia, malaria and diarrhoea, and often also the detection of malnutrition, aims to reduce childhood mortality and increase access to care for hard-to-reach populations. It also has the potential to encourage the rational use of antibiotics at community level and control the development or spread of antibiotic resistance through training, the use of simple job aids with effective clinical algorithms, diagnostic tools, availability of blister packaged (and sometimes colour coded) appropriate antibiotics, and supportive supervision, with CHWs acting as gatekeepers to these essential medicines. Furthermore, CHWs may be able to encourage and counsel caregivers to adhere to correct treatment regimens, for example, giving instructions regarding the number of doses to be given per day and the number of tablets per dose. However, there remains a perceived risk that CHWs distributing antibiotics may add to selective pressure among pathogens [10], i.e. the misuse of antibiotics selectively encourages the growth of antibiotic resistant bacteria, and thus more evidence is needed surrounding the use of antibiotics at community level and specifically relating to iCCM. This study sought to provide evidence on the use of antibiotics by CHWs and caregivers for the treatment of suspected pneumonia in children under five to inform future implementation of iCCM programmes to help preserve the effectiveness of current antibiotics, whilst at the same time continuing to provide access to care for remote and marginalised populations [6]. This paper presents quantitative data on CHWs’ adherence to guidelines in prescribing antibiotics, the rational use of antibiotics including dosage prescribed, and caregivers’ adherence in administering the course of treatment prescribed. In addition, qualitative data helps to provide context, exploring some of the reasons that may influence adherence by both populations respectively, for example, community perceptions on CHW prescribing practices, and caregivers’ expectation of care provided. Methods Study setting Data collection was conducted in Samfya and Kawambwa districts of Luapula province in northern Zambia between October and December 2012. The two study districts are situated in the south and the north of the province, respectively, with populations of 191,980 and 130,680 [11], and are representative of the other districts in the province. iCCM was implemented in Luapula province from 2009 to 2013, funded by the Canadian International Development Agency (CIDA), with a total of 440 CHWs trained in iCCM in these two districts. In this context, CHWs are typically non-clinical unpaid community volunteers, who have received little formal training. At a minimum, all CHWs involved in this study had received the 6-day iCCM training supported by Malaria Consortium, as part of the iCCM programme. Study design Nine field researchers observed 1497 CHW consultations conducted by 90 CHWs, in CHWs’ usual work context (at their own or their patient’s home), using a structured checklist. For the 698 children with suspected pneumonia, as assessed by the CHW, the field researcher filmed the CHW measuring the respiratory rate of the child. Every field researcher visited ten CHWs for 3 days each. On the first day of each 3 day visit, the community was sensitised to encourage them to bring their sick children to the CHW during the next 2 days. Subsequently, the CHW’s consultations were observed for these 2 days. The video recordings were reviewed retrospectively by two pairs of a total of four experts to obtain a reference standard assessment of respiratory rate. Experts were national-level certified master trainers in Integrated Management of Childhood Illnesses (IMCI) in Zambia. Data comparing the expert and CHW assessment of fast breathing, and the proportion of children appropriately treated has been reported separately [12], as the focus of this manuscript is primarily on the rational use of antibiotics in this context, rather than the quality of care received. However, in summary, the data support existing literature that CHWs are capable of measuring respiratory rates and providing appropriate treatment for suspected pneumonia. Following each observation period, the field researchers conducted follow-up home visits with caregivers whose children had been previously prescribed antibiotics by the 90 CHWs involved in the observation component of the study. The antibiotics prescribed were a 5-day course of dispersible amoxicillin in age-specific and colour-coded blister packs. Caregivers were identified from CHW records. A total of 55 caregiver interviews were conducted; 42 interviews (76 %) in Samfya district, and 13 in Kawambwa district. During each home visit, field researchers conducted a pill count of the tablets remaining in the pack, and a structured interview to collect a self-report of caregiver adherence. To further elaborate on the quantitative findings, six focus group discussions (FGD) with CHWs and caregivers, and nine in depth interviews (IDI) with CHWs were conducted. All qualitative study components were audio-recorded and conducted in the local language, Bemba. The data were subsequently translated and transcribed into English, either verbatim for the IDIs, or using the Fairnotes approach [13] for the FGDs. IDIs were conducted at the end of the 2 days of observation with the CHW to be interviewed, and FGDs were conducted once the observation of all 90 CHWs was completed. The IDIs took approximately 40 min to conduct, whilst the FGDs were about 2 h in length. The field researchers were all Zambian nationals from Luapula province with no clinical experience. The field researchers participated in 7 days of interactive training on all the study components, including detailed review of the question guides to ensure shared understanding of the translated material. Participants and sampling The sample size for the observation component of the study was 90 CHWs, calculated to identify the proportion of CHWs who adhere to iCCM guidelines. This assumed 50 % of CHWs adhered to iCCM guidelines, at a precision of 10 %, within a finite population of 440 CHWs that were trained in iCCM in the two study districts and included an adjustment of 10 % to account for CHW non-response and data excluded from final analysis for any reason. The findings presented on correct use of antibiotics are a result of secondary analysis of the data collected. CHWs were selected for inclusion in the observation component of the study by simple random sampling, from the total number of CHWs trained in iCCM in the two study districts. The sample size for the follow-up visits to assess caregiver adherence was calculated to identify the proportion of caregivers who adhere to the antibiotic treatment regimen. The total number of caregivers required was 132, assuming 50 % adherence to antibiotic regimens, at a precision of 10 %, an adjustment for clustering of caregivers by CHW, with a design effect of 1.25. Caregivers of all simple pneumonia cases prescribed antibiotics 6 days prior to the day when data collection was done were eligible for inclusion in this study component. Unfortunately, due to logistical issues relating to the amount of distance to be travelled by each field researcher, it was not possible to interview all who were eligible and reach the required sample size. A total of 55 caregivers were interviewed. Whilst this is a limitation of the data presented, the data remains of value given the limited evidence available on antibiotic adherence by patients in low-income country contexts. For the IDIs, nine CHWs were systematically sampled, with the eighth CHW observed by each of the nine field researchers invited to participate at the end of the second day of observation. For the six FGDs, three of which were conducted with CHWs, and three with caregivers, 24 CHWs and 24 caregivers were selected by convenience sampling. All CHWs and caregivers who had participated in the observations and day six follow-up interviews, respectively, were eligible to participate, as well as other caregivers whose children had recently received treatment for suspected pneumonia from the CHW. Analysis Quantitative data were analysed using Stata version 12 (StataCorp LP). Of the 1497 CHW consultations observed, seven were excluded due to an absence of data being recorded (n = 2), the child being older than 5 years of age (n = 1) or the age of the child not specified (n = 4). Subsequently, of the 698 videos recorded of the respiratory rate count conducted by the CHW during these consultations, 161 videos were excluded due to poor video quality, missing information or if the child was breastfeeding or crying or where the expert ratings of respiratory rate differed by more than five counts and were not recounted. Therefore, data from 1490 consultations and 537 videos were analysed to assess CHW adherence to guidelines and antibiotic use. Videos were retrospectively reviewed by the experts, assessing CHWs’ counts against a reference standard. Descriptive results are summarised as frequencies, percentages or as a median with inter-quartile range [IQR]. A number of indicators were calculated to assess the percentage of children with fast and normal breathing, as classified by the CHW, who were treated in line with the iCCM guidelines. Indicators were also developed to determine whether antibiotics were given for the correct purpose i.e. fast breathing pneumonia as assessed by an expert, and at the correct dosage. Each indicator is presented as a percentage with 95 % confidence intervals adjusting for clustering of children by CHW. Pearson chi-squared test accounting for clustering was used to assess association between each indicator and various factors (e.g. younger and older children). Data from all 55 caregiver interviews was analysed to explore caregiver adherence to antibiotic treatment. Adherence was measured using the pill count of remaining antibiotics and a self-report on the frequency of giving antibiotics to the child. From these, a combined indicator of adherence was created. Full adherence was defined as a caregiver who reported giving the child antibiotics three times a day, at the correct times (morning, noon and night) for 5 days and who had no tablets remaining in the pack. In all other scenarios, caregivers were classified as non-adherent. Results are presented as summary statistics (means or percentages). The percentage adherent is presented with confidence intervals, accounting for the cluster effect. Qualitative data was managed using NVivo 10 (QSR International) and analysed by thematic analysis. The coding frame was developed by three researchers (RK, KG, CS) based on analysis and discussion of three CHW IDI transcripts. It consisted of both a priori and grounded codes. The remaining IDI and FGD transcripts were analysed by KG, with further discussion as needed regarding any changes to the coding frame. Summaries of the qualitative data were reviewed by CS and RK. Results and discussion Participants As presented in Table 1, approximately 80 % of the 90 CHWs who were observed were male, 77 % had a secondary school education, and 90 % originated from the community in which they worked. The CHWs had a wide range of experience, with 23 % having worked for longer than 10 years, 21 % between 5 to 10 years and 53 % less than 5 years. The median number of children observed per CHW was 14 [IQR: 10–21] during the two days observation period. Half the children were female and 25 % were aged 2–11 months. The characteristics of the children whose respiratory rate was recorded on video were similar to those whose respiratory rate was not counted.Table 1 CHW characteristics CHW N = 90  Median Age, [IQR] 44 [38,49]   Sex, n(%)     Female 17 (19)   Highest level of education, n(%)     Primary 19 (21)     Secondary 69 (77)     College 1 (1)     Not specified 1 (1)  Originate from community where they work, n(%) 81 (90)   Years working as CHW, n(%)      < 1 1 (1)     1 to 2 22 (24)     2 to 5 26 (29)     5 to 10 19 (21)      > 10 21 (23)     Not specified 1 (1) Of the 55 caregivers interviewed, 93 % were mothers and the remaining 7 % were close family members (Table 2). Almost half of the caregivers were above 30 years of age and the majority (82 %) had a primary school education. The age of the sick child was recorded by field researchers in 42 interviews and the age ranged from 3 months to 4 years, with 38 % in the younger age category (2–11 months).Table 2 Caregiver characteristics Caregiver characteristics Mothers N = 51 Others N = 4 Total N = 55 Age of caregiver (years)   < 30 29 (57 %) 0 (0 %) 29 (53 %)   > = 30 22 (43 %) 4 (100 %) 26 (47 %) Highest Education  None 3 (6 %) 1 (25 %) 4 (7 %)  Primary 42 (82 %) 3 (75 %) 45 (82 %)  Secondary or above 6 (12 %) 0 (0 %) 6 (11 %) Age of child  2–11 months 16 (31 %) 0 (0 %) 16 (29 %)  12–59 months 24 (47 %) 2 (50 %) 26 (47 %)  Not specified 11 (22 %) 2 (50 %) 13 (24 %) The majority of CHWs who participated in the IDIs were male. The CHW FGDs involved 17 males and seven females, with eight participants per group. All 24 participants in the caregiver FGDs were female, and brought children with them. Rational use of antibiotics by CHWs Results are reported for the 537 observations with a video recording, where it was possible for experts to verify the respiratory rate count determined by the CHW. Adherence to treatment guidelines Of all the children whose respiratory rate was counted by the CHW, 92 % [95 % CI: 87, 95] were treated according to the treatment guidelines, receiving antibiotics for fast breathing and no antibiotics for a normal respiratory rate (Table 3). There was no significant difference in CHW adherence to the guidelines for prescribing antibiotics between the two age groups (p = 0.31).Table 3 Medication prescribed by CHW according to assessment of fast breathing by CHW and child’s age All children N = 537 Children aged 2–11 months N = 147 Children aged 12–59 months N = 390 Normal breathing (%) Fast breathing (%) Normal breathing (%) Fast breathing (%) Normal breathing (%) Fast breathing (%) N = 324 N = 213 N = 68 N = 79 N = 256 N = 134 Received Antibiotics, n  No 308 (95) 27 (13) 65 (89) 6 (8) 243 (95) 21 (16)  Yes 16 (5) 186 (87) 3 (4) 73 (92) 13 (5) 113 (84) Colour of Pack Given, n  Pink (for children aged 2–11 months) 7 (44) 78 (41) 2 (67) 66 (90) 5 (38) 12 (11)  Green (for children aged 12–59 months) 8 (50) 99 (53) 0 (0) 3 (4) 8 (62) 96 (85)  NSa 1 (6) 9 (5) 1 (33) 4 (6) 0 (0) 5 (4)  % received antibiotics [95 % CI] 5 [2,10] 87 [78,93] 4 [1,14] 92 [80,97] 5 [2,12] 84 [75,91]   p-value* <0.01 <0.01 <0.01 *P-value comparing proportion receiving antibiotics by fast breathing status (CHW), for all children and by age group a NS not specified CHWs’ ability to prescribe antibiotics according to the guidelines and corresponding to their assessment suggests that CHWs are capable of treating pneumonia at community level. Overall, as shown in Table 3, 87 % [95 % CI: 78, 93] of children with fast breathing, as assessed by the CHW, were prescribed antibiotics. The difference between the number receiving antibiotics for fast breathing in the two age groups was not significant (p = 0.06). These data are in line with studies by Mukanga et al. [14] and Kalyango et al. [15], with the latter stating that 82 % of children with fast breathing had received an antibiotic. To note, 6/27 of the children with fast breathing who were not given amoxicillin were referred. Despite these children not being prescribed antibiotics, they may still have been appropriately managed by the CHW, for example if the CHW lacked the appropriate medication. The proportion receiving antibiotics was significantly lower among those with normal breathing (p < 0.01), in the two age groups, when analysed separately, and for all children (Table 3). When the data was further analysed to take into account the dosage of antibiotics given, 79 % [95 % CI: 70, 87] of the children with fast breathing as assessed by the CHW were prescribed antibiotics and at the correct dose for the child’s age. A significantly higher proportion of children in the younger age group received antibiotics for fast breathing and at the correct dose compared to older age group (p = 0.02). Of all the children prescribed antibiotics by CHWs, regardless of whether or not the child had fast breathing, 10 % received antibiotics that were at the incorrect dose for the child’s age (Table 3). 14 % of the older children received the pink pack of antibiotics intended for the younger children, whilst 4 % of the younger children received the green pack of antibiotics intended for the older children. The pack given was not specified for ten children, five in the younger and five in the older age group. The median age of the children in the older group, who received the dosage of antibiotics intended for the younger group, was 24 months [IQR 14 to 48]. As drug stock levels were not assessed as part of this study, it is not possible to determine if those children receiving an incorrect dose for their age group were as a result of stock-outs of the correct age-specific blister pack, or that the importance of giving the correct dose for age was not understood by some CHWs. Unfortunately, this was not specifically raised during the CHW FGDs, and it is acknowledged that this is a limitation of the data collected. However, overall, the findings suggest that CHWs are able to prescribe the appropriate dose of antibiotics according to the child’s age. Interestingly, the study showed that CHWs referred to the flip chart job aid in only 25 % of consultations, and used the sick child reporting form in only 33 %. Although the low utilisation of these tools may be due to the presence of the observer [16], these data could suggest that the CHWs are confident in their ability to recall the diagnostic algorithm and danger signs, given the high level of adherence to the guidelines. In discussing and analysing data relating to the ability of CHWs, it must be acknowledged that health facility workers with more training may demonstrate a lower level of adherence to guidelines than CHWs for the care of children with cough. For example, one study from Ghana that evaluated the clinical assessments of children presenting to primary health facilities found that the respiratory rate was measured in only 4 % of children presenting with cough, instead of all children as recommended [17]. Correct use of antibiotics by CHWs When the data was analysed with a focus on whether the antibiotics prescribed by CHWs were used for the correct purpose, it was found that of the antibiotics given to 202 children in this study, 35 % [95 % CI: 26, 45] were given to children who did not have fast breathing, as assessed by an expert and therefore antibiotics should not have been prescribed. There was a significant difference between the two age groups with regard to the rational use of antibiotics, with antibiotics given to children who were aged 12–59 months more likely to be incorrectly prescribed (45 % [95 % CI: 34, 57] vs 18 % [95 % CI: 11, 30], p = <0.01), although it is not clear why this would be the case. The apparent discrepancy between the high number of CHWs able to prescribe antibiotics according to the guidelines and corresponding to their assessment (92 %); and the incorrect use of antibiotics by CHWs (35 %), is in part due to CHWs prescribing antibiotics for children that did not have fast breathing according to their own assessment, but primarily due to their incorrect measurement of the respiratory rate. According to the studies by Cardemil et al. [18] and Miller et al. [19], 24 % and 6 % of antibiotics, respectively, were used irrationally i.e. for the wrong purpose. In terms of irrational use of antibiotics, where antibiotics were prescribed for the correct purpose i.e. fast breathing pneumonia as assessed by an expert, but not at the correct dosage, 5 % [95 % CI: 2, 11] of all antibiotics given were at the incorrect dose for the child’s age. The difference between the two age groups was not significant (p = 0.51). Dosage generally does not appear to have been separately assessed in other research studies, or perhaps is not reported in the literature. Given that 92 % of children, based on the CHW assessment, were treated as per the guidelines, it is likely that the percentage of antibiotics used irrationally could be substantially reduced with improved tools for the measurement of respiratory rate and an emphasis in training and supervision on the importance of prescribing the correct dose for age. Although it has been recently debated following an article by Druetz et al. [10], the data from this study support the opinion that CHWs are capable of prescribing appropriate treatment in line with guidelines. Qualitative findings regarding CHW adherence to guidelines and correct use of antibiotics The level of knowledge relating to how pneumonia should be treated was fairly good, although the FGDs indicated a higher level of knowledge, than the interviews. This may be simply because the minority of CHWs who contributed in the discussion groups were more vocal because they knew more, or because the group environment helped to stimulate a more detailed discussion. All of the FGDs correctly mentioned that the medicine for pneumonia is given according to age, with a pink pack for children aged 2–11 months, with one tablet given three times a day (morning, afternoon and evening). For children aged 12–59 months, a green pack is given, with two tablets to be administered three times a day. All groups mentioned that the course of treatment was for 5 days for both age groups. Two of the groups specifically mentioned the name of drug, amoxicillin. One participant in one of the FGDs also mentioned that the CHWs tell “caregivers not to share medicine” (CHW FGD 2). In terms of the interviews with CHWs, the responses were briefer and less focused. Four of the seven participants who described how they would treat a child with pneumonia mentioned amoxicillin, whilst the others referred simply to medicine given. Four of the participants mentioned that the medicine was for 5 days and two made reference to completing the course of treatment given. Two of the interviews also highlighted the importance of explaining to the caregiver that the child has pneumonia and/or how to give the medicine. Only one participant mentioned the dosage of tablets to be given, although did so incorrectly.“…child had 1 year 1 month. I gave her amoxicillin 250 a pack of 15 for 5 days”. (CHW IDI 2) The child was 13 months old, and therefore should have received a pack of 30 tablets for 5 days, allowing for administration of two tablets per dose, three times a day. None of the CHWs in the IDI mentioned the different packs of tablets to be given depending on the age of the child. In response to the question, ‘What do you think are the things that matter most to your clients regarding the kind of care you provide to them for pneumonia or cough?’, two interview participants specifically referred to the necessity of finishing the course of antibiotics. This was also highlighted by another three participants during the interviews.“I just encourage the caregivers to give medicine to the patient and finish all the medicine they have been given.” (CHW IDI 8) Qualitative data regarding the influences on the prescription and use of antibiotics by CHW It appears that whilst caregivers are involved in the consultation process, CHWs do not feel influenced when treating children with pneumonia and prescribing medicine, even though some caregivers are perceived to apply pressure to receive medicine.“When they come, they do not influence us. It’s us to see using our knowledge. When we diagnose, using the knowledge we were taught, then from there we able to know the problem.” (CHW IDI 7) “I should not base the treatment on the parent’s instructions like our experience today when one parent wanted me to give malaria treatment before I could even attend to the child. We have been trained never to give medicine before checking the child”. (CHW IDI 2) Whilst CHWs may be unlikely to admit that they are influenced, the data support the quantitative findings. Of those children with normal breathing, as assessed by the CHW, 95 % were not given antibiotics. The qualitative data, however, does indicate that a large proportion of CHWs feel that caregivers would expect to receive medicine when they attend the CHW and often would pressure the CHW to prescribe medicine or antibiotics. Whilst there are studies exploring antibiotic use and provider prescribing behaviour, there are very few from low and middle income countries, and even less that are focused on pneumonia or lower respiratory tract infections, and there appears to be none exploring prescribing practices by CHWs. Studies in Peru and Thailand conducted at hospitals and health facilities, relating to treatment for diarrhoea, note that patient expectations can impact antibiotic prescribing behaviour [20, 21]. However, in exploring factors related to the overuse of antimalarials at hospital level in Tanzania, Chandler et al. [22] found that patient satisfaction did not appear to be dependent on prescription of antimalarials. Patients did not explicitly request antimalarials during observed consultations, but there was an expectation that they would be tested and subsequently receive appropriate treatment. In relation to patent medicine vendors in Nigeria, Akuse et al. [23] document a general expectation or preference by caregivers to receive antibiotics as treatment. Chandler et al. [22] also point out that it is likely that social pressures on provider prescribing behaviour may play a greater role in lower level health facilities, which could therefore be further increased for CHWs at community level. In our study, there was a clear desire to receive treatment of some kind, although the data suggest that the emphasis is on treatment that will make the child better, rather than on receiving antibiotics specifically. Furthermore, CHWs did not generally seem to be influenced in our study, despite the perceived social pressure to prescribe medication. Whilst the care given by the CHW may not be influenced by the caregiver, it appears that this does impact the morale of both CHWs and caregivers. Factors impacting CHWs’ morale should be acknowledged given that their role in Zambia at the time of the study, and other countries, is often voluntary. The FGD that took place with CHWs in Kawambwa appeared to show a greater appreciation for the CHW’s training and knowledge, resulting in less pressure to prescribe medicine. The behaviour change component of the iCCM programme had just begun to be implemented in this district, however it is not possible to definitively attribute this difference in caregiver behaviour to the same communities where the CHWs involved in our study were responsible for providing care. In our study, the caregivers generally seemed to be satisfied with the care provided by the CHW for children with pneumonia, bearing in mind their apparent discontent if medication was not prescribed. Encouragingly, the reasons for satisfaction expressed by the caregivers were matched by the CHWs’ perceptions of their satisfaction, indicating a good understanding of the community in which the CHWs live and work. Aside from receiving medication and its availability, reasons for satisfaction included availability and accessibility of CHWs, as well as the importance of CHWs’ attitude towards caregivers and quality of care received, similar to reasons noted in a study by Hanson et al. [24] relating to preferences for hospital care in northern Zambia for acute pneumonia in children. Of note, one participant highlighted that the amoxicillin they receive from the CHW is different to that given by the clinic, which means it is easier for children to swallow, indicating a different formulation. As noted in the background, as part of the iCCM programme in Luapula province, the Pharmaceutical Regulatory Authority in Zambia granted a waiver for the use of dispersible amoxicillin, although it had not been approved for routine use in Zambia. Rational use of antibiotics by caregivers In general, the caregiver who was responsible for administering the drug, was also the same individual that had brought the sick child to the CHW and had received information from the CHW on how to give the antibiotics. All caregivers interviewed stated that they understood the instructions as to how to give the antibiotics, regardless as to how they received them. Although it is not possible to determine exactly what instructions were given to these caregivers from the CHW or another individual, data from the observation of consultations (1490 children) shows that 95 % [95 % CI: 90, 97] of caregivers given amoxicillin received some instructions from the CHW, with 74 % [95 % CI: 62, 83] of those receiving instructions receiving all five key points; number of doses to be given per day, timing, number of tablets per dose, length of treatment course and that all tablets should be taken. Table 4 displays the results on self-reported administration of antibiotics and pill count. Of the 42 interviews where the age of the child is known, the number of tablets given per dose was generally in line with the iCCM treatment guidelines. Children aged 2–11 months received one tablet per dose and those aged 12–59 months received two tablets as per the guidelines, except for five children who received only one tablet. The median age of these children was 32 months. It is not possible to determine whether this was non-adherence where they only gave one of the two tablets that should have been given, or if they were given the wrong blister pack for their age by the CHW and therefore could only give one tablet per dose. Two children received two or more tablets per dose, but in this instance the child’s age was not specified.Table 4 Self-reported administration of antibiotics by caregivers and pill count Age of child 2–11 months 12–59 months Not specified Total N = 16 N = 25 N = 13 N = 54a Number of tablets given per dose:  1 (for children aged 2–11 months) 16 (100 %) 5 (20 %) 3 (23 %) 24 (44 %)  2 (for children aged 12–59 months) 0 20 (80 %) 8 (62 %) 28 (52 %)   > 2 0 0 2 (15 %) 2 (4 %)  Gave amoxicillin 3 times a day 15 (94 %) 23 (92 %) 13 (100 %) 51 (94 %)  Gave amoxicillin morning, noon and night 15 (94 %) 23 (92 %) 12 (92 %) 50 (93 %) Number of days amoxicillin given:  2 1 (6 %) 1 (4 %) 0 (0 %) 2 (4 %)  3 16 (%) 6 (24 %) 1 (8 %) 8 (15 %)  4 2 (12 %) 2 (8 %) 1 (8 %) 5 (9 %)  5 11 (69 %) 12 (48 %) 8 (62 %) 31 (57 %)   > 5 1 (6 %) 4 (16 %) 3 (23 %) 8 (15 %) Number of tablets remaining in the pack: N = 15 N = 23 N = 12 N = 50b  0 9 (60 %) 15 (65 %) 9 (75 %) 33 (66 %)  1–6 5 (33 %) 4 (17 %) 2 (17 %) 11 (22 %)  >6 1 (7 %) 4 (17 %) 1 (8 %) 6 (12 %) aOnly 54/55 caregivers completed information on timing and number of tablets given b50 caregivers able to present packet of pills In terms of the timing of doses given, 93 % of caregivers reported that they gave amoxicillin three times a day and at morning, noon and night. However, only 57 % reported that they gave treatment for five consecutive days. 81 % gave treatment for 3 to 5 days, with eight caregivers giving for more than 5 days. As shown in Table 5, combining the number of days, number of times a day and time when treatment was given into one adherence indicator based on caregiver recall alone indicated that only 54 % of caregivers gave amoxicillin three times a day, at morning, noon and night for 5 days, as required. Adherence increased to 76 %, if those who gave treatment for 3 to 5 days were included (i.e. gave amoxicillin three times a day, at morning, noon and night for 3 to 5 days). A recent Cochrane review [25] found that antibiotics administered for 3 days was as effective as a 5 day course for treatment of non-severe pneumonia in children under five. Our data suggest that a shorter course would potentially improve adherence, considering that caregivers were more likely to be adherent for 3 days than for 5 days. A shorter course of antibiotics could thereby reduce the risk of antibiotic resistance and substantially lower the costs of treatment. However, it must be noted that all the studies in the Cochrane review were from Pakistan and further evidence is needed to support a shorter course of antibiotics for pneumonia in Africa, especially in settings where HIV prevalence is high.Table 5 Indicators of treatment adherence by caregivers n N % Adherent [95 % CI] Self-report indicator:  Gave amoxicillin, 3 times a day at morning, noon and night for 5 days. 29 54 54 [36 to 64]  Gave amoxicillin, 3 times a day at morning, noon and night for 3, 4 or 5 days 41 54 76 [63 to 85] Pill-count indicator:  O tablets 33 50 66 [50 to 79] Combined indicator:  0 tablets + gave amoxicillin, 3 times a day at morning, noon and night for 5 days. 23 50 46 [30 to 63] Looking at pill count alone, of the 50 caregivers who were able to present the packet of pills, 66 % had no tablets remaining. The primary reason that treatment was discontinued, mentioned by nine individuals, was forgetfulness. The second most common explanation was that the patient was better or cured. In the study by Kalyango et al. [26], the same factors were highlighted, with forgetfulness mentioned by 38 % of caregivers and caregivers’ perception of improvement or recovery in the child mentioned by 14 %. It could be concerning that one caregiver specified that the CHW had advised them to stop when the child recovers. However, in the interviews with CHWs, five of the nine participants highlighted the importance of finishing the course of antibiotics and encouraging the caregiver to do so. A combined indicator for adherence, including both caregiver recall and pill count, indicates that 46 % were fully adherent. Full adherence is defined by giving amoxicillin three times a day, at morning, noon and night for 5 days, based on caregiver recall, and no tablets remaining, as determined by pill count. It is difficult to find data on comparable indicators from studies assessing adherence in higher income countries, however, Kardas and Ratajczyk-Pakalska [27] found that only 23.5 % of patients (aged 18 and over) fully adhered to a 5-day, three times a day, antibiotic treatment regimen for respiratory tract infections in Poland. Adherence was also assessed by pill-count and questionnaire. High quality pneumonia case management by CHWs is essential for successful treatment outcomes and to minimise any contribution to the development of antibiotic resistance. However, adherence of caregivers to the treatment given is also pivotal for both and to ensure iCCM has the maximum impact in reducing childhood mortality from pneumonia. Whilst there have been a number of studies on the ability of CHWs to diagnose and treat pneumonia, adherence to courses of antibiotics by caregivers or patients has rarely been explored in a low-income country context (Karin Källander, unpublished observations). However, studies have explored adherence to antimalarials at community level, especially following the change in first line treatment of malaria from chloroquine or sulfadoxine-pyrimethamine to artemisinin combination therapies (ACTs), which required more complex dosing regimens than previous treatments [28–31]. Adherence to a six dose treatment regimen for artemether-lumefrantrine (3 days, twice daily) has been shown to be between 81 and 97 % in some African countries [28, 30, 31], but closer to 40 % in Ethiopia and South Sudan [29]. ACTs are routinely used as part of iCCM in multiple countries. Adherence has been found to be inversely proportional to dose frequency [32, 33]. However, although children may be treated for more than one disease through iCCM with different dosing regimens, Kalyango et al. [26] found that there was no difference between treatment adherence to antimalarials alone compared with adherence to antimalarials and antibiotics, both 3 day courses of treatment. Qualitative data relating to the factors influencing antibiotic use by caregivers There are multiple factors that may affect the adherence of caregivers to treatment prescribed by a CHW, including knowledge and understanding related to the specific disease, trust and relationship with the CHW, expectation of care and culture or traditional beliefs. The qualitative data from the focus group discussions with caregivers indicate, perhaps unsurprisingly, that when asked, caregivers all agree that they would follow the instructions given, either in order for their child to be healthy or because they understood the CHW to be appropriately trained.“The CHWs are well trained and can only give professional advice” (Caregiver FGD 1). It was implied by one that other caregivers in the community would not finish the course of antibiotics. Also, the caregivers all agreed that they would not share treatment with others. In terms of caregivers’ knowledge and understanding of pneumonia, one group from one district did seem to have a generally higher level of knowledge, regarding prevention, treatment and prevalence of pneumonia, which may be due to an additional behaviour change communication intervention that had begun to be implemented in certain districts of Luapula province at the time of study. Notably, this group felt that the prevalence of pneumonia had reduced since CHWs had started working in the area, although their perceptions on the prevalence of pneumonia are only likely to reflect their understanding of the disease itself. Whilst knowledge on the causes of pneumonia appeared to be lacking across all the groups, there was an understanding in this group that pneumonia could be prevented, with some suggestions given as to how to prevent pneumonia. In terms of diagnosing pneumonia, the data perhaps imply a level of trust in health workers’ abilities, with caregivers commenting that the health workers would advise as to whether their child needed treatment and that the clinic would not give out medicine without testing the child. It is not clear whether the group was referring solely to health workers based in health facilities specifically, or CHWs. Regarding treatment, the majority of caregivers emphasised that seeking medical help or intervention was the most effective way to treat pneumonia, referring to attendance at the clinic or the hospital. In terms of type of medicine given for pneumonia, there was a large range of comments, but again the group who generally had a higher level of knowledge were more aware of the treatment given, with one individual having a very good understanding, including the number of tablets to be given for children of different ages. However, there were generally mixed responses given on these different topics in the focus group discussions. Improving caregivers’ understanding of pneumonia and its treatment may improve adherence to treatment and could also impact caregivers’ interactions with CHWs, resulting in greater trust in, recognition of, and increased morale for CHWs. In addition, such BCC interventions may also reduce the likelihood that caregivers will seek medication from elsewhere [34]. Conclusions This research provides valuable data on the rational use of antibiotics at community level, by both CHWs and caregivers, through iCCM. From the literature, it appears no other studies have been conducted exploring both the assessment of respiratory rate measurement by CHWs (data presented in a separate manuscript [12]) with evaluation of CHW antibiotic prescribing behaviour, alongside assessment of caregivers’ adherence to treatment in the same setting for iCCM. The timely identification of children with fast breathing and chest indrawing could prevent the majority of deaths from acute respiratory infections, if appropriate services are available [35]. iCCM improves access to health services for children under five in remote and rural locations, for diseases that have the highest mortality in this age group. Although there is an ongoing debate, this study further supports literature that CHWs providing iCCM are capable of prescribing treatment based on their assessment, in line with guidelines. However, greater accuracy of respiratory rate assessment through improved diagnostic tools would further facilitate the rational use of antibiotics, as well as ensuring the effectiveness of treatment. It is therefore recommended that there is further investment in the development of more user-friendly and refined diagnostic devices for assessing symptoms of pneumonia for use in remote settings in low and middle income countries. Caregiver adherence to the full 5-day course of antibiotics in this study is less than optimal, despite the use of age-specific blister packs, although evidence suggests that complete adherence could be far higher with a shorter course of treatment [26], also resulting in more rational use of antibiotics. Assessment of effectiveness of a 3-day, compared to a 5-day course of antibiotics for treatment of pneumonia in high HIV settings, such as Africa, would therefore inform decision making on iCCM, helping to maximise adherence and minimise the development of antibiotic resistance. Furthermore, communities’ perspectives, understanding of a given health issue and their needs should be given sufficient priority and investment in the implementation of health services to ensure interventions can be effectively delivered, for example, through information, education and communication activities. Abbreviations CHWCommunity health worker CIConfidence interval CIDACanadian International Development Agency FGDFocus group discussion iCCMIntegrated community case management IDIIn depth interview IMCIIntegrated management of childhood illness IQRInterquartile range Acknowledgements The authors would like to acknowledge the hard work of the community health workers in Luapula Province, Zambia, as well as participation of the caregivers and children in the study. We also appreciate the support of Dr Penelope Kalesha, from the Department of Mother and Child Heath at the Ministry of Health in Zambia, in helping to identify appropriate experts to retrospectively assess respiratory rate. Thanks also to Dr Sylvia Meek for providing valuable oversight. Funding This work was conducted through COMDIS-HSD, a Research Programme Consortium funded by the UK government. However the views expressed do not necessarily reflect the UK Government’s official policies. This research study was conducted alongside implementation of iCCM in Luapula Province in Zambia, funded by CIDA (now Global Affairs Canada) from 2009 to 2012. Some of the activities and materials for the study were therefore supported by funding from CIDA. The funding bodies had no role in the study design, data collection, analysis, interpretation of data, in the preparation of the manuscript or the decision to submit. Availability of data and materials Consent from participants was only granted to use the data for future research purposes where ethical approval has been granted. For qualifying researchers, access to the data will be guaranteed by contacting Dr Karin Källander, Malaria Consortium Senior Research Advisor, at k.kallander@malariaconsortium.org. Authors’ contributions CS, PH, MM and JT contributed towards the conception and design of the study; CS, KG and SM were responsible for the acquisition of data, KG, CS, SN, RK analysed and interpreted the data, with significant input from HC, KK and PH. KG drafted the manuscript and all authors were involved in revising it critically for intellectual content, and have given final approval of the version to be published. Competing interests The authors declare that they have no competing interests. Consent for publication Not applicable. Ethics approval and consent to participate Ethics approval was received from the School of Medicine Research Ethics Committee at the University of Leeds and the University of Zambia Biomedical Research Ethics Committee. Informed consent was obtained from each CHW prior to the observation of any consultations, and from caregivers of children under five presenting to the CHW during the 2 days. Consent was also obtained from the caregivers who were selected for the follow-up visit to assess adherence, prior to the pill count or questionnaire. Further consent was obtained from the CHWs who were subsequently selected for the IDIs, and the CHWs and caregivers who participated in the FGDs. ==== Refs References 1. World Health Organization Antimicrobial resistance: global report on surveillance 2014 Geneva World Health Organization 2. World Health Organization Global action plan on antimicrobial resistance 2015 Geneva World Health Organization 3. Leung E Weil DE Raviglione M Nakatani H World Health Organization World Health Day Antimicrobial Resistance Technical Working Group The WHO policy package to combat antimicrobial resistance Bull World Health Organ 2011 89 5 390 2 10.2471/BLT.11.088435 21556308 4. World Health Organization WHO global strategy for containment of antimicrobial resistance 2001 Geneva World Health Organization 5. Hart CA Kariuki S Antimicrobial resistance in developing countries Br Med J 1998 317 7159 647 50 10.1136/bmj.317.7159.647 9727995 6. Weerasuriya KSJ O’Brien TF Containing antimicrobial resistance: a renewed effort Bull World Health Organ 2010 88 12 878 10.2471/BLT.10.084236 21124706 7. Walker CLF Rudan I Liu L Nair H Theodoratou E Bhutta ZA O’Brien KL Campbell H Black RE Global burden of childhood pneumonia and diarrhoea Lancet 2013 381 9875 1405 16 10.1016/S0140-6736(13)60222-6 23582727 8. Liu L Oza S Hogan D Perin J Rudan I Lawn JE Cousens S Mathers C Black RE Global, regional, and national causes of child mortality in 2000–13, with projections to inform post-2015 priorities: an updated systematic analysis Lancet 2014 385 9996 430 40 25280870 9. Laxminarayan R Duse A Wattal C Zaidi AK Wertheim HF Sumpradit N Vlieghe E Hara GL Gould IM Goossens H Antibiotic resistance-the need for global solutions Lancet Infect Dis 2013 13 12 1057 98 10.1016/S1473-3099(13)70318-9 24252483 10. Druetz T, Siekmans K, Goossens S, Ridde V, Haddad S. The community case management of pneumonia in Africa: a review of the evidence. Health Policy Plan. 2013. doi: 10.1093/heapol/czt104. 11. Republic of Zambia Central Statistical Office Zambia 2010 census of population and housing: preliminary population figures 2011 Lusaka Republic of Zambia Central Statistical Office 12. Sinyangwe C Graham K Nicholas S King R Mukupa S Kallander K Counihan H Montague M Tibenderana J Hamade P Assessing the quality of care for pneumonia in integrated community case management: a cross-sectional mixed methods study PLoS One 2016 11 3 e0152204 10.1371/journal.pone.0152204 27011331 13. Halcomb EJ Davidson PM Is verbatim transcription of interview data always necessary? Appl Nurs Res 2006 19 1 38 42 10.1016/j.apnr.2005.06.001 16455440 14. Mukanga D Babirye R Peterson S Pariyo GW Ojiambo G Tibenderana JK Nsubuga P Kallander K Can lay community health workers be trained to use diagnostics to distinguish and treat malaria and pneumonia in children? Lessons from rural Uganda Tropical Med Int Health 2011 16 10 1234 42 10.1111/j.1365-3156.2011.02831.x 15. Kalyango JN Rutebemberwa E Alfven T Ssali S Peterson S Karamagi C Performance of community health workers under integrated community case management of childhood illnesses in eastern Uganda Malar J 2012 11 282 10.1186/1475-2875-11-282 22905758 16. Counihan H Harvey SA Sekeseke-Chinyama M Hamainza B Banda R Malambo T Masaninga F Bell D Community health workers use malaria rapid diagnostic tests (RDTs) safely and accurately: results of a longitudinal study in Zambia AmJTrop Med Hyg 2012 87 1 57 63 10.4269/ajtmh.2012.11-0800 17. Baiden F Owusu-Agyei S Bawah J Bruce J Tivura M Delmini R Gyaase S Amenga-Etego S Chandramohan D Webster J An evaluation of the clinical assessments of under-five febrile children presenting to primary health facilities in rural Ghana PLoS One 2011 6 12 e28944 10.1371/journal.pone.0028944 22174932 18. Cardemil CV Gilroy KE Callaghan-Koru JA Nsona H Bryce J Comparison of methods for assessing quality of care for community case management of sick children: an application with community health workers in Malawi AmJTrop Med Hyg 2012 87 5 Suppl 127 36 10.4269/ajtmh.2012.12-0389 19. Miller NP Amouzou A Tafesse M Hazel E Legesse H Degefie T Victora CG Black RE Bryce J Integrated Community Case Management of Childhood Illness in Ethiopia: Implementation Strength and Quality of Care Am J Trop Med Hyg 2014 91 2 424 34 10.4269/ajtmh.13-0751 24799369 20. Paredes P de la Peña M Flores-Guerra E Diaz J Trostle J Factors influencing physicians’ prescribing behaviour in the treatment of childhood diarrhoea: knowledge may not be the clue Soc Sci Med 1996 42 8 1141 53 10.1016/0277-9536(95)00387-8 8737432 21. Howteerakul N Higginbotham N Freeman S Dibley MJ ORS is never enough: physician rationales for altering standard treatment guidelines when managing childhood diarrhoea in Thailand Soc Sci Med 2003 57 6 1031 44 10.1016/S0277-9536(02)00478-1 12878103 22. Chandler CIR Mwangi R Mbakilwa H Olomi R Whitty CJM Reyburn H Malaria overdiagnosis: is patient pressure the problem? Health Policy Plan 2008 23 3 170 8 10.1093/heapol/czm046 18321889 23. Akuse R, Eseigbe E, Ahmed A, Brieger W. Combating Antibiotic Resistance In Tropical Countries‚ Don’t Ignore Patent Medicine Vendors. Internet J Prev Med. 2012;1(2):1–11. 24. Hanson K McPake B Nakamba P Archard L Preferences for hospital quality in Zambia: results from a discrete choice experiment Health Econ 2005 14 7 687 701 10.1002/hec.959 15619273 25. Haider BA, Lassi ZS, Bhutta ZA. Short-course versus long-course antibiotic therapy for non-severe community-acquired pneumonia in children aged 2 months to 59 months. The Cochrane Database Syst Rev. 2008(2). doi: 10.1002/14651858.CD005976.pub2 26. Kalyango JN Rutebemberwa E Karamagi C Mworozi E Ssali S Alfven T Peterson S High adherence to antimalarials and antibiotics under integrated community case management of illness in children less than 5 years in eastern Uganda PLoS One 2013 8 3 e60481 10.1371/journal.pone.0060481 23555980 27. Kardas P Ratajczyk-Pakalska E Patient adherence in respiratory tract infections: ceftibuten versus other antibiotics (PARTICULAR study) Pol Merkur Lekarski 2001 10 60 445 9 11503260 28. Chinbuah AM Gyapong JO Pagnoni F Wellington EK Gyapong M Feasibility and acceptability of the use of artemether-lumefantrine in the home management of uncomplicated malaria in children 6–59 months old in Ghana Trop Med Int Health 2006 11 7 1003 16 10.1111/j.1365-3156.2006.01654.x 16827701 29. Lemma H Lofgren C San Sebastian M Adherence to a six-dose regimen of artemether-lumefantrine among uncomplicated Plasmodium falciparum patients in the Tigray Region, Ethiopia Malar J 2011 10 349 10.1186/1475-2875-10-349 22141751 30. Fogg C Bajunirwe F Piola P Biraro S Checchi F Kiguli J Namiiro P Musabe J Kyomugisha A Guthmann J-P Adherence to a six-dose regimen of artemether-lumefantrine for treatment of uncomplicated plasmodium falciparum malaria in Uganda Am J Trop Med Hyg 2004 71 5 525 30 15569777 31. Ajayi IO Browne EN Bateganya F Yar D Happi C Falade CO Gbotosho GO Yusuf B Boateng S Mugittu K Effectiveness of artemisinin-based combination therapy used in the context of home management of malaria: a report from three study sites in sub-Saharan Africa Malar J 2008 7 190 10.1186/1475-2875-7-190 18822170 32. Osterberg L Blaschke T Adherence to medication N Engl J Med 2005 353 5 487 97 10.1056/NEJMra050100 16079372 33. Claxton AJ Cramer J Pierce C A systematic review of the associations between dose regimens and medication compliance Clin Ther 2001 23 8 1296 310 10.1016/S0149-2918(01)80109-0 11558866 34. Yu M Zhao G Stålsby Lundborg C Zhu Y Zhao Q Xu B Knowledge, attitudes, and practices of parents in rural China on the use of antibiotics in children: a cross-sectional study BMC Infect Dis 2014 14 1 112 10.1186/1471-2334-14-112 24576064 35. Hadi A Management of acute respiratory infections by community health volunteers: experience of Bangladesh Rural Advancement Committee (BRAC) Bull World Health Organ 2003 81 3 183 9 12764514

PMC005xxxxxx/PMC5002324.txt

==== Front J Neuroeng RehabilJ Neuroeng RehabilJournal of NeuroEngineering and Rehabilitation1743-0003BioMed Central London 18510.1186/s12984-016-0185-yResearchEffects of a balance-based exergaming intervention using the Kinect sensor on posture stability in individuals with Parkinson’s disease: a single-blinded randomized controlled trial Shih Meng-Che solardescendant@hotmail.com 1Wang Ray-Yau rywang@ym.edu.tw 1Cheng Shih-Jung csjneuro@gmail.com 2Yang Yea-Ru 886-2-28267279yryang@ym.edu.tw 11 Department of Physical Therapy and Assistive Technology, National Yang-Ming University, 155, Sec 2, Li Nong St., Beitou, Taipei, Taiwan 2 Department of Neurology, Mackay Memorial Hospital, Taipei, Taiwan 27 8 2016 27 8 2016 2016 13 1 784 2 2016 13 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background The present study examined the effects of a balance-based exergaming intervention using the Kinect sensor on postural stability and balance in people with Parkinson’s disease (PD). Methods We conducted a subject-blinded, randomized controlled study. Twenty people with PD (Hoehn and Yahr stages I through III) were recruited and randomly assigned to either a balance-based exergaming group (N = 10) or a balance training group (N = 10) for an 8-week balance training period. Postural stability was assessed using the limits of stability (LOS) and one-leg stance (OLS) tests. Balance was assessed using the Berg Balance Scale (BBS) and the timed up and go (TUG) test. Participants were assessed pre- and post-training. Results After training, participants in the balance-based exergaming group showed significant improvements in LOS performance, and in the eyes-closed condition of the OLS test. Both training programs led to improvements in BBS and TUG performance. Furthermore, balance-based exergaming training resulted in significantly better performance in directional control in the LOS test (78.9 ± 7.65 %) compared with conventional balance training (70.6 ± 9.37 %). Conclusions Balance-based exergaming training resulted in a greater improvement in postural stability compared with conventional balance training. Our results support the therapeutic use of exergaming aided by the Kinect sensor in people with PD. Trial registration ClinicalTrials.gov.NCT02671396 Electronic supplementary material The online version of this article (doi:10.1186/s12984-016-0185-y) contains supplementary material, which is available to authorized users. Keywords Balance trainingExergamingPostural stabilityParkinson’s diseasehttp://dx.doi.org/10.13039/501100001868National Science CouncilNSC100-2314-B-010-021-MY2Yang Yea-Ru issue-copyright-statement© The Author(s) 2016 ==== Body Background People with idiopathic Parkinson’s disease (PD) commonly exhibit postural instability during daily activities [1]. PD-related balance impairment is associated with a loss of mobility and increased likelihood of falls, and can cause marked disability [2, 3]. To ameliorate postural instability, techniques using external feedback with cueing or sensory stimuli have been investigated [4, 5]. Several studies suggest that external feedback may initiate other neural pathways and play a significant role in the volitional control of movements for people with PD [6, 7]. Virtual reality (VR) technologies such as exergaming may have therapeutic value in the treatment of postural instability [8–10]. VR is a technology that allows the user to interact directly with a computer-simulated environment [11]. Exergames are computer games that are controlled by body movements. VR and exergaming can provide augmented feedback in real time, while a person performs specific motor tasks [12]. Opportunities for repeated accurate performance can be incorporated into VR and exergaming to enhance motor learning [7, 13]. Moreover, VR games can be effective for retaining participants’ interest and motivation. A recent meta-analysis suggested that exergaming may provide an appropriate training approach to improve balance and functional mobility in healthy older people [14]. These findings raise the possibility that exergaming might also provide an approach for improving postural instability for people with PD. A previous study examined the effects a 6-week home-based balance training program using the Wii Fit game for a total of 18 training sessions on balance and functional abilities in people with PD, compared with a group of paired healthy participants [15]. Another study investigated the effects of Wii-based training compared with conventional balance training for 7 weeks (a total of 14 training sessions) on activities of daily living in people with PD [16]. Both studies revealed positive effects of exergaming on balance, functional abilities and activities of daily living among people with PD. However, positive effects were found only within groups, with no between-group differences observed in a comparison with the control group. The absence of between-group differences may have resulted from an inability to capture the full-body motion involved in postural control, or the lack of a sufficiently sensitive sensor to accurately measure motion. The shortcomings of the Wii system’s sensors may limit its potential as an effective intervention [17]. A new exergaming system was recently developed using the Kinect sensor. The Kinect sensor is a low-cost device that can provide measurements for most of the main human joints. Previous studies reported that a kinematic measurement method using the Kinect sensor was accurate and reliable for measuring postural control [18, 19]. These findings suggest that the Kinect sensor could provide a useful tool for therapeutic use. However, there has been little research into the therapeutic use of the Kinect sensor to date. The present study sought to test a therapeutic application of exergaming using the Kinect sensor. We examined the effects of an 8-week balance-based exergaming program developed in our lab, compared with an 8-week period of conventional balance training (16 training sessions), on postural stability and balance in people with PD. We hypothesized that participants who underwent an 8-week balance-based exergaming intervention would demonstrate superior performance on measures of postural stability and balance, compared with those who received balance training. Methods Participants Participants were recruited from Mackay Memorial Hospital in Taipei. Outpatients with PD were informed about the study by a neurologist. Eligibility required a diagnosis of idiopathic PD according to the United Kingdom Brain Bank Criteria [20] by the same neurologist. Information on age, gender, the more affected side, and disease duration were obtained through patient interviews and from medical charts. All participants met the following inclusion criteria: (1) Hoehn and Yahr stages I through III, (2) a score of ≥ 24 on the mini-mental state examination, (3) stable medication usage and (4) standing unaided to perform the measurement and training. The exclusion criteria were as follows: (1) histories of other neurological, cardiovascular, or orthopedic diseases affecting postural stability and (2) uncontrolled chronic diseases. In total, 48 individuals were identified as potential participants for this study. Of these, 22 participants gave informed consent and participated in the study. Study design This study was a subject-blinded, randomized controlled trial. The study protocol was approved by the Institutional Review Board of Mackay Memorial Hospital (reference number: 13MMHIS120) and was explained to all participants before their participation. The study was performed in accordance with the Declaration of Helsinki. Block randomization was used to assign participants to either the balance-based exergaming (BE) or the conventional balance training (BT) group. Assignment was performed by an independent person who selected one of a set of sealed envelopes 30 min before the intervention began. Participants in the BE and BT groups received an 8-week balance-based exergaming intervention, and conventional balance training, respectively. Measures of postural stability and functional balance were measured pre- and post-training. The measurement and intervention were conducted with participants in the “on” state, when they were moving freely and easily without dystonia, excessive rigidity or tremor. The data were collected in a university laboratory. Intervention Participants in both groups underwent balance training for 50 min per session, two sessions every week, for 8 weeks. Each training session began with a 10-min warm-up and ended with a 10-min cool-down. Both the warm-up and cool-down periods focused on stretching exercises of the trunk and extremities. Participants in the BE group received a 30-min balance-based exergaming intervention using the Kinect sensor (Microsoft Corporation, Redmond, WA, USA). The Kinect sensor incorporates infrared light and a video camera, which creates a 3D map of the area in front of it. This device provides full-body 3D motion capture. Four exergaming programs were used for training (Fig. 1), designed to incorporate an appropriate level of challenge to match the ability and fitness of people with PD. The first program was called Reaching task 1. In this task, participants were asked to reach toward a stationary target at a given location. The second program was called Reaching task 2. Participants were asked to track a moving object by lengthening the arm and immersing the hand into the object as it flew in 3D space. The third program was called Obstacle avoidance. Participants were instructed to avoid upcoming obstacles that approached from varying directions at random, by moving the body right/left or up/down. The final task was called Marching. Participants were instructed to step alternately without going forward or backward while following dynamic bars that were automatically rising and falling at a predetermined speed and frequency. During the training duration, the challenge level was increased progressively by adjusting the amplitude, frequency, speed, complexity and number of hints. The details of the exergaming programs are shown in Table 1.Fig. 1 Screen shots of interaction with the exergaming program. Four exergaming programs, Reaching task 1 (a), Reaching task 2 (b), Obstacle avoidance (c) and Marching (d), were designed and used for training Table 1 Program of balance-based exergaming intervention Program Action Progression Motor demand Reaching task 1 Standing in a given area and reaching toward a stationary target at different heights, depths and in different directions • Reaching length • Number of targets • Range of distribution • Amount of repetition • Weight shifting • Challenging limits of stability • Functional transitions Reaching task 2 Standing in a given area and tracking a moving object while extending arm and immersing the hand into the object as it flew in 3D space • Speed • Moving range • Pathway pattern • Remembered sequence or course of trajectory • Weight shifting • Arm coordination • Advance motor planning Obstacle avoidance Standing in a given area and preparing to avoid upcoming obstacles that randomly approached from varying directions by moving body sideways or up/down • Obstacle hitting ratio • Speed • Dual task • Hitting direction • Quick change strategy • Movement adaption • Agility Marching Alternating steps without going forward while following dynamic bars that automatically rose and fell at a predetermined speed and frequency • Frequency • Gap between steps • Functional stepping • Leg coordination • Single limb support Participants in the BT group underwent a 30-min conventional balance training session. The training program included reaching activities, weight-shifting activities and marching activities. The general training protocols used for the BT group were the same as those used for the BE group. The challenge level was increased progressively by changing the base of support, speed, complexity and deprivation of sensory inputs. Outcome measures Postural stability The limits of stability (LOS) and one-leg stance (OLS) tests were used to assess postural stability in this study. Participants were harnessed into a suspension system to prevent falls when performing the tasks. LOS performance was measured using the Smart Balance Master (NeuroCom International Inc., Clackamas, OR, USA) instrument to extract quantitative data [21–24]. The LOS test provides an assessment of the ability to intentionally displace the center of gravity (COG) to the participant’s stability limits without losing balance. In this task, participants were asked to quickly transfer their COG, while standing on stable force plates, toward eight targets spaced at 45° intervals around the COG, represented on a computer monitor. All participants underwent one practice trial followed by one test trial. In the LOS test, we measured reaction time (the time from the presentation of a start cue to the onset of the voluntary shifting of the participant’s COG toward the target position), movement velocity (average speed of COG movement based on the middle 90 % of the distance, measured in degrees per second), end point excursion (percentage of the distance achieved toward a target on the initial movement) and directional control (100 % being a straight line from the center of pressure to the intended target). The validity and reliability of the LOS test in people with neurological disease has been well established [25–27]. The OLS test is an assessment of postural steadiness [15, 28–31]. Participants were asked to cross their arms over the chest, and to stand on either the less or more affected leg, with the other leg raised so that the raised foot was near but not touching the ankle of the stance leg. The assessor timed the OLS test until participants either: (1) uncrossed the arms, (2) moved the stance leg, (3) moved the raised leg touching the floor or the stance leg, (4) opened the eyes on eyes-closed trials or (5) reached a maximum of 30 s. Each participant performed three trials with the eyes open, and three trials with the eyes closed. Data were averaged from the three trials. A previous study found a high degree of reliability (ICC = 0.87) in the OLS test in older adults [32]. Functional balance The Berg Balance Scale (BBS) and the timed up and go (TUG) test were used to assess functional balance. The BBS comprises a set of 14 balance-related tasks, ranging from standing up from a sitting position, to standing on one foot. The degree of success in each task is given a score from zero (unable) to four (independent), and the final measure is the sum of all scores. The highest possible score on the BBS is 56, which indicates excellent balance. The validity and reliability (ICC > 0.95) of BBS scores in people with PD has been established in several studies [33–35]. The TUG test is a mobility test requiring both static and dynamic balance. During the test, the assessors measured the time participants took to rise from a chair, walk 3 meters, turn around, walk back to the chair, and sit down. Each participant performed three trials of the TUG test. Data were averaged from the three trials. The TUG test has previously been found to have high validity and reliability (ICC > 0.87) for assessing balance in people with PD [36, 37]. Sample size The sample size calculation was based on a pilot study that tested eight participants at Hoehn and Yahr stages 1 and 2, indicating a difference of 0.2 s between pre- and post-training on reaction time in the LOS test. Based on this difference, a sample size calculation indicated that 20 participants would be sufficient for 85 % power (α = 0.05). Statistical analysis All analyses were performed using the SPSS 20.0 statistical package (SPSS Inc., Chicago, IL, USA). Descriptive statistics were generated for all variables, and distributions of variables were expressed as the mean ± standard deviation. Because of the relatively small number of participants included in the current study (N < 30) and since the results of a Shapiro-Wilk test did not allow us to assume that the data were normally distributed, nonparametric tests were employed. Comparison of two groups for general characteristics was made using chi-square or Mann-Whitney U test for categorical or continuous variables, respectively. The Friedman test, followed by a post hoc test, was used to determine differences in each dependent variable. The Wilcoxon signed-rank post hoc test was performed for within-group comparisons and the Mann-Whitney U post hoc test was performed for between-group comparisons. The statistical significance was set at P ≤ 0.05. Results A total of 48 individuals were screened and 22 enrolled between 2013 and 2014. Of these, 11 were assigned to the BT group, and 11 were assigned to the BE group. Of 22 participants, two did not complete the intervention (one in the BT group and one in the BE group). A flow diagram of the study protocol is shown in Fig. 2. The 20 participants who completed the intervention attended all intervention sessions. None of the participants reported any adverse events.Fig. 2 Flowchart of the experimental design The demographic characteristics of participants in both groups are presented in Table 2. Demographic differences between the two groups were not significant. Moreover, differences in all pre-intervention-selected outcome measures in the two groups were not significant (Table 3).Table 2 Baseline demographics and clinical characteristics of the subjects Balance-based exergaming group (N = 10) Balance training group (N = 10) P Age (years) 67.5 ± 9.96 68.8 ± 9.67 0.67 Sex (male/female) 9/1 7/3 0.58 Disease duration (years) 4.03 ± 3.74 5.22 ± 4.85 0.34 Hoehn and Yahr stage 1.6 ± 0.84 1.4 ± 0.52 0.73 Mini-Mental State Examination 27.4 ± 2.59 28.2 ± 1.99 0.40 More affected side (right/left) 8/2 5/5 0.35 Data are presented as the mean ± standard deviation or proportion Table 3 Outcome measures for each group Balance-based exergaming group (N = 10) Balance training group (N = 10) Friedman test Pre-training Post-training Pre-training Post-training P Limits of stability  Reaction time (sec) 0.96 ± 0.33 0.74 ± 0.24* 0.88 ± 0.24 0.79 ± 0.18 <0.001  Movement velocity (deg/sec) 3.37 ± 1.35 3.83 ± 0.97 4.19 ± 1.54 4.57 ± 1.41 0.07  Endpoint excursion (%) 75.2 ± 12.48 84 ± 12.04* 79.7 ± 13.84 81.8 ± 11.37 0.04  Directional control (%) 75.7 ± 8.78 78.9 ± 7.65*,† 70.9 ± 10.85 70.6 ± 9.37 0.02 One-leg stance  Less affected with eyes open (sec) 17.39 ± 12.87 15.16 ± 10.53 9.14 ± 9.63 12.98 ± 11.08 0.47  More affected with eyes open (sec) 15.06 ± 11.23 15.58 ± 11.58 13.72 ± 12.43 14.54 ± 9.65 0.09  Less affected with eyes closed (sec) 3.35 ± 2.85 6.1 ± 8.65* 2.71 ± 2.54 5.31 ± 7.68 0.002  More affected with eyes closed (sec) 3.06 ± 2.55 4.13 ± 2.74 5.88 ± 7.56 6.66 ± 8.41 0.16 Berg Balance Scale 50.9 ± 5.32 53.2 ± 2.86* 50.4 ± 4.79 53 ± 1.89* 0.001 Timed up and go (sec) 9.5 ± 2.45 8.71 ± 1.8* 10.05 ± 4.66 9.18 ± 3.42* 0.007 Data are presented as mean ± standard deviation *and †are P ≤ 0.05 for within-group and between-group comparisons, respectively The results of the interventions are presented in Table 3. Analysis of selected outcomes using the Friedman test revealed a significant effect of intervention type on reaction time, endpoint excursion and directional control in the LOS test, and in the less affected leg in the eyes-closed condition in the OLS test, the BBS and the TUG test. Within-group post hoc analysis revealed that balance-based exergaming training significantly improved LOS performance (improving reaction time from 0.96 ± 0.33 to 0.74 ± 0.24 s, end point excursion from 75.2 ± 12.48 to 84 ± 12.04 % and directional control from 75.7 ± 8.78 to 78.9 ± 7.65 %) and OLS on the less affected leg in the eyes-closed condition (from 3.35 ± 2.85 to 6.1 ± 8.65 s). Compared with the BT group (70.6 ± 9.37 %), the BE group (78.9 ± 7.65 %) exhibited better performance in directional control of LOS post-training. Functional balance in both groups, as measured by the BBS and the TUG test, was improved significantly post-training compared with pre-training. However, no significant differences were found between groups. Discussion This study produced two main findings: (1) balance-based exergaming training had a greater effect on postural stability compared with conventional balance training; and (2) both training programs improved functional balance in people with PD. The current study tested two balance training programs with similar training protocols. A recent meta-analysis examined the BBS, postural sway, TUG, and Functional Reach test as measures of postural stability, reporting that exercise therapy is an important treatment option for improving postural stability in people with PD [38]. The findings suggested that exercises containing a balance component were most beneficial in improving postural stability in people with PD [38]. In the current study, we used the LOS and OLS tests to measure postural stability, and the BBS and TUG tests to measure functional balance. The current findings were in line with the findings of Klamroth et al., who reported that balance training was beneficial for performance in the BBS and TUG tests [38]. Our findings revealed that only balance-based exergaming training produced positive effects on LOS and OLS, with particularly strong effects on directional control in LOS. These findings suggest that exergaming training using the Kinect sensor contributed to the beneficial gains we observed. As a therapeutic tool, the Kinect sensor can provide specific motor practice using full-body motion capture, which offers precise real-time information to guide performance and monitor body movement. Previous clinical trials indicated that exergaming programs using the Kinect sensor resulted in accurate capture of movement components [39, 40]. Our results revealed within-group improvements on most measures of postural stability during the exergaming intervention training period. Our exergaming programs involved various balance challenges. This may have contributed to our positive findings, involving actions focused on agility, challenging postural or locomotor-like skills, and reaching away from the base of support. All of these are involved in whole-body movements. In addition, the repetitive, real-time feedback and graded complexity in our exergaming programs may have contributed to the positive effects of training reflected in LOS performance. However, the movement velocity of LOS remained unchanged after exergaming training. Persistent bradykinesia [41] and a choice to focus on improving accuracy rather than faster motor performance among people with PD are possible reasons for our movement velocity findings [21]. The current results also revealed better OLS performance in the eyes-closed condition after exergaming training. A previous study using a Wii-based system reported similar results [15]. Because participants needed to focus on each joint position while carrying out the fine motor plan necessary for many of the tasks in the exergaming training, stimulation of proprioceptive feedback or an improvement in the internal representation of balance may have enhanced OLS performance. Little evidence is available regarding the minimal clinically important differences in postural stability and balance outcomes in people with PD. Evidence of minimal clinically important differences for LOS and OLS test in PD is lacking. Steffen and Seney reported a minimal detectable change of 5 points on the BBS for people with PD [34]. In the current study, we recorded a 2.45-point improvement after balance training for BBS. The minimally detectable change in TUG performance in people with PD has previously been reported to be 3.5 s [42], which is greater than the 0.83-second improvement observed in the present study. The small but significant changes observed in this study support the therapeutic use of exergaming interventions. However, a greater evidence base is required to support the clinical significance of these results. Several important characteristics have been identified for useful interventions in PD, suggesting that interventions should be task-specific, progressive, variable in terms of practice, and highly challenging [43, 44]. The exergaming programs designed for the current study involved each of these components. For specificity, the full-body motion capture method can be tailored for the needs of balance strategies. To create an appropriate practice resource and construct the progression and variability of program, we implemented enriched setting parameters by increasing speed, repetition and the addition of tasks. Additionally, the novel motor training gave participants more experience and an opportunity to explore or learn to negotiate the new challenges. Although only directional control in the LOS test showed a significant between-group difference, exergaming training using the Kinect system may provide additional benefits. Participants are able to practice free motions without wearing a sensor that could cause discomfort and inconvenience. Reduced staff intervention and the affordability of the device are important economic benefits of the system. Finally, considering the clinical implications of our findings, the current results suggest that the Kinect system can provide an assistive modality with therapeutic potential as a training tool under the supervision of a therapist. The current study involved several limitations. First, the sample size was small, limiting the strength to interpret our results. Second, calibration variability was observed during the preparation of each exergaming session. This issue may have influenced the effect of training because calibration was used to normalize each participant’s body information. This formed the basis of the exergaming programs that were tailored for individuals with varying levels of ability. Third, most participants in this study exhibited only mild impairment, and performance at baseline was relatively high. This may have limited the benefits received from training, and the generalizability of our findings to the target population. Finally, the absence of kinematic data meant we were unable to examine spatio-temporal changes in detailed movements. Conclusion The current study revealed that an 8-week period of balance-based exergaming training using the Kinect sensor resulted in a greater improvement of postural stability than conventional balance training. Both exergaming and conventional balance training had positive effects on functional balance. This trial supports the potential therapeutic use of exergaming aided by the Kinect sensor for people with PD. Importantly, the significant changes in BBS and TUG performance observed after both the exergaming and conventional balance training did not reach the minimal detectable change in patients with PD. Further studies on the use of exergaming are needed to verify the clinical implications of these results. Additional file Additional file 1: Dataset. (SAV 6 kb) Abbreviations BBSBerg Balance Scale BEBalance-based exergaming BTBalance training COGCenter of gravity LOSLimits of stability OLSOne-leg stance PDParkinson's disease TUGTimed up and go VRVirtual reality The authors wish to acknowledge Mr. Ray Chen and the Long Good team to develop exergaming program software for rehabilitation. Funding This work was supported by the National Science Council (grant number NSC100-2314-B-010-021-MY2). Availability of data and materials The dataset supporting the conclusion of this article is included within the article and its Additional file 1. Authors’ contributions MCS collected date, analyzed date and prepared the manuscript. SJC confirmed the pathological diagnosis of subjects and recruited subjects. YRY and RYW made substantial study design and revised the manuscript. All authors read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. Consent for publication The consent for publication had been obtained from all participants. Ethics approval and consent to participate The study protocol was approved by the Institutional Review Board of Mackay Memorial Hospital (reference number: 13MMHIS120) and was explained to all participants before their participation. All participants gave their informed consent. ==== Refs References 1. Morris ME Movement disorders in people with Parkinson disease: a model for physical therapy Phys Ther 2000 80 578 97 10842411 2. Bloem BR Grimbergen YA Cramer M Willemsen M Zwinderman AH Prospective assessment of falls in Parkinson’s disease J Neurol 2001 248 950 8 10.1007/s004150170047 11757958 3. Wood BH Bilclough JA Bowron A Walker RW Incidence and prediction of falls in Parkinson’s disease: a prospective multidisciplinary study J Neurol Neurosurg Psychiatry 2002 72 721 5 10.1136/jnnp.72.6.721 12023412 4. Smith BA Jacobs JV Horak FB Effects of amplitude cueing on postural responses and preparatory cortical activity of people with Parkinson disease J Neurol Phys Ther 2014 38 207 15 10.1097/NPT.0000000000000058 25198870 5. Lee BC Thrasher TA Fisher SP Layne CS The effects of different sensory augmentation on weight-shifting balance exercises in Parkinson’s disease and healthy elderly people: a proof-of-concept study J Neuroeng Rehabil 2015 12 75 10.1186/s12984-015-0064-y 26329918 6. Konczak J Corcos DM Horak F Poizner H Shapiro M Tuite P Proprioception and motor control in Parkinson’s disease J Mot Behav 2009 41 543 52 10.3200/35-09-002 19592360 7. Nieuwboer A Rochester L Müncks L Swinnen SP Motor learning in Parkinson’s disease: limitations and potential for rehabilitation Parkinsonism Relat Disord 2009 15 S53 8 10.1016/S1353-8020(09)70781-3 20083008 8. Henderson A Korner-Bitensky N Levin M Virtual reality in stroke rehabilitation: a systematic review of its effectiveness for upper limb motor recovery Top Stroke Rehabil 2007 14 52 61 10.1310/tsr1402-52 17517575 9. Dobkin BH Training and exercise to drive poststroke recovery Nat Clin Pract Neurol 2008 4 76 85 10.1038/ncpneuro0709 18256679 10. Langhorne P Coupar F Pollock A Motor recovery after stroke: a systematic review Lancet Neurol 2009 8 741 54 10.1016/S1474-4422(09)70150-4 19608100 11. Butler DP Willett K Wii-habilitation: is there a role in trauma? Injury 2010 41 883 5 10.1016/j.injury.2010.03.024 20398899 12. Adamovich SV Fluet GG Tunik E Merians AS Sensorimotor training in virtual reality: a review Neuro Rehabilitation 2009 25 29 44 19713617 13. Kakei S Hoffman DS Strick PL Sensorimotor transformations in cortical motor areas Neurosci Res 2003 46 1 10 10.1016/S0168-0102(03)00031-2 12725907 14. Donath L, Rössler R, Faude O. Effects of virtual reality training (exergaming) compared to alternative exercise training and passive control on standing balance and functional mobility in healthy community-dwelling seniors: a meta-analytical review. Sports Med. 2016. Feb 17. [Epub aheadof print] after 2016. 15. Esculier JF Vaudrin J Bériault P Gagnon K Tremblay LE Home-based balance training programme using Wii Fit with balance board for Parkinsons’s disease: a pilot study J Rehabil Med 2012 44 144 50 10.2340/16501977-0922 22266676 16. Pompeu JE Mendes FA Silva KG Lobo AM Oliveira Tde P Zomignani AP Effect of Nintendo Wii-based motor and cognitive training on activities of daily living in patients with Parkinson’s disease: a randomised clinical trial Physiotherapy 2012 98 196 204 10.1016/j.physio.2012.06.004 22898575 17. Lange B Chang CY Suma E Newman B Rizzo AS Bolas M Development and evaluation of low cost game-based balance rehabilitation tool using the Microsoft Kinect sensor Conf Proc IEEE Eng Med Biol Soc 2011 2011 1831 4 22254685 18. Clark RA Pua YH Fortin K Ritchie C Webster KE Denehy L Validity of the Microsoft Kinect for assessment of postural control Gait Posture 2012 36 372 7 10.1016/j.gaitpost.2012.03.033 22633015 19. Clark RA Pua YH Oliveira CC Bower KJ Thilarajah S McGaw R Reliability and concurrent validity of the Microsoft Xbox One Kinect for assessment of standing balance and postural control Gait Posture 2015 42 210 3 10.1016/j.gaitpost.2015.03.005 26009500 20. Hughes AJ Daniel SE Kilford L Lees AJ Accuracy of clinical diagnosis of idiopathic Parkinson’s disease: a clinico-pathological study of 100 cases J Neurol Neurosurg Psychiatry 1992 55 181 4 10.1136/jnnp.55.3.181 1564476 21. Jessop RT Horowicz C Dibble LE Motor learning and Parkinson disease: refinement of movement velocity and endpoint excursion in a limits of stability balance task Neurorehabil Neural Repair 2006 20 459 67 10.1177/1545968306287107 17082501 22. Pickerill ML Harter RA Validity and reliability of limits-of-stability testing: a comparison of 2 postural stability evaluation devices J Athl Train 2011 46 600 6 22488184 23. Kara B Genc A Colakoglu BD Cakmur R The effect of supervised exercises on static and dynamic balance in Parkinson’s disease patients NeuroRehabilitation 2012 30 351 7 22672951 24. Li F Harmer P Fitzgerald K Eckstrom E Stock R Galver J Tai chi and postural stability in patients with Parkinson’s disease N Engl J Med 2012 366 511 9 10.1056/NEJMoa1107911 22316445 25. Liston RA Brouwer BJ Reliability and validity of measures obtained from stroke patients using the Balance Master Arch Phys Med Rehabil 1996 77 425 30 10.1016/S0003-9993(96)90028-3 8629916 26. Newstead AH Hinman MR Tomberlin JA Reliability of the Berg Balance Scale and balance master limits of stability tests for individuals with brain injury J Neurol Phys Ther 2005 29 18 23 10.1097/01.NPT.0000282258.74325.cf 16386157 27. Chien CW Hu MH Tang PF Sheu CF Hsieh CL A comparison of psychometric properties of the smart balance master system and the postural assessment scale for stroke in people who have had mild stroke Arch Phys Med Rehabil 2007 88 374 80 10.1016/j.apmr.2006.11.019 17321832 28. Jacobs JV Horak FB Tran VK Nutt JG Multiple balance tests improve the assessment of postural stability in subjects with Parkinson’s disease J Neurol Neurosurg Psychiatry 2006 77 322 6 10.1136/jnnp.2005.068742 16484639 29. Schilling BK Karlage RE LeDoux MS Pfeiffer RF Weiss LW Falvo MJ Impaired leg extensor strength in individuals with Parkinson disease and relatedness to functional mobility Parkinsonism Relat Disord 2009 15 776 80 10.1016/j.parkreldis.2009.06.002 19560392 30. Mak MK Pang MY Parkinsonian single fallers versus recurrent fallers: different fall characteristics and clinical features J Neurol 2010 257 1543 51 10.1007/s00415-010-5573-9 20449601 31. Springer BA Marin R Cyhan T Roberts H Gill NW Normative values for the unipedal stance test with eyes open and closed J Geriatr Phys Ther 2007 30 8 15 10.1519/00139143-200704000-00003 19839175 32. da Silva RA Bilodeau M Parreira RB Teixeira DC Amorim CF Age-related differences in time-limit performance and force platform-based balance measures during one-leg stance J Electromyogr Kinesiol 2013 23 634 9 10.1016/j.jelekin.2013.01.008 23403137 33. Qutubuddin AA Pegg PO Cifu DX Brown R McNamee S Carne W Validating the Berg Balance Scale for patients with Parkinson’s disease: a key to rehabilitation evaluation Arch Phys Med Rehabil 2005 86 789 92 10.1016/j.apmr.2004.11.005 15827933 34. Steffen T Seney M Test-retest reliability and minimal detectable change on balance and ambulation tests, the 36-item short-form health survey, and the unified Parkinson disease rating scale in people with parkinsonism Phys Ther 2008 88 733 46 10.2522/ptj.20070214 18356292 35. Schlenstedt C Brombacher S Hartwigsen G Weisser B Möller B Deuschl G Comparing the Fullerton Advanced Balance Scale with the Mini-BESTest and Berg Balance Scale to assess postural control in patients with Parkinson disease Arch Phys Med Rehabil 2015 96 218 25 10.1016/j.apmr.2014.09.002 25261718 36. Morris S Morris ME Iansek R Reliability of measurements obtained with the Timed “Up & Go” test in people with Parkinson disease Phys Ther 2001 81 810 8 11175678 37. Verheyden G Kampshoff CS Burnett ME Cashell J Martinelli L Nicholas A Psychometric properties of 3 functional mobility tests for people with Parkinson disease Phys Ther 2014 94 230 9 10.2522/ptj.20130141 24231225 38. Klamroth S Steib S Devan S Pfeifer K Effects of exercise therapy on postural instability in Parkinson disease: a meta-analysis J Neurol Phys Ther 2016 40 3 14 10.1097/NPT.0000000000000117 26655098 39. Chang YJ Chen SF Huang JD A Kinect-based system for physical rehabilitation: a pilot study for young adults with motor disabilities Res Dev Disabil 2011 32 2566 70 10.1016/j.ridd.2011.07.002 21784612 40. Sin H Lee G Additional virtual reality training using Xbox Kinect in stroke survivors with hemiplegia Am J Phys Med Rehabil 2013 92 871 80 10.1097/PHM.0b013e3182a38e40 24051993 41. Dimitrova D Horak FB Nutt JG Postural muscle responses to multidirectional translations in patients with Parkinson’s disease J Neurophysiol 2004 91 489 501 10.1152/jn.00094.2003 12944541 42. Huang SL Hsieh CL Wu RM Tai CH Lin CH Lu WS Minimal detectable change of the timed “up & go” test and the dynamic gait index in people with Parkinson disease Phys Ther 2011 91 114 21 10.2522/ptj.20090126 20947672 43. King LA Horak FB Delaying mobility disability in people with Parkinson disease using a sensorimotor agility exercise program Phys Ther 2009 89 384 93 10.2522/ptj.20080214 19228832 44. Conradsson D Löfgren N Ståhle A Hagströmer M Franzén E A novel conceptual framework for balance training in Parkinson’s disease-study protocol for a randomised controlled trial BMC Neurol 2012 12 111 10.1186/1471-2377-12-111 23017069

PMC005xxxxxx/PMC5002325.txt

==== Front Virol JVirol. JVirology Journal1743-422XBioMed Central London 60110.1186/s12985-016-0601-8Short ReportIdentification and genomic characterization of a novel species of feline anellovirus Zhang Wen zhangwen@ujs.edu.cn 1Wang Hua wangjiahan1979@163.com 1Wang Yan wangyan_jtu@126.com 1Liu Zhijian zhijianchange1008@outlook.com 1Li Jingjiao 847306075@qq.com 2Guo Lianghua 306664973@qq.com 2Yang Shixing johnsonyang1979@163.com 1Shen Quan shenquanfly@yahoo.com 1Zhao Xiaoying 1508139522@qq.com 1Cui Li l_cui@sina.cn 2Hua Xiuguo hxg@sjtu.edu.cn 21 School of Medicine, Jiangsu University, 301 Xuefu Road, Zhenjiang, Jiangsu 212013 People’s Republic of China 2 School of Agriculture and Biology, Shanghai Jiaotong University, 800 Dongchuan Road, Shanghai, 200240 China 27 8 2016 27 8 2016 2016 13 1 1469 6 2016 15 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Here, a novel feline anellovirus strain (named FelineAV621 and GenBank no. KX262893) was detected in two cats with diarrhea. The complete genome of FelineAV621 is 2409 nt long with a G+C content of 56.67 %, including three open reading frames (ORFs). Phylogenetic analysis based on the amino acid sequence of the putative capsid protein (ORF1) indicated that FelineAV621 belonged to a novel anellovirus species inside a clade containing the seal anellovirus, canine TTVs, and porcine TTVs, but was distant from all the previous feline anelloviruses. Keywords CatAnellovirusComplete genomeViral metagenomicsPhylogenetic analysishttp://dx.doi.org/10.13039/501100001809National Natural Science Foundation of China3130210731270186Zhang Wen http://dx.doi.org/10.13039/501100004608Natural Science Foundation of Jiangsu ProvinceBK20130542BK20140537Wang Hua Yang Shixing http://dx.doi.org/10.13039/501100004608 Natural Science Foundation of Jiangsu ProvinceBK20130502Wang Yan China Postdoctoral Special Foundation2015T80503Zhang Wen issue-copyright-statement© The Author(s) 2016 ==== Body Findings Anelloviruses belong to non-enveloped, circular, single-stranded DNA viruses with genome of 2.1–3.9 kb in length depending on the isolate analyzed [1]. Anelloviruses are subgrouped into Torque teno virus (TTV), Torque teno mini virus (TTMV), Torque teno midi virus (TTMDV) and small anellovirus (SAV) [2]. Although anelloviruses are widely prevalent in humans and animals and suspected to be associated with some diseases of humans [3–6] or animals [7–9], the etiological role of anelloviruses has not yet been clearly identified. The first report of discovering feline anellovirus was published by Okamoto et al. [10] and subsequently reported in several other publications [11–14]. Phylogenetic analysis based on the amino acid sequences of ORF1 of feline anelloviruses available in GenBank revealed that these known feline anellovirus clustered together and were classified into two species, Torque teno felis virus 1 (FcTTV1) and Torque teno felis virus 2 (FcTTV2), belonging to the genus Etatorquevirus within the family Anelloviridae [13]. The aim of this study is to characterize the genome sequence of a divergent feline anellovirus which is present in the fecal and blood samples from domestic cats with diarrhea. In a previous study, where viral metagenomic method was used to investigate the fecal virome of cats in a shelter in California, USA, several sequencing reads showed amino acid similarity to viruses in the family Anelloviridae [14]. In this study, nested primers designed based on a 621 bp contig assembled by geneious software from the data of that study [14] are used to investigate whether this anellovirus is present in cat population in China. The primers used here are Anel621FO (5′- ACTTCCTCCTGGTCGGGCGT-3′) and Anel621RO (5′- TGGGGAGGGGTTGATGCCCA-3′) for the 1st round PCR, and Anel621FI (5′- GGCTCTCGTGCGGTTTGGGA-3′) and Anel621RI (5′- TCCCGTCGTCCCACCACCAT-3′) for the 2nd round PCR. The PCR product size of the 2nd round PCR is about 215 bp. PCR screening was performed to detect the anellovirus-like sequence in a total 22 samples, including 11 blood and 11 fecal samples collected from 11 cats with diarrhea from Jan. 2014 to Dec. 2015. Results indicated that two blood samples and one fecal samples were positive, including one cat positive both in blood and fecal samples, and one positive blood sample. The specific DNA band was T-A cloned and sequencing result indicated that the two sequences from blood and fecal samples of the same cat were identical and had one nucleotide difference from the sequence from the other blood sample in the present study. Sequence analysis indicated that the two sequences from cats in the present study shared 97.7 % sequence identity with the sequences from the California cats [14], with five different nt over the 215 bp sequence fragment. To sequence the entire genome of the feline anellovirus, the viral nucleic acid was extracted form one of the cat the blood samples. Circular viral DNA molecules were preferentially amplified by rolling circle amplification (RCA) using random hexamer primers and the Illustra TempliPhi amplification kit (GE Healthcare). Outwardly pointing specific PCR primers FelAVF (5′-ATGCTCAACACCACAAACGC-3′) and FelAVR (5′-TGTAATCCCAAACCGCACGA-3′) were designed from the viral metagenomic sequence [14] and then used to amplify the small circular genomes by inverse PCR. The PCR products were then sequenced by primer walking and Sanger sequencing. Sequencing results indicated that the complete genome of the feline anellovirus strain (named FelineAV621 and GenBank no. KX262893) was 2409 nt long, with a G+C content of 56.67 %. The genome organization of FelineAV621 is consistent with that of other anelloviruses, containing three open reading frames (ORFs) (Fig. 1). The ORF1, the largest ORF, does not display any significant nucleotide identity to any virus sequence using BLASTn in GenBank, but shares low identity with a seal anellovirus (GenBank no.: KM262782) (30.2 % identity) [15] and a pine marten torque teno virus (GenBank no.: JN704611) (29.4 % identity) [16] based on the amino acid sequence of ORF1. ORF1 encodes a 630 amino acid long putative capsid protein which presents three potential glycosylation sites; no signal peptides were identified. As in other anelloviruses, the capsid protein is arginine rich in its N terminus, with 31 arginine residues in the first 60 amino acids (51.67 %). ORF2 encodes 109 amino acids and shares 29–40 % amino acid sequence identities with several ORF2 proteins of anelloviruses in GenBank. ORF3 encodes147 amino acids and shares no amino acid sequence identity with proteins in GenBank.Fig. 1 Genomic organization of the feline anellovirus (named FelineAV621 and GenBank no. KX262893) To determine the relationship of FelineAV621 to other anelloviruses, phylogenetic analysis was performed based on the ORF1 amino acid sequences of FelineAV621, its best BLASTp matches in GenBank and the representative members of related viruses. Sequence alignment was performed using CLUSTALW with the default settings. A phylogenetic tree with 1000 bootstrap resamples of the alignment data sets was generated using the maximum likelihood method based on Jones-Taylor-Thornton (JTT) model in MEGA5.0. The anellovirus tree topology was consistent with results of earlier studies [1, 9, 17] and revealed that FelineAV621 is a novel anellovirus species that belongs to a clade that contains the seal anellovirus, canine TTVs, and porcine TTVs, but distant from the previous four feline anelloviruses (Fig. 2), suggesting cat can carry at least two different species of anelloviruses.Fig. 2 Phylogenetic tree based on the full-length amino acid sequence of ORF1 depicting relationships among the members of the family Anelloviridae. The newly discovered feline anellovirus is indicated by a diamond shape. The host is indicated in animal profiles and GenBank accession numbers are shown Taken together, a novel species of feline anellovirus was detected in two cats with diarrhea, including both blood and fecal samples of one cat and the blood sample of the other one. The complete genome of this novel anellovirus was sequenced and characterized from one of the blood sample. Phylogenetic analysis revealed that this feline anellovirus belonged to a novel anellovirus species which is significantly divergent from the previous feline anellovirus strains. Due to the limited sample number and lacking serologic evidence in the present study, whether FelineAV621 can really infect cats and cause diarrhea will require further epidemiologic study based on a larger sample size and testing cat sera for specific antibodies. Abbreviations BLASTBasic local alignment search tool ORFOpen reading frame PCRPolymerase chain reaction SAVSmall anellovirus TTMVTorque teno mini virus, TTMDV, Torque teno midi virus TTVTorque teno virus Funding This work was partly supported by the National Natural Science Foundation of China No. 31302107 and 31270186, the Natural Science Foundation of Jiangsu Province No. BK20130542, BK20140537, and BK20130502, and China Postdoctoral Special Foundation No. 2015T80503. Authors’ contributions WZ and XH conceived the study. WZ and HW performed all the experiments. WZ wrote the paper. All authors read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. ==== Refs References 1. Biagini P Classification of TTV and related viruses (anelloviruses) Curr Top Microbiol Immunol 2009 331 21 33 19230555 2. Bernardin F Operskalski E Busch M Delwart E Transfusion transmission of highly prevalent commensal human viruses Transfusion 2010 50 2474 83 10.1111/j.1537-2995.2010.02699.x 20497515 3. Tajiri H Tanaka T Sawada A Etani Y Kozaiwa K Mushiake S Mishiro S Three cases with TT virus infection and idiopathic neonatal hepatitis Intervirology 2001 44 364 9 10.1159/000050073 11805444 4. Gergely P Perl A Poór G Possible pathogenic nature of the recently discovered TT virus: does it play a role in autoimmune rheumatic diseases? Autoimmun Rev 2006 6 5 9 10.1016/j.autrev.2006.03.002 17110309 5. Ishimura M Ohga S Ichiyama M Kusuhara K Takada H Hara T Takahashi M Okamoto H Hepatitis-associated aplastic anemia during a primary infection of genotype 1a torque teno virus Eur J Pediatr 2010 169 899 902 10.1007/s00431-009-1116-8 19998044 6. Miyamoto M Takahashi H Sakata I Adachi Y Hepatitis-associated aplastic anemia and transfusion-transmitted virus infection Intern Med 2000 39 1068 70 10.2169/internalmedicine.39.1068 11197792 7. Krakowka S Hartunian C Hamberg A Shoup D Rings M Zhang Y Allan G Ellis JA Evaluation of induction of porcine dermatitis and nephropathy syndrome in gnotobiotic pigs with negative results for porcine circovirus type 2 Am J Vet Res 2008 69 1615 22 10.2460/ajvr.69.12.1615 19046009 8. Teixeira TF Cibulski SP dos Santos HF Wendlant A de Sales Lima FE Schmidt C Franco AC Roehe PM Torque teno sus virus 1 (TTSuV1) and 2 (TTSuV2) viral loads in serum of postweaning multisystemic wasting syndrome (PMWS)-affected and healthy pigs in Brazil Res Vet Sci 2015 101 38 41 10.1016/j.rvsc.2015.05.016 26267087 9. Li L Giannitti F Low J Keyes C Ullmann LS Deng X Aleman M Pesavento PA Pusterla N Delwart E Exploring the virome of diseased horses J Gen Virol 2015 96 2721 33 10.1099/vir.0.000199 26044792 10. Okamoto H Takahashi M Nishizawa T Tawara A Fukai K Muramatsu U Naito Y Yoshikawa A Genomic characterization of TT viruses (TTVs) in pigs, cats and dogs and their relatedness with species-specific TTVs in primates and tupaias J Gen Virol 2002 83 Pt 6 1291 7 10.1099/0022-1317-83-6-1291 12029143 11. Zhu CX Shan TL Cui L Luo XN Liu ZJ Tang SD Liu ZW Yuan CL Lan DL Zhao W Hua XG Molecular detection and sequence analysis of feline Torque teno virus (TTV) in China Virus Res 2011 156 13 6 10.1016/j.virusres.2010.12.008 21185891 12. Biagini P Uch R Belhouchet M Attoui H Cantaloube J-F Brisbarre N de Micco P Circular genomes related to anelloviruses identified in human and animal samples by using a combined rolling-circle amplification/sequence-independent single primer amplification approach J Gen Virol 2007 88 Pt 10 2696 701 10.1099/vir.0.83071-0 17872521 13. Jarošová V Hrazdilová K Filipejová Z Schánilec P Celer V Whole genome sequencing and phylogenetic analysis of feline anelloviruses Infect Genet Evol 2015 32 130 4 10.1016/j.meegid.2015.02.015 25724090 14. Zhang W Li L Deng X Kapusinszky B Pesavento PA Delwart E Faecal virome of cats in an animal shelter J Gen Virol 2014 95 Pt 11 2553 64 10.1099/vir.0.069674-0 25078300 15. Bodewes R Contreras GJS García AR Hapsari R van de Bildt MWG Kuiken T Osterhaus ADME Identification of DNA sequences that imply a novel gammaherpesvirus in seals J Gen Virol 2015 96 Pt 5 1109 14 10.1099/vir.0.000029 25524165 16. Van den Brand JMA van Leeuwen M Schapendonk CM Simon JH Haagmans BL Osterhaus ADME Smits SL Metagenomic analysis of the viral flora of pine marten and European badger feces J Virol 2012 86 2360 5 10.1128/JVI.06373-11 22171250 17. Jones MS Kapoor A Lukashov VV Simmonds P Hecht F Delwart E New DNA viruses identified in patients with acute viral infection syndrome J Virol 2005 79 8230 6 10.1128/JVI.79.13.8230-8236.2005 15956568

PMC005xxxxxx/PMC5002326.txt

==== Front BMC Pulm MedBMC Pulm MedBMC Pulmonary Medicine1471-2466BioMed Central London 28910.1186/s12890-016-0289-yResearch ArticleNoninvasive versus invasive mechanical ventilation for immunocompromised patients with acute respiratory failure: a systematic review and meta-analysis Wang Tao 1Zhang Lixi 2Luo Kai 3He Jianqiang 1Ma Yong 1Li Zongru 4Zhao Na 5Xu Qun 3Li Yi billliyi@126.com 1Yu Xuezhong erxuezhongyu@sina.cn 11 Emergency Department, Peking Union Medical College Hospital, Beijing, 100730 China 2 Department of Cardiology, Peking Union Medical College Hospital, Beijing, 100730 China 3 Department of Epidemiology and Biostatistics, Institute of Basic Medical Sciences Chinese Academy of Medical Sciences, School of Basic Medicine Peking Union Medical College, Beijing, 100005 China 4 Department of Pneumology, Peking Union Medical College Hospital, Beijing, 100730 China 5 Department of Anesthesiology, Beijing Obstetrics and Gynecology Hospital, Capital Medical University, Beijing, 100026 China 27 8 2016 27 8 2016 2016 16 1 12917 5 2016 19 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background To determine the effects of noninvasive mechanical ventilation (NIV) compared with invasive mechanical ventilation (IMV) as the initial mechanical ventilation on clinical outcomes when used for treatment of acute respiratory failure (ARF) in immunocompromised patients. Methods We searched PubMed, EMBASE, the Cochrane Central Register of Controlled Trials (CENTRAL), the Chinese Biomedical Literature Database (CBM) and other databases. Subgroup analyses by disease severity and causes of immunodeficiency were also conducted. Results Thirteen observational studies with a total of 2552 patients were included. Compared to IMV, NIV was shown to significantly reduce in-hospital mortality (OR 0.43, 95 % CI 0.23 to 0.80, P value = 0.007) and 30-day mortality (OR 0.34, 95 % CI 0.20 to 0.61, P value < 0.0001) in overall analysis. Subgroup analysis showed NIV had great advantage over IMV for less severe, AIDS, BMT and hematological malignancies patients in reducing mortality and duration of ICU stay. Conclusions The overall evidence we obtained shows NIV does more benefits or at least no harm to ARF patients with certain causes of immunodeficiency or who are less severe. Electronic supplementary material The online version of this article (doi:10.1186/s12890-016-0289-y) contains supplementary material, which is available to authorized users. Keywords Noninvasive mechanical ventilationInvasive mechanical ventilationAcute respiratory failureImmunocompromised patientsSystematic reviewMeta-analysisissue-copyright-statement© The Author(s) 2016 ==== Body Background The immunocompromised condition is defined as a state of subnormal immune response of the host to a foreign antigen, which could be congenital (primary) or acquired (secondary) [1], with primary immunodeficiencies caused by gene mutations and secondary caused by malignancy, chemotherapies of malignancy, malnutrition, aging, viral infection, immunosuppressive medication for treatment of a variety of disorders such as autoimmune disease and organ transplantation. The number of immunocompromised patients has increased dramatically over recent decades [2]. In spite of better antimicrobial agents and preventive measures, infections continue to be one of the most frequent complications in immunocompromised patients and have a high mortality rate of 30 to 90 % [2], with the highest when acute respiratory failure (ARF) occurs. Thus, early diagnosis and proper intervention are essential for better outcomes. Noninvasive mechanical ventilation (NIV) and invasive mechanical ventilation (IMV) are two approaches for providing supplemental oxygen for patients with relatively severe ARF. NIV has gained more and more popularity since its first application in 1980s [3], and is now widely accepted as a first-line intervention for certain forms of ARF, including acute exacerbation of chronic obstructive pulmonary disease (AECOPD) and cardiogenic pulmonary edema [4–14]. Although the use of NIV as a first-line strategy for immunocompromised patients with acute respiratory failure was recommended by Canadian Critical Care Society Noninvasive Ventilation Guidelines Group(Grade 2B: weak recommendation and moderate evidence quality) [11], recent studies showed noninvasive ventilation might not be the appropriate choice for all immunocompromised patients [15–17]. The choice of NIV versus IMV for immunocompromised patients, especially for relatively severe ARF, is still under debate. On the other hand, there are studies indicating that the use of NIV as the initial treatment for certain population may delay intubation, which may increase mortality and cost of health care [17–20]. Identifying the proper candidates and evidence for the effect of NIV is of great importance for better outcome of immunocompromised patients with ARF. This review will provide a systematic review of the evidences to determine the effects of NIV compared to IMV on clinical outcomes when used for treatment of acute respiratory failure in immunocompromised patients. Methods Search strategy We searched the following databases (Additional file 1): PubMed, EMBASE, the Cochrane Central Register of Controlled Trials (CENTRAL), and Chinese Biomedical Literature Database (CBM, in Chinese). We also searched other resources including Web of Science (WOS), National technical information service conference proceedings (NTIS), Open Grey (OG), and conference proceedings for relevant abstracts, online clinical trial registers for ongoing and recently completed studies including Controlled Clinical Trials (http://www.controlled-trials.com/), government registries (http://www.clinicaltrials.gov), and World Health Organization registries (http://www.who.int/trialsearch/). There was no study type, language, date, or publication type restrictions. We searched the bibliography of all included studies and requested original data from the primary authors when necessary. The most recent search was conducted on April 15th, 2016. Inclusion criteria We included randomized controlled trials (RCTs) and non-RCTs. We accepted the definition of acute respiratory failure as the state when ratio of the partial pressure of arterial oxygen to the fraction of inspired oxygen (PaO2/FiO2) ≤ 300 mmHg. For studies enrolling patients with other causes for mechanical ventilation, we stipulated that a minimum of 85 % of patients must have ARF to meet the inclusion criteria. Alternatively, patients were considered as immunocompromised when clinically diagnosed as: 1) HIV-infected individuals (with or without acquired immune deficiency syndrome, AIDS), 2) individuals on immunosuppressive therapy (e.g., cytotoxic agents, glucocorticoids, etc.), 3) transplantation individuals (i.e., solid organ transplantation or bone marrow transplantation), 4) hematologic cancers, 5) certain trauma or surgery (splenectomy), 6) secondary to metabolic diseases (e.g., malnutrition, noncontrolled diabetes, uremia). The intervention group included patients who received NIV as the initial mechanical ventilation technique, in addition to standard medical care, irrespective of whether IMV was also used later during the hospital stay. The control group included patients who received IMV as the initial MV technique. Patients with absolute contraindications of NIV including respiratory arrest and inability to fit the mask, or with underlying pathologies where NIV and IMV is both contraindicated, such as facial trauma, were excluded. We included studies in which at least one of the review-defined outcomes was identified. The primary outcomes were all-cause mortality, including mortality in hospital or intensive care unit (ICU), and 30-day mortality after ICU admission. The secondary outcomes included duration of hospitalization and ICU stay, nosocomial infections, and duration of mechanical ventilation. Data extraction and study quality Two authors (JH and YM) independently assessed studies for inclusion retrieved from electronic searches and other resources and extracted data from the included studies. In cases of ambiguity or insufficient data, we requested additional information from study authors. Disagreements were ultimately resolved by a third author (YL). We used a standardised data extraction form to collect the following data: 1) General information: study ID, title, authors, source, language and year of publication, country, and source of funding. 2) Study characteristics: study type, hospital settings, and dates of study. 3) Participants: age, sex, diagnosis standard and cause of ARF and the immune status, sample size, baseline physiological variables including Acute Physiology and Chronic Health Evaluation (APACHE II) scores, Simplified Acute Physiology Scores (SAPS II) [21] and details of respiratory state on admission. 4) Interventions: intervention used in each group and number of each group; indications, contraindications, settings and duration of intervention. 5) Outcomes: mortality (in ICU, in hospital and 30-day), duration of mechanical ventilation, ICU stay and hospitalization, number of participants with development of major complications or nosocomial infections, numbers experiencing each outcome, and information of follow-up. 6) Methodological quality: items in the Cochrane Collaboration’s tool [22] and Newcastle Ottawa Scale for assessing risk of bias (Additional file 1) [23]. Two review authors (TW, LXZ) independently assessed the risks of bias in included studies. We used the adapted Newcastle Ottawa Scale to assess the risks of bias, including selection, comparability and outcome/exposure (Additional file 1) [23]. We also explored other risks of bias such as reporting bias. We then analyzed the quality of evidence following the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology [24, 25]. Statistical analysis We used random-effect models throughout the analysis to take account of the between-study variance in our findings. All analyses were performed on R3.2 with the meta and metafor packages. We considered P values < 0.05 to be statistically significant. When outcomes were dichotomous, we used unadjusted odds ratio (OR) for non-RCTs (cohort and case-control studies); when outcomes were continuous, we used mean difference (MD); each was provided with 95 % confidence intervals (95 % CIs). We followed the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions [22] in assessing the impact of heterogeneity. In general, we interpreted an I2 value > 60 % as having substantial heterogeneity and meta analyses cannot be done; an I2 value ≤ 40 % would suggest little concern about statistical heterogeneity. If data was sufficient, we would undertake the following subgroup analyses for each comparison:Severity of disease: based on SAPS II [21] at admission. Studies were classified into less severe group when a mean of SAPS II < 60 and more severe group when a mean of SAPS II ≥ 60. Participants in more severe group were considered to have more severe disease than those in less severe group. In addition, disease severity could also be assessed by Sequential Organ Failure Assessment (SOFA) [26] and PaO2/FiO2. Causes of the immunocompromised status. Results We retrieved 3359 records from the electronic database searches (Fig. 1 for a flow diagram of studies identified). Thirteen non-RCTs [17, 20, 27–37] met all of the inclusion criteria with a detailed description of each available in Table 1. There were one prospective cohort study [20], three retrospective cohort studies [28, 31, 34] and nine retrospective case-control studies [17, 27, 29, 30, 32, 33, 35–37]. A total of 2552 patients were included in our final analysis. Sample size in each study ranged between 15 and 1302. One study included children with a mean age of 9 years old [33]. The rest recruited adults with age ranged from 17 to 82. All studies met the diagnosis criteria for acute respiratory failure and immunocompromised status. Main types of ARF included adult respiratory distress syndrome/acute lung injury (ARDS/ALI), infectious pneumonia, chronic obstructive pulmonary disease (COPD), and pulmonary edema. Causes of patients’ immunocompromised status included AIDS, hematologic malignancies, solid tumors, bone marrow transplantation (BMT), chemotherapy and receipts of immunosuppressive medications. All included studies compared NIV with IMV as the first mechanical ventilation technique in immunocompromised patients with ARF. Settings in the NIV group were as follows: (1) Ventilation modes: two studies [27, 32] used continuous positive airway pressure (CPAP) ventilation only; three [33, 34, 36] solely used bi­level positive airway pressure (Bi­PAP); another five [17, 20, 28, 30, 37] solely used pressure support ventilation (PSV); another one [31] used both CPAP and Bi-PAP; no description was given in the remaining studies. (2) Interfaces: full face mask, face mask, nasal mask or helmet. (3) Duration and continuity: eight studies applied NIV continuously for the first 24 h in the NIV group. Four studies didn’t report the duration of mechanical ventilation. (4) Positive End Expiratory Pressure (PEEP): in most studies, PEEP ranged between 3–10 cmH2O, adjusted with ventilation flow, pressure support and other settings to achieve a proper clinical outcome, including a pulse oximetry oxyhemoglobin saturation ≥ 95 %, an oxygen saturation ≥ 90 %, FiO2 ≤ 0.6, an exhaled tidal volume of 8 to 10 ml/ kg or a respiratory rate ≤ 25 breaths/min. A portion of patients who initially received NIV switched to IMV due to NIV failure, with a mean intubation rate of 49.5 %, ranging from 25.8 to 78.3 %. For the IMV group, however, only three studies gave a description for ventilation settings, two of which used volume-controlled ventilation, the other used pressure controlled ventilation. Supportive treatments were described in three studies, including use of antimicrobial agents, diuretics, bronchodilators, granulocyte-colony stimulating factor, dialysis, and parenteral nutrition, etc.Fig. 1 Study flow diagram Table 1 Characteristics of the included studies Study Study design Settings Sample size (NIV/IMV) Cause of immunodeficiency SAPS II (SD orrange) Number of NIV patients that switched to IMV (%) Outcomes Azoulay 2001 Cohort, single-centered France, ICU 96 (48/48) Hematologic malignancy or solid tumors NIV: 47 (38–60) Not specified 1. 30-day mortality IMV: 44.5 (36–59) 2. Nosocomial infections Azoulay 2003 Case-control, single-centered France, ICU 15 (7/8) Hematological malignancy NS 4 (57.1) 1. Mortality (in ICU) Azoulay 2004 Case-control, single-centered France, ICU 148 (79/69) Hematological malignancy, allogeneic BMT, solid tumors, chemotherapy NS 45 (57.0) 1. Mortality (in hospital) B-M 2013 Case-control, single-centered Spain, ICU 41 (35/6) Hematological malignancy 63 (18) 14 (40.0) 1. Mortality (in hospital) 2. Mortality (in ICU) 3. Duration of ICU stay 4. Duration of hospitalization 5. Nosocomial infections 6. Duration of mechanical ventilation Confalonieri 2002 Cohort, single-centered Italy, ICU 48 (24/24) AIDS NIV: 37 (9) 8 (33.0) 1. Mortality (in ICU) IMV: 38 (5) 2. Duration of ICU stay 3. Duration of hospitalization 4. Nosocomial infections 5. Duration of mechanical ventilation Depuydt 2004 Cohort, single-centered Belgium, ICU 78 (26/52) Hematological malignancy and allogeneic BMT NIV: 46 18 (69.2) 1. Mortality (in hospital) IMV: 46 Depuydt 2010 Cohort, single-centered Belgium, ICU and general medical units 91 (24/67) Hematological malignancy and allogeneic BMT NIV: 52 (15) 18 (75.0) 1. Mortality (in ICU) IMV: 65 (18) 2. Mortality (in hospital) 3. Duration of ICU stay Gachot 1992 Case-control, single-centered France, ICU 45 (36/9) AIDS NS 11 (30.6) 1. Mortality (in ICU) 2. Mortality (in hospital) Gristina 2011 Case-control, multicenter Italy, ICU 1302 (274/1028) Hematologic malignancy NIV: 49 (16) 126 (46.0) 1. Mortality (in ICU) IMV: 58 (18) 2. Mortality (in hospital) 3. Duration of ICU stay 4. Duration of hospitalization 5. Duration of mechanical ventilation 6. Nosocomial infections Molina 2012 Case-control, multicenter Spain, ICU 300 (131/169) Hematological malignancy and BMT NS 79 (60.3) 1. Mortality (in ICU) Pancera 2008 Case-control, single-centered Italy, PICU 239 (120/119) Hematologic malignancy or solid tumors NS 31 (25.8) 1. Mortality (in ICU) 2. 30-day mortality Rabitsch 2005 Case-control, single-centered Austria, ICU 82 (35/47) Autologous or allogeneic BMT for hematological malignancies NIV: 62 (49–84) 24 (68.6) 1. Mortality (in hospital) IMV: 68 (51–87) Turkoglu 2013 Case-control, single-centered Turkey, ICU 67 (46/21) Hematological malignancies NS 36 (78.3) 1. Mortality (in ICU) Abbreviations: AIDS acquired immune deficiency syndrome; BMT bone marrow transplantation; ICU Intensive Care Unit; PICU Pediatric Intensive Care Unit; USA United States of America; NIV Noninvasive mechanical ventilation; IMV Invasive mechanical ventilation; SAPS II Simplified Acute Physiologic Score II; NS Not stated; SD Standard deviation Five studies were excluded for the fact that there was no or improper invasive mechanical ventilation group set as control group among these studies [38–42]. Another common reason was that invasive mechanical ventilation was not studied as a comparison but rather an outcome of the non-invasive ventilation [43, 44]. Two more studies were excluded since attempts to obtain original data from the author concerning subgroup data were unsuccessful [45, 46]. Two studies were excluded since less than 85 % of participants who were diagnosed as ARF on admission [47, 48]. One was excluded because participants were not restricted within patients with ARF [49]. In addition, baseline of PaO2/FiO2) was higher than 300, which has already exceeded the upper limit of current definition of ARF. One was excluded since it used a different definition of 30-day mortality [50]. Study quality We used the adapted Newcastle Ottawa Scale to assess the risk bias in cohort/case-control studies in our review, as described in the Methods section. The four cohort studies scored between 7 and 8 (out of a 9 points), with one study at low risk of bias and three at medium risk. The other nine case-control studies scored between 3 and 6 (out of 8 points), all assessed as high risk of bias but one as medium (see Additional file 1). We defined main primary outcomes to assess the quality of evidence using the GRADE methodology (Table 2). The main factor that may downgrade the levels of quality was the non-RCT study design in all included studies, which share the inherent limitations of the design and implementation compared to RCTs. Another factor was the inconsistency of results across the small number of included trials. The substantial heterogeneity may be attributed to methodological variations among studies and clinical variations among participants. Although the definition of ARF fulfilled the inclusion criteria, substantial heterogeneity existed in the types of ARF. The paucity of studies made us unable to conduct subgroup analyses or meta regression analyses to explore causes of such heterogeneity. Few studies reported estimates of effect after adjustment by multiple variables. In addition, some included studies had small sample sizes, which made them less representative of the exposed population. There were also factors that upgraded the levels of quality for the outcome (30-day mortality) due to relatively large outcome events (Table 2).Table 2 Summary of main findings Patient or population: Immunocompromised patients with acute respiratory failure Setting: ICUs, General medical units. Intervention: Noninvasive mechanical ventilation Comparison: Invasive mechanical ventilation Outcomes Anticipated absolute effectsa (95 % CI) Relative effect (95 % CI) No. of participants (studies) Quality of the evidence (GRADE) Comments Risk with invasive mechanical ventilation Risk with Noninvasive mechanical ventilation Mortality in hospital 624 per 1000 416 per 1000 (276 to 570) OR 0.43 (0.23 to 0.80) 1787 (7 observational studies) ⨁◯◯◯ VERY LOWb Mortality in hospital- Less severe subgroup 584 per 1000 496 per 1000 (431 to 558) OR 0.70 (0.54 to 0.90) 1380 (2 observational studies) ⨁⨁◯◯ LOW Mortality in ICU 549 per 1000 339 per 1000 (226 to 464) OR 0.42 (0.24 to 0.71) 2148 (9 observational studies) ⨁⨁◯◯ LOWb,c Mortality in ICU- AIDS subgroup 576 per 1000 230 per 1000 (98 to 440) OR 0.22 (0.08 to 0.58) 93 (2 observational studies) ⨁⨁⨁◯ MODERATEc Mortality in ICU-Hematological malignancy and BMT subgroup 543 per 1000 443 per 1000 (348 to 543) OR 0.67 (0.45 to 1.00) 1816 (6 observational studies) ⨁◯◯◯ VERY LOWd Mortality in ICU Hematological malignancy and solid tumors subgroup 613 per 1000 222 per 1000 (137 to 337) OR 0.18 (0.10 to 0.32) 239 (1 observational study) ⨁⨁◯◯ LOW 30-day mortality 749 per 1000 503 per 1000 (396 to 616) OR 0.34 (0.22 to 0.54) 335 (2 observational studies) ⨁⨁⨁◯ MODERATEc GRADE Working Group grades of evidence High quality: We are very confident that the true effect lies close to that of the estimate of the effect Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect CI confidence interval; OR odds ratio; MD mean difference aThe risk in the intervention group (and its 95 % confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95 % CI) bSubstantial heterogeneity may be due to methodological variations among studies and clinical variations among participants cUpgraded due to large sample size and/or large outcome events d95 % confidence interval up to 1 Primary outcomes We found a significantly lower mortality in hospital in NIV compared to IMV (OR 0.43, 95 % CI 0.23 to 0.80, P value = 0.007), but the heterogeneity was substantial (I2 statistic = 62 %). Subgroup analysis of less severe group showed similar result (OR 0.70, 95%CI 0.54 to 0.90, P value = 0.007; I2statistic = 0 %, Fig. 2). One study (Depuydt 2010) [34], with overlapped mean of SAPS II (52–65), was allocated in the more severe subgroup due to a substantially lower mean value of PaO2/FiO2 and relatively higher SOFA score than other studies. In AIDS subgroup, only one study was included, which showed NIV significantly reduced the mortality in hospital (OR 0.17, 95 % CI 0.03 to 0.81, P value = 0.03). In addition, multivariate logistic regression analysis reported in one study (Gristina 2011) [17], after adjustments for all available demographic characteristics and clinical variables, an initial NIV was associated with a significantly lower mortality in hospital (OR 0.73, 95 % CI 0.53 to 1.00, P value = 0.05).Fig. 2 Mortality in hospital by disease severity. CI confidence interval, I2 percentage of total variation across studies from between-study heterogeneity rather than chance. Vertical solid line null effect, Vertical dotted line overall effect As for mortality in ICU, meta-analysis was not possible in both overall and subgroup analysis of disease severity due to substantial heterogeneity (overall analysis: I2 statistic = 72 %). Subgroup analysis of causes of immunodeficiency showed that NIV was associated with a significant reduction of mortality in ICU in all subgroups, AIDS subgroup (OR 0.20, 95 % CI 0.07 to 0.54, P value = 0.001, I2 statistic = 0 %, Fig. 3), hematological malignancy and BMT subgroup (OR 0.67, 95 % CI 0.45 to 1.00, P value = 0.05, I2 statistic = 34 %), and hematological malignancy and solid tumors subgroup (OR 0.18, 95 % CI 0.10 to 0.32, P value < 0.00001).Fig. 3 Mortality in ICU by cause of immunodeficiency The meta-analysis of two studies showed a statistically significant reduction of 30-day mortality between NIV versus IMV (OR 0.34, 95 % CI 0.22 to 0.54, P value < 0.0001, I2 statistic = 0 %, Fig. 4). Subgroup analysis of disease severity was not possible due to lacking of SAPS II data. Both studies were in the haematological malignancy and solid tumors subgroup, thus, significantly favour NIV. One study (Azoulay 2001) [28] reported a multivariate logistic regression analysis, showing that the probability of 30-day mortality was significantly decreased in NIV compared with IMV, with an odds ratio of 0.31 (95 % CI 0.12 to 0.82).Fig. 4 30-day mortality Secondary outcomes Four studies revealed no significant difference in the rate of nosocomial infections in patients on NIV in comparison with IMV. Similar results were shown in subgroup analysis of disease severity. Significant difference was found only in the hematological malignancy and BMT subgroup favoring NIV (OR 0.58, 95 % CI 0.36 to 0.93, P value = 0.03, I2 statistic = 0 %, Fig. 5).Fig. 5 Nosocomial infections by cause of immunodeficiency Duration of mechanical ventilation was reported in three studies, but data couldn’t be pooled in the meta-analysis due to substantial heterogeneity (overall I2 statistic = 98 %). One study (Confalonieri 2002) [20] reported a statistically significant shortening of mechanical ventilation favoring NIV versus IMV (MD ± SD: 6 ± 2 versus 7 ± 1 days, P value = 0.034). Another (Gristina 2011) [17] reported that the mean duration of mechanical ventilation was 4 days (SD 4 days) in NIV group, compared to 11 days (SD 4 days) in IMV group (P value < 0.0001). The rest (B-M 2013) [36] revealed no difference among the two groups (P value = 0.08). Duration of ICU stay was reported in four studies. Overall and subgroup analyses of cause of immunodeficiency were not possible due to substantial heterogeneity (overall: I2 statistic = 62 %). In less severe subgroup, we found a statistically significant shortening of ICU stay in favor of NIV (MD −3.00 days; 95 % CI −4.27, −1.73 days, P value < 0.00001, I2 statistic = 0 %, Fig. 6).Fig. 6 Duration of ICU stay by disease severity. SD standard deviation Duration of hospitalization could only be pooled in the subgroup analysis of cause of immunodeficiency. One study (Confalonieri 2002) [20] in AIDS subgroup reported a statistically significant shortening of mechanical ventilation favoring NIV versus IMV (MD ± SD: 13 ± 5 versus 24 ± 17 days, P value = 0.004, Fig. 7).Fig. 7 Duration of hospitalization by cause of immunodeficiency Discussion To our knowledge this is the first systematic review aiming at compilation of the clinical evidence of the effect of NIV compared to IMV on ARF in immunocompromised patients. The overall evidence we obtained supports NIV over IMV in treating ARF in certain group of immunocompromised patients. Compared to IMV, NIV was shown to significantly reduce mortality in overall analysis, and mortality and duration of hospitalization/ICU stay in less severe (mainly reflected by SAPS II < 60), AIDS, hematological malignancy subgroups. Compared with IMV, NIV is more frequently associated with minor complications because NIV avoids endotracheal tube or tracheostomy hence leaves the upper airway intact, and preserves airway defense mechanisms. Various complications directly related to the process of intubation and IMV such as nosocomial pneumonia, aspiration of gastric contents, ventilator-associated events, trauma of the teeth, hypopharynx, esophagus, larynx, and trachea could thus be avoided by application of NIV [19, 51, 52]. These complications are far more common in patients who are immunocompromised or critically ill. In addition, NIV causes less mask-related discomfort, unrecognized patient-ventilator asynchrony due to leaks, and milder gastric insufflations. At the meantime, NIV achieves the same physiological benefits of reduced work of breathing and improved gas exchange for certain groups of patients [53]. Although NIV was recommended as the first-line strategy for immunocompromised patients with ARF by guidelines in several countries [11, 54], our study along with several previous studies showed that NIV might not be the appropriate choice for all immunocompromised patients [15–17]. NIV has been stated by German and Canadian guidelines as the first-line treatment, with different recommendation levels, for immunocompromised patients with ARF [11, 54], though they were based on the same two RCTs [41, 55]. However, the control against NIV in both RCTs was standard oxygen therapy instead of IMV. In addition, participants in both RCTs were less severe than those in our review. The mean SAPS II in Hilbert 2001 [41] was no more than 45, and mean SAPS in Antonelli 2000 [55] was 13. Our result, along with several other reviews raised new puzzles and debates for NIV as the first-line approach for immunocompromised patients with ARF [16, 56]. We make the speculation that spectrum of choice for oxygen supplement strategy varies among standard oxygen therapy, NIV or IMV depending on different disease severity, and NIV or standard oxygen therapy should be applied in less severe patients. For the more severe patients, however, the choice might be limited within NIV or IMV. Our review showed that even in relatively more severe patients (45 < SAPS II < 60) than those in the two RCTs, NIV still showed significant advantages against IMV. As is shown in our analysis, effects of NIV on patients vary with different levels of disease severity, different causes of immunocompromised status and types of ARF. Strict patient selection is of critical importance for the effect of NIV and other oxygen therapies. As for disease severity, our analysis showed NIV had clear advantage in less severe patients (mainly reflected by SAPS II < 60). For cause of immunodeficiency, NIV showed great advantage over IMV among AIDS patients in reducing mortality in both hospital and ICU, duration of both hospital and ICU stay, and also duration of mechanical ventilation. NIV was also related to better prognosis in hematological malignancies and BMT patients. Data available, though weak, also favored NIV in patients with solid tumors. One systematic review by Laura et al. [57] included thirteen studies to examine the effect of initial NIV versus IMV in hematological patients with ARF. Eight studies of which were also included in our review. NIV is associated with a lower risk of death in hematological patients (RR 0.74, 95 % CI 0.65 to 0.84, p < 0.0001), which was similar with our results. Unfortunately, there was little evidence showing the effects of NIV on different types of ARF due to lacking of data. Two studies [17, 38] showed poor prognosis of NIV in ALI/ARDS. NIV was strongly recommended as the first-line approach for acute exacerbation of chronic obstructive pulmonary disease (AECOPD), facilitation of weaning/extubation in patients with COPD and cardiogenic pulmonary edema in immunocompetent patients [11, 54]. The criteria between immunocompetent and immunocompromised patients are not identical, but similar. We expect similar efficacy of NIV in immunocompromised patients with these types of ARF. In addition, another review [58] emphasized the importance of patient selection and suggested both absolute and relative situations in which NIV is contraindicated. However, additional studies, especially randomized controlled trials, are needed to explore the possible benefits and/or risks of NIV in these groups of patients. Considering that our analysis showed NIV reduced mortality in less severe patients, early recognition and prediction of NIV failure and timely initiation of IMV are of great importance for patients’ overall survival, especially for more severe patients. Laura et al. [57] showed that failure of NIV might worsen the prognosis, mainly in less severe patients. Predictors of failure of NIV in immunocompromised patients were summarised in one review [16], including higher illness severity at baseline reflected by SAPS II, higher respiratory rate under NIV, later initiation of NIV after ICU admission, need for vasopressors, renal replacement therapy, and presence of ALI/ARDS. Criteria used for NIV discontinuation and endotracheal intubation in immunocompromised patients were also suggested, including persistent dyspnea, severe hemodynamic or electrocardiographic instability, etc. The cut-off point for switching NIV to IMV in immunocompromised patients is still a puzzling issue which calls for more high-qualified studies [59]. The presence of patients who initially received NIV switched to IMV due to NIV failure did not substantially change our overall results and final conclusion. Almost all included studies (12 out of 13, excluding one study where the rate was not specified) reported such switching to IMV after initial NIV treatment. The mean intubation rate after initial NIV treatment was 49.5 %, greater than or equal to 50 % in seven studies and lower than 50 % in five studies (Table 1). We did a subgroup meta-analysis based on intubation rate in NIV group (Additional file 1), which showed that when intubation rate was lower than 50 %, NIV was more favorable in terms of mortality in ICU, 30-day mortality, duration of ICU stay as well as nosocomial infections. No significant difference was found between the NIV and IMV groups in terms of mortality in hospital, mortality in ICU and duration of ICU stay when intubation rate was greater than or equal to 50 %. It is thus reasonable to deduct that the final analysis results would be in even more favour of NIV if the overall intubation rate of all the studies included are lower than 50 %. In addition, a research where NIV patients are kept from switching to IMV once their conditions worsen would be unethical and in turn, impractical. The major limitation of the present systematic review was that all included studies were exclusively observational. Given to the limitations of current data, RCTs of good methodological design are needed to address the effect of NIV versus IMV in treating acute respiratory failure in immunocompromised patients. Researchers should consider RCTs with a sample size large enough to demonstrate a meaningful result. Although it would be impossible to conduct double-blinded trials in the future due to the nature of intervention, it would be important to undertake blind assessment of the participants to ensure quality of the trials and minimize the risk of bias, with well classified causes of ARF and immunocompromised status and well stratified disease severity. Settings of intervention such as time between onset of acute respiratory failure and ICU admission, clear indications and contra-indications for NIV and or high flow oxygen therapy, type of interface and equipment use, strict NIV and high flow oxygen use protocol, early recognition of NIV or high flow oxygen failure, clear indications of intubation and invasive protective mechanical ventilation should also be well matched [56]. In place of RCTs, we suggest well-conducted observational studies, such as strictly matched prospective cohort studies. Conclusions The overall evidence we obtained from 2552 immunocompromised patients with ARF shows NIV was associated with a significant lower mortality rates, especially in less severe patients and those who was immunosuppressed by AIDS, haematological malignancies and bone marrow transplant. For more severe patients, however, NIV didn’t show clear advantages over IMV. The advantages of NIV were also shown in reducing duration of hospitalization and ICU stay, as well as rates of nosocomial infection. Future studies need to be methodologically sound and include patients immunocompromised by other causes such as chemotherapy and receipt of glucocorticoids. Additional file Additional file 1: Supplement 1. Risk of bias graph: review authors’ judgements about each risk of bias item presented as percentages across all included studies. Supplement 2. Risk of bias summary: review authors’ judgements about each risk of bias item for each included study. Supplement 3. CENTRAL search strategy. Supplement 4. PubMed search strategy. Supplement 5. EMBASE search strategy. Supplement 6. CBM search strategy. Supplement 7. Assessment of risk of bias in cohort studies. Supplement 8. Assessment of risk of bias in case-control studies. Supplement 9. Newcastle-Ottawa Grading Results. Supplement 10. Subgroup meta-analysis based on intubation rate in NIV group. Supplement 11. Mortality in hospital by intubation rate. Supplement 12. Mortality in ICU by intubation rate. Supplement 13. 30-day mortality by intubation rate. Supplement 14. Duration of hospitalization by intubation rate. Supplement 15. Duration of ICU stay by intubation rate. Supplement 16. Nosocomial infection by intubation rate. Supplement 17. Duration of mechanical ventilaiton by intubation rate. (DOC 194 kb) Abbreviations 95 % CIs95 % confidence intervals AECOPDAcute exacerbation of chronic obstructive pulmonary disease AIDSAcquired immune deficiency syndrome APACHE IIAcute physiology and chronic health evaluation II scores ARDS/ALIAdult respiratory distress syndrome/acute lung injury ARFAcute respiratory failure Bi-PAPBi­level positive airway pressure BMTBone marrow transplantation CPAPContinuous positive airway pressure GRADEThe grading of recommendations assessment, development and evaluation ICUIntensive care unit IMVInvasive mechanical ventilation MDMean difference NIVNoninvasive mechanical ventilation OROdds ratio PaO2/FiO2Ratio of the partial pressure of arterial oxygen to the fraction of inspired oxygen PEEPPositive end expiratory pressure PSVPressure support ventilation RCTsRandomized controlled trials SAPS IISimplified acute physiology scores II SDStandard deviation SOFASequential organ failure assessment Acknowledgements The authors would like to thank Dr. Pieter O. Depuydt, Dr. Alberto Belenguer and Dr. Eric M. Mortensen for providing relevant first-hand information regarding their papers. Funding Not applicable. Availability of data and material The datasets analysed during the current study are available at https://osf.io/b74wr/. Authors’ contributions YL, TW and LZ were responsible for setting up the initial idea for writing the paper, JH and YM collected the data regarding the paper, TW, LZ, NZ, KL, and QX analyzed the data, YL and TW wrote the original paper in English, YL, ZL and XY made revisions, worked on the language, and finally made the final version of the manuscript, which was reviewed by all the authors. LZ and TW were considered to equally contribute to this paper. All authors read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. Consent for publication Not applicable. Ethics approval and consent to participate Not applicable. ==== Refs References 1. Hayashi Y, Paterson DL. Infections in the Immunocompromised Patient. In: Textbook of Critical Care. edn. sixth: Saunders. Elsevier; 2011. p. 1040-46. https://www.clinicalkey.com/#!/content/book/3-s2.0-B9781437713671001373. 2. Linden PK Approach to the immunocompromised host with infection in the intensive care unit Infect Dis Clin North Am 2009 23 3 535 56 10.1016/j.idc.2009.04.014 19665082 3. Brochard L Isabey D Piquet J Amaro P Mancebo J Messadi AA Brun-Buisson C Rauss A Lemaire F Harf A Reversal of acute exacerbations of chronic obstructive lung disease by inspiratory assistance with a face mask N Engl J Med 1990 323 22 1523 30 10.1056/NEJM199011293232204 2122253 4. Boldrini R Fasano L Nava S Noninvasive mechanical ventilation Curr Opin Crit Care 2012 18 1 48 53 10.1097/MCC.0b013e32834ebd71 22186215 5. Olivieri C Carenzo L Vignazia GL Campanini M Pirisi M Della Corte F Does noninvasive ventilation delivery in the ward provide early effective ventilation? Respir Care 2015 60 1 6 11 10.4187/respcare.03294 25352686 6. Antonelli M Conti G Rocco M Bufi M De Blasi RA Vivino G Gasparetto A Meduri GU A comparison of noninvasive positive-pressure ventilation and conventional mechanical ventilation in patients with acute respiratory failure N Engl J Med 1998 339 7 429 35 10.1056/NEJM199808133390703 9700176 7. Lightowler JV Wedzicha JA Elliott MW Ram FS Non-invasive positive pressure ventilation to treat respiratory failure resulting from exacerbations of chronic obstructive pulmonary disease: Cochrane systematic review and meta-analysis BMJ (Clinical research ed) 2003 326 7382 185 10.1136/bmj.326.7382.185 8. Lindenauer PK Stefan MS Shieh MS Pekow PS Rothberg MB Hill NS Outcomes associated with invasive and noninvasive ventilation among patients hospitalized with exacerbations of chronic obstructive pulmonary disease JAMA Intern Med 2014 174 12 1982 93 10.1001/jamainternmed.2014.5430 25347545 9. Plant PK Owen JL Elliott MW Early use of non-invasive ventilation for acute exacerbations of chronic obstructive pulmonary disease on general respiratory wards: a multicentre randomised controlled trial Lancet (London, England) 2000 355 9219 1931 5 10.1016/S0140-6736(00)02323-0 10. Vital FM Ladeira MT Atallah AN Non-invasive positive pressure ventilation (CPAP or bilevel NPPV) for cardiogenic pulmonary oedema Cochrane Database Syst Rev 2013 5 CD005351 23728654 11. Keenan SP Sinuff T Burns KE Muscedere J Kutsogiannis J Mehta S Clinical practice guidelines for the use of noninvasive positive-pressure ventilation and noninvasive continuous positive airway pressure in the acute care setting CMAJ 2011 183 3 E195 214 10.1503/cmaj.100071 21324867 12. Wang S Singh B Tian L Biehl M Krastev IL Kojicic M Li G Epidemiology of noninvasive mechanical ventilation in acute respiratory failure--a retrospective population-based study BMC Emerg Med 2013 13 6 10.1186/1471-227X-13-6 23570601 13. Antonelli M Conti G Esquinas A Montini L Maggiore SM Bello G A multiple-center survey on the use in clinical practice of noninvasive ventilation as a first-line intervention for acute respiratory distress syndrome Crit Care Med 2007 35 1 18 25 10.1097/01.CCM.0000251821.44259.F3 17133177 14. Ferrer M Esquinas A Leon M Gonzalez G Alarcon A Torres A Noninvasive ventilation in severe hypoxemic respiratory failure: a randomized clinical trial Am J Respir Crit Care Med 2003 168 12 1438 44 10.1164/rccm.200301-072OC 14500259 15. Lemiale V Mokart D Resche-Rigon M Pene F Mayaux J Faucher E Effect of Noninvasive Ventilation vs Oxygen Therapy on Mortality Among Immunocompromised Patients With Acute Respiratory Failure: A Randomized Clinical Trial JAMA 2015 314 1711 9 10.1001/jama.2015.12402 26444879 16. Bello G De Pascale G Antonelli M Noninvasive ventilation for the immunocompromised patient: always appropriate? Curr Opin Crit Care 2012 18 54 60 10.1097/MCC.0b013e32834e7c21 22143051 17. Gristina GR Antonelli M Conti G Ciarlone A Rogante S Rossi C Bertolini G GiViTi Noninvasive versus invasive ventilation for acute respiratory failure in patients with hematologic malignancies: a 5-year multicenter observational survey Crit Care Med 2011 39 10 2232 9 10.1097/CCM.0b013e3182227a27 21666446 18. Bach JR Intintola P Alba AS Holland IE The ventilator-assisted individual. Cost analysis of institutionalization vs rehabilitation and in-home management Chest 1992 101 1 26 30 10.1378/chest.101.1.26 1729079 19. Nava S Evangelisti I Rampulla C Compagnoni ML Fracchia C Rubini F Human and financial costs of noninvasive mechanical ventilation in patients affected by COPD and acute respiratory failure Chest 1997 111 6 1631 8 10.1378/chest.111.6.1631 9187186 20. Confalonieri M Calderini E Terraciano S Chidini G Celeste E Puccio G Gregoretti C Meduri GU Noninvasive ventilation for treating acute respiratory failure in AIDS patients with Pneumocystis carinii pneumonia Intensive Care Med 2002 28 9 1233 8 10.1007/s00134-002-1395-2 12209270 21. Le Gall JR Lemeshow S Saulnier F A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study Jama 1993 270 24 2957 63 10.1001/jama.1993.03510240069035 8254858 22. Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from: http://handbook.cochrane.org/. 23. Wells GA, Shea B, O’Connell D, Peterson J, WelchV, Losos M, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. In: Available from: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp; 2010. 24. Schünemann H, Brożek J, Guyatt G, Oxman A. GRADE handbook for grading quality of evidence and strength of recommendations. Updated October 2013: The GRADE Working Group, 2013. Available from: http://gdt.guidelinedevelopment.org/app/handbook/handbook.html. 25. GRADEpro Guideline Development Tool [Software]. In: McMaster University, 2015 (developed by Evidence Prime, Inc) Available from: https://gradepro.org/. Accessed 24 July 2016. 26. Minne L Abu-Hanna A de Jonge E Evaluation of SOFA-based models for predicting mortality in the ICU: A systematic review Crit Care 2008 12 6 R161 10.1186/cc7160 19091120 27. Gachot B Clair B Wolff M Regnier B Vachon F Continuous positive airway pressure by face mask or mechanical ventilation in patients with human immunodeficiency virus infection and severe Pneumocystis carinii pneumonia Intensive Care Med 1992 18 3 155 9 10.1007/BF01709239 1644963 28. Azoulay E Alberti C Bornstain C Leleu G Moreau D Recher C Chevret S Le Gall JR Brochard L Schlemmer B Improved survival in cancer patients requiring mechanical ventilatory support: impact of noninvasive mechanical ventilatory support Crit Care Med 2001 29 3 519 25 10.1097/00003246-200103000-00009 11373414 29. Azoulay E Fieux F Moreau D Thiery G Rousselot P Parrot A Le Gall JR Dombret H Schlemmer B Acute monocytic leukemia presenting as acute respiratory failure Am J Respir Crit Care Med 2003 167 10 1329 33 10.1164/rccm.200206-554OC 12574074 30. Azoulay E Thiery G Chevret S Moreau D Darmon M Bergeron A Yang K Meignin V Ciroldi M Le Gall JR The prognosis of acute respiratory failure in critically ill cancer patients Medicine 2004 83 6 360 70 10.1097/01.md.0000145370.63676.fb 15525848 31. Depuydt PO Benoit DD Vandewoude KH Decruyenaere JM Colardyn FA Outcome in noninvasively and invasively ventilated hematologic patients with acute respiratory failure Chest 2004 126 4 1299 306 10.1378/chest.126.4.1299 15486396 32. Rabitsch W Staudinger T Locker GJ Kostler WJ Laczika K Frass M Knoebl P Greinix HT Kalhs P Keil F Respiratory failure after stem cell transplantation: improved outcome with non-invasive ventilation Leuk Lymphoma 2005 46 8 1151 7 10.1080/10428190500097649 16085555 33. Pancera CF Hayashi M Fregnani JH Negri EM Deheinzelin D de Camargo B Noninvasive ventilation in immunocompromised pediatric patients: eight years of experience in a pediatric oncology intensive care unit J Pediatr Hematol Oncol 2008 30 7 533 8 10.1097/MPH.0b013e3181754198 18797201 34. Depuydt PO Benoit DD Roosens CD Offner FC Noens LA Decruyenaere JM The impact of the initial ventilatory strategy on survival in hematological patients with acute hypoxemic respiratory failure J Crit Care 2010 25 1 30 6 10.1016/j.jcrc.2009.02.016 19682849 35. Molina R Bernal T Borges M Zaragoza R Bonastre J Granada RM Rodriguez-Borregan JC Nunez K Seijas I Ayestaran I Ventilatory support in critically ill hematology patients with respiratory failure Crit Care 2012 16 4 R133 10.1186/cc11438 22827955 36. Belenguer-Muncharaz A Albert-Rodrigo L Ferrandiz-Selles A Cebrian-Graullera G Ten-year evolution of mechanical ventilation in acute respiratory failure in the hematogical patient admitted to the intensive care unit Med Intensiva 2013 37 7 452 60 10.1016/j.medin.2012.12.011 23890541 37. Turkoglu M Erdem GU Suyani E Sancar ME Yalcin MM Aygencel G Aki Z Sucak G Acute respiratory distress syndrome in patients with hematological malignancies Hematology 2013 18 3 123 30 10.1179/1607845412Y.0000000038 23321772 38. Adda M Coquet I Darmon M Thiery G Schlemmer B Azoulay E Predictors of noninvasive ventilation failure in patients with hematologic malignancy and acute respiratory failure Crit Care Med 2008 36 10 2766 72 10.1097/CCM.0b013e31818699f6 18766110 39. Chin K Uemoto S Takahashi K Egawa H Kasahara M Fujimoto Y Noninvasive ventilation for pediatric patients including those under 1-year-old undergoing liver transplantation Liver Transpl 2005 11 188 95 10.1002/lt.20297 15666379 40. Hilbert G Gruson D Vargas F Valentino R Chene G Boiron JM Pigneux A Reiffers J Gbikpi-Benissan G Cardinaud JP Noninvasive continuous positive airway pressure in neutropenic patients with acute respiratory failure requiring intensive care unit admission Crit Care Med 2000 28 9 3185 90 10.1097/00003246-200009000-00012 11008980 41. Hilbert G Gruson D Vargas F Valentino R Gbikpi-Benissan G Dupon M Reiffers J Cardinaud JP Noninvasive ventilation in immunosuppressed patients with pulmonary infiltrates, fever, and acute respiratory failure N Engl J Med 2001 344 7 481 7 10.1056/NEJM200102153440703 11172189 42. Piastra M De Luca D Pietrini D Pulitano S D’Arrigo S Mancino A Conti G Noninvasive pressure-support ventilation in immunocompromised children with ARDS: a feasibility study Intensive Care Med 2009 35 8 1420 7 10.1007/s00134-009-1558-5 19547952 43. Rocco M Conti G Antonelli M Bufi M Costa MG Alampi D Ruberto F Stazi GV Pietropaoli P Non-invasive pressure support ventilation in patients with acute respiratory failure after bilateral lung transplantation Intensive Care Med 2001 27 10 1622 6 10.1007/s001340101063 11685303 44. Yu H Ping G Zhengxian C Noninvasive versus invasive mechanical ventilation in severe pneumonia patients after kidney transplant J Pract Med 2009 18 3108 9 45. Festic E Gajic O Limper AH Aksamit TR Acute respiratory failure due to pneumocystis pneumonia in patients without human immunodeficiency virus infection: outcome and associated features Chest 2005 128 573 9 10.1378/chest.128.2.573 16100140 46. Schnell D Timsit JF Darmon M Vesin A Goldgran-Toledano D Dumenil AS Noninvasive mechanical ventilation in acute respiratory failure: trends in use and outcomes Intensive Care Med 2014 40 582 91 10.1007/s00134-014-3222-y 24504643 47. Grgic Medic M Gornik I Gasparovic V Hematologic malignancies in the medical intensive care unit--Outcomes and prognostic factors Hematology (Amsterdam, Netherlands) 2015 20 5 247 53 48. Rabbat A Chaoui D Montani D Legrand O Lefebvre A Rio B Roche N Lorut C Marie JP Huchon G Prognosis of patients with acute myeloid leukaemia admitted to intensive care Br J Haematol 2005 129 3 350 7 10.1111/j.1365-2141.2005.05459.x 15842658 49. Azevedo LC Caruso P Silva UV Torelly AP Silva E Rezende E Netto JJ Piras C Lobo SM Knibel MF Outcomes for patients with cancer admitted to the ICU requiring ventilatory support: results from a prospective multicenter study Chest 2014 146 2 257 66 10.1378/chest.13-1870 24480886 50. Johnson CS Frei CR Metersky ML Anzueto AR Mortensen EM Non-invasive mechanical ventilation and mortality in elderly immunocompromised patients hospitalized with pneumonia: a retrospective cohort study BMC Pulm Med 2014 14 7 10.1186/1471-2466-14-7 24468062 51. Hess DR Noninvasive ventilation for acute respiratory failure Respir Care 2013 58 6 950 72 10.4187/respcare.02319 23709194 52. Namendys-Silva SA Hernandez-Garay M Herrera-Gomez A Noninvasive ventilation in immunosuppressed patients Am J Hosp Palliat Care 2010 27 2 134 8 10.1177/1049909109346833 19776370 53. Vitacca M Ambrosino N Clini E Porta R Rampulla C Lanini B Physiological response to pressure support ventilation delivered before and after extubation in patients not capable of totally spontaneous autonomous breathing Am J Respir Crit Care Med 2001 164 4 638 41 10.1164/ajrccm.164.4.2010046 11520729 54. Schonhofer B Kuhlen R Neumann P Westhoff M Berndt C Sitter H Clinical practice guideline: non-invasive mechanical ventilation as treatment of acute respiratory failure Deutsches Arzteblatt Int 2008 105 24 424 33 55. Antonelli M Conti G Bufi M Costa MG Lappa A Rocco M Gasparetto A Meduri GU Noninvasive ventilation for treatment of acute respiratory failure in patients undergoing solid organ transplantation: a randomized trial Jama 2000 283 2 235 41 10.1001/jama.283.2.235 10634340 56. Barbas CS Serpa Neto A New puzzles for the use of non-invasive ventilation for immunosuppressed patients J Thorac Dis 2016 8 1 E100 3 26904233 57. Amado-Rodriguez L, Bernal T, Lopez-Alonso I, Blazquez-Prieto J, Garcia-Prieto E, Albaiceta GM. Impact of Initial Ventilatory Strategy in Hematological Patients With Acute Respiratory Failure: A Systematic Review and Meta-Analysis. Crit Care Med. 2016;44(7):1406–1413. 58. Nava S Hill N Non-invasive ventilation in acute respiratory failure Lancet (London, England) 2009 374 9685 250 9 10.1016/S0140-6736(09)60496-7 59. Quintin L Cut-off point for switching from non- -invasive ventilation to intubation in severe ARDS. Fifty shades of grey? Anaesthesiol Intensive Ther 2016 48 1 62 4 10.5603/AIT.2016.0011 26966115

PMC005xxxxxx/PMC5002327.txt

==== Front BMC Fam PractBMC Fam PractBMC Family Practice1471-2296BioMed Central London 52610.1186/s12875-016-0526-8Research ArticleCommunity health center provider ability to identify, treat and account for the social determinants of health: a card study Lewis Joy H. JHLewis@atsu.edu 1Whelihan Kate kwhelihan@atsu.edu 1Navarro Isaac inavarro@atsu.edu 1Boyle Kimberly R. krbclinicalsolutions@gmail.com 2SDH Card Study Implementation Team 1 A.T. Still University of Health Sciences School of Osteopathic Medicine in Arizona (ATSU SOMA), 5850 E. Still Circle, Mesa, AZ 85206 USA 2 KRB Clinical Solutions LLC, 24 Underwood Road, Montville, NJ 07045 USA 27 8 2016 27 8 2016 2016 17 1 12120 4 2016 25 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background The social determinants of health (SDH) are conditions that shape the overall health of an individual on a continuous basis. As momentum for addressing social factors in primary care settings grows, provider ability to identify, treat and assess these factors remains unknown. Community health centers care for over 20-million of America’s highest risk populations. This study at three centers evaluates provider ability to identify, treat and code for the SDH. Methods Investigators utilized a pre-study survey and a card study design to obtain evidence from the point of care. The survey assessed providers’ perceptions of the SDH and their ability to address them. Then providers filled out one anonymous card per patient on four assigned days over a 4-week period, documenting social factors observed during encounters. The cards allowed providers to indicate if they were able to: provide counseling or other interventions, enter a diagnosis code and enter a billing code for identified factors. Results The results of the survey indicate providers were familiar with the SDH and were comfortable identifying social factors at the point of care. A total of 747 cards were completed. 1584 factors were identified and 31 % were reported as having a service provided. However, only 1.2 % of factors were associated with a billing code and 6.8 % received a diagnosis code. Conclusions An obvious discrepancy exists between the number of identifiable social factors, provider ability to address them and documentation with billing and diagnosis codes. This disparity could be related to provider inability to code for social factors and bill for related time and services. Health care organizations should seek to implement procedures to document and monitor social factors and actions taken to address them. Results of this study suggest simple methods of identification may be sufficient. The addition of searchable codes and reimbursements may improve the way social factors are addressed for individuals and populations. Electronic supplementary material The online version of this article (doi:10.1186/s12875-016-0526-8) contains supplementary material, which is available to authorized users. Keywords Social determinants of healthPrimary careHealth care providersCommunity Health CentersElectronic Health RecordDiagnosis codesProcedure codesICD-10issue-copyright-statement© The Author(s) 2016 ==== Body Background Public health officials have long recognized social and environmental factors play a role in impacting an individual’s health. In fact, the social determinants of health (SDH) have been an ever present theme in the works and initiatives set forth by the World Health Organization (WHO) since its inception in 1948 [1]. Defined by the WHO as “the conditions in which people are born, grow, live, work, and age” [2] the social determinants of health are conditions that shape the overall health of an individual on a continuous basis. A significant upswing in momentum for SDH focus was the development of the WHO’s Commission on Social Determinants of Health on March 18, 2005 [3]. The commission expands on the WHO definition of SDH and states that “avoidable health inequalities arise because of the circumstances in which people grow, live, work, and age, AND the systems put in place to deal with illness” (emphasis added) [4]. Identifying social factors and relating them to health conditions for individuals and for populations is of the utmost importance. This importance has been further emphasized by the Institute of Medicine (IOM) with the development of a framework for educating health professionals to address the social determinants of health [5]. Despite the increased focus on SDH, the US suffers from relatively poor health status and worse health outcomes when compared to other high-income nations. SDH can both directly affect health and limit the ability of individuals to make healthy choices [6]. SDH are commonly considered factors associated with economic stability, neighborhood and physical environment, education, food and nutrition, social contexts and health care access [7]. Specific examples of SDH include substandard education, food insecurity, unstable housing and unsafe neighborhoods, environmental conditions and low income [8, 9]. The evidence base for the impact of these factors on health outcomes is growing. Chronic diseases such as diabetes, cardiovascular disease and asthma can be linked to social determinants such as low income and education level [10]. Food insecurity puts people at an increased risk for hypertension, hyperlipidemia and diabetes [11, 12]. The “American Health Care Paradox” - an unfavorable balance between health care costs and health of the population [13] - may be evidence of the lack of attention health care providers give to SDH, despite increasing evidence which suggests that SDH impact health conditions more than the care provided in medical offices [13]. Attention is spent at the point of care to evaluate and motivate patients towards positive health behaviors. However, without addressing social factors, it is difficult to affect positive change. In 2008, Forde and Raine mention, “social factors such as poverty and its sequelae substantially affect people’s abilities to adopt healthy behaviors” [14]. The existence of SDH and the health inequities they create are estimated to be responsible for 30 % of direct medical excess costs for Blacks, Hispanics and Asian Americans [15]. In an increasingly diverse population, efforts to address the SDH are essential to improving quality of care and health outcomes [15]. A study by Milstein, et al. points out that only when a multi-layered approach to health care improvement is completed will the result of better outcomes along with lower cost be achieved [16]. The implementation of the Patient Protection and Affordable Care Act (ACA) will allow for many factors impacting the overall health of the US population to be addressed. The ACA will allow for widespread dissemination of preventive practices that should take SDH into account. The Prevention and Public Health Fund will assure preventive practices related to housing, transportation, food security, and environment while the expansion of Medicaid and the advent of the Health Exchange Marketplace will help assure widespread access to necessary health care delivery [17]. This multifactorial approach has the potential to make a profound impact on the overall health status of the average US citizen. To address the increased need for SDH services in health care, the Institute of Medicine’s Committee on the Recommended Social and Behavioral Measures for Electronic Health Records (EHR) was convened to develop measures for SDH that could be incorporated into EHRs [18]. They were also tasked with minimizing the barriers to the use of these measures [18]. The committee recognized the importance of addressing SDH in clinical practice and the need for an efficient, systematic way to do so. They prioritized SDH with the “greatest clinical usefulness and feasibility for capture in the clinical workflow” [10]. The panel of SDH measures they developed is intended to be brief and readily adoptable into any EHR system to allow clinicians everywhere to incorporate these measures into the treatment of all patients [10]. The ability to treat the SDH is particularly important for providers at Community Health Centers (CHCs). In the United States, CHCs provide primary health care services for nearly 22 million of the nation’s poorest and highest risk individuals [18–20]. CHCs are primary health care organizations typically located in designated Medically Underserved Areas. Governed by a board of directors consisting of local community members with a makeup of at least 51 % health center patients, these organizations provide comprehensive services including enabling services such as education, translation, and transportation. CHCs are ‘safety net’ organizations; as such they provide services to all community members regardless of ability to pay [21]. Over 70 % of health center patients have a family income at or below poverty. A reported 38 % of health center patients are uninsured and 39 % have Medicaid as their insurance [22]. CHCs also serve many who would likely not have other access to care such as rural populations, migrant workers and the homeless [23]. Most CHCs have programs focused on helping patients with some aspects of the SDH, but it is not clear if all CHC primary care providers understand the relationship among the programs, SDH and patient health [24]. Most providers likely recognize social factors but they do not necessarily understand the relationship between these factors and health outcomes. Providers may not be confident regarding how to address the SDH [25]. Additionally, it is not clear if CHC providers are able to code for their provision of services to address specific SDH or if they are able to utilize diagnosis codes for identified SDH. With the current atmosphere related to the expansion of health care coverage under the ACA and with new opportunities for CHCs to refine the methods in which they service their communities, now is the time for CHCs to review their existing methods and make necessary changes. A first step to achieving this change is a better understanding of the specific issues faced by patients who visit CHCs and the way in which these issues are addressed. This study was a collaboration involving multiple CHCs in varying locations (CA, NY and IL) in order to record SDH factors determined by providers to be affecting patients and to evaluate provider perceptions of the impact of these factors on their patients’ health. Further, this study aimed to assess providers’ abilities to address, bill for services focused on the SDH and code for identified SDH. Ultimately, this study sought to document specific information about what is happening, related to the SDH, at the point of care at varying CHC locations. The ultimate intent was to inform policy. Methods The A.T. Still University School of Osteopathic Medicine in Arizona Practice Based Research Network (ATSU SOMA PBRN) implemented a card study from May, 2014 – December, 2015 related to the SDH. The ATSU SOMA PBRN is made up of primary care providers at 12 CHCs across the United States [26]. The PBRN mission is directly related to this project: “focusing on the social determinants of health and the needs of America’s underserved and vulnerable populations we will conduct clinically relevant research to develop and share sustainable programs to improve health care access and quality” [26]. This project was reviewed and deemed exempt by the ATSU Arizona Institutional Review Board. Card study Card studies are observational studies that collect patient level data at the point of care. They have been used for more than 30 years and have been proven effective for describing clinical problems, management issues and outcomes in primary care settings [27]. The card study method can be used to gather information such as observational phenomena, disease incidence and prevalence, practice methods and patterns and clinical practice behaviors. Pocket-sized cards are carried by practitioners and completed at the point of contact with each patient. Cards should be designed to take no longer than 60 s to complete; they are, in essence, a survey that is to be completed while the practitioner is providing care. The cards are designed so the providers can fill them out without having to refer directly to the patient chart [27]. They are used to document, and not to influence, routine care. Although simple in design and implementation, card studies can be valuable in acquiring in-the-moment data about patients, practitioners, and care delivery systems. Card studies are designed to gain information from practitioners who engage primarily in patient care and not in research [27]. Choice of social factors The researchers, with members of ATSU SOMA faculty and the ATSU SOMA PBRN, selected a list of 16 SDH factors to include on the card. These factors were developed based on the Institute of Medicine’s (IOM) Committee on the Recommended Social and Behavioral Measures for Electronic Health Records report [13] and the Healthy People 2020 Social Determinants of Health Framework [8] along with consideration of the WHO’s definition of SDH [2] and factors deemed to be important for patients seen in each CHC setting. After the initial run of the study in California, minor protocol changes were made to better capture SDH factors and improve data collection. For the New York and Illinois sites, three SDH factors were added: Insufficient or Lack of Insurance, Homelessness, and Unstable Housing. These factors were added based on the needs of the communities served by the NY and IL CHCs, where homelessness and insufficient insurance are more common than in the community served by the CA CHC. The CA site is rural and serves a significant number of migrant farm workers. The NY and IL sites are urban. The categories and listing of all SDH factors used in the study are included in Table 1.Table 1 Categorizations of social factors included Category Factors Access to resources Educational limitations Language barrier Poverty Near poverty Insufficient or lack of insurancea Food insecurity Lack of autonomy Homelessnessa Unstable housinga Unstable work schedule Transportation issues Family care demands/Issues Social and home life Cultural beliefs Immigrant status Migrant status Neighborhood safety Optional Other Table 1 shows the SDH factors included in this study aFactors not included in the California study The selected factors were chosen to reflect SDH which are not included in current routine social history screening practices. Therefore, behavioral factors, such as alcohol and tobacco use, and psychological domains such as mood and stress, were not included. The option for “other” was included to allow providers the ability to record factors not on the provided list. Study settings This study was performed with three distinct CHC organizations located in California, New York and Illinois. Each CHC organization has a number of clinic sites; select locations were recruited to participate based on the population served, the number or providers at the sites and the interest of the medical directors for each site. The three CHC organizations recruited for this study are all Federally Qualified Health Centers with an established relationship with ATSU SOMA and are members of the National Association of Community Health Centers (NACHC). A description of each CHC network and the number of sites which participated in the study is included in Table 2.Table 2 Descriptions of CHC locations and patient demographics State Area type Patients Patient demographics At or below 100 % FPI Total sites Sites included in study California Rural 127,128 18.4 % Non-Hispanic White 80.0 % 16 3 76.7 % Hispanic/Latino 1.5 % Black/African American 2.2 % Asian 0.4 % Am. Indian/Alaska Native 0.3 % Hawaiian/Pacific Islander New York Urban 119,734 17.8 % Non-Hispanic White 71.5 % 9 3 47.2 % Hispanic/Latino 20.2 % Black/African American 9.5 % Asian 0.5 % Am. Indian/Alaska Native 0.3 % Hawaiian/Pacific Islander Illinois Urban 34,076 7.1 % Non-Hispanic White 77.4 % 10 6 23 % Hispanic/Latino 68.7 % Black/African American 2.1 % Asian 0.3 % Am. Indian/Alaska Native 0.0 % Hawaiian/Pacific Islander Table 2 provides descriptions of each of the CHC organizations participating in this study Data collected from: 2014 Health Center Profile, HRSA [28, 31, 33] The California CHC network is located in central California and has a patient population which is primarily Hispanic, with 42 % of patients speaking a language other than English [28]. The CHC provides services to a number of migrant, seasonal and undocumented farm workers [29]. Approximately 80 % of the CHC’s patients live at or below 100 % of the federal poverty line. The CHC provides enabling services for patients such as transportation assistance, translation services, health education, and nutritional counseling [28]. The New York CHC organization is located throughout the city of Brooklyn and is one of the largest CHC networks in the nation offering a full range of health and dental care, behavioral health, specialty services and community based programs [30]. Over 71 % of the CHC’s patient population is living at or below 100 % of the federal poverty line and 20.3 % is uninsured. Almost 50 % of the CHC’s patient population is Hispanic and 31.4 % speak a primary language other than English [31]. The Illinois CHC organization is located in Chicago and targets low-income communities throughout the city, which are characterized by an absence of basic services and health care resources [32]. The CHC’s patient population is predominately African American with approximately 77 % of patients living at or below the federal poverty line and 40 % is uninsured [33]. The CHC offers various clinical services, nutrition education, and social support programs [32]. Recruitment and timeline Primary care providers from all 3 CHC organizations were recruited through presentations by study Investigators. In total, 43 out of 71 total eligible providers (61 %) consented to participate in the study; 9 out of 12 (75 %) from California, 21 out of 31 (68 %) from New York and 14 out of 28 (50 %) from Illinois. The active study period was set to run for 4 weeks. In California it ran from May 28th - June 23rd, 2014; in New York from March 5th - April 3rd, 2015; and in Illinois from October 2nd - November 25th, 2015. The study period was extended in Illinois to encourage more participation. Survey and education Upon consent, providers were given a pre-study questionnaire related to the SDH. This 5-item survey was used to assess the following:Providers’ perceptions regarding their familiarity with the SDH concept and how well they feel they can identify SDH for their patients. How important the providers feel SDH are for their patients, and how much the providers feel these factors affect their patients’ health. Provider perceptions of CHC resources to address SDH and how often they refer patients for these services. The survey items were measured using five options for each response. The questionnaire is exhibited in Additional file 1 - Provider Questionnaire. Following the survey, the providers were given a brief (20 min) lecture to define the SDH concept and describe the role SDH play in patient health. Additionally, providers were introduced to the study and given training related to the specific “card” used. This included a description of how to fill out the cards and definitions of the SDH factors. Card study After pre-survey activities were completed, providers were asked to participate in the card study in which qualitative data were collected pertaining to their ability to recognize, document and charge reimbursement fees related to the identified SDH. The cards presented a grid with lines for each potential social factor and also provided space for providers to add their opinions regarding how the identified social factors affected each specific patient’s health. The providers were instructed to complete a card for every patient they saw on one designated day each week for the 4-week time period. Providers were instructed to practice and code in their normal manner. They were specifically instructed not to change their routine or to look up specific codes for this study. The cards in this study were designed to assess the following:Common SDH within the CHC’s catchment area. Provider ability to provide counseling or other interventions for identified SDH. Provider impressions related to the health effects of identified SDH. Provider ability to bill for SDH directed interventions. Provider ability to enter SDH specific diagnosis codes. The card instructions and grids were printed on 9 × 6 in. index cards and were color-coded based on provider specialty (California) or clinic location (New York and Illinois). Cards were to be completed in an anonymous fashion with no identifying information being added about the patients or practitioners. At the end of each assigned day, practitioners placed cards in a designated drop box and the cards were collected by investigators at the end of each week. A copy of the study card is exhibited in Additional file 2 - Study Card. Results Survey All 43 of the participating providers completed the survey prior to beginning the card study. Table 3 presents the results of the survey, which indicate the providers were familiar with the SDH concept and were comfortable with identifying SDH at the point of care. The most frequently chosen response for the first two survey items “How familiar are you with the Social Determinants of Health Concept?” and “How comfortable are you identifying Social Determinants at the point of care?” was “Very Familiar.” Additionally, providers agreed that SDH factors contribute to the health of their patients and they often refer patients to CHC resources to address identified SDH. However, most providers were neutral on whether their CHC had adequate resources to address SDH. Overall, providers at the California CHC responded most favorably when describing their understanding and ability to address SDH factors, whereas providers from the Illinois CHC network responded the least favorably.Table 3 Summary of most frequent responses from provider survey California New York Illinois Combined Question  1. How familiar are you with the SDH concept? Very familiar Very familiar Somewhat familiar Very familiar  2. How comfortable are you identifying SDH at the point of care? Very familiar Very familiar Somewhat familiar Very familiar  3. To what extent do you feel social factors contribute to your patient’s medical condition? Extremely Very much Very much Very much  4. As part of your treatment plan, how often do you refer patients to CHC resources to address SDH? Often Sometimes Often Often  5. My CHC has adequate resources available to address specific SDH affecting patient health. Neutral Neutral Neutral Neutral Table 3 indicates that most providers agreed with being familiar with the SDH concept, that they were comfortable identifying SDH at the point of care and that SDH have an impact on patients’ health. Survey questions each offered 5 options as answers. For items 1 and 2 these were; not familiar, somewhat familiar, neutral, very familiar and extremely familiar. For item 3 they were; not at all, somewhat, neutral, very much and extremely. For item 4 they were; never, rarely, sometimes, often and all of the time. For item 5 they were; strongly disagree, disagree, neutral, agree and strongly agree Card study Among all three CHC networks, there were a total of 747 patient encounters for which cards were completed; 267 in California, 394 in New York and 86 in Illinois. Only 34 patient encounters had no SDH factors identified by providers, with the remaining patient encounters having from 1 to 12 social factors identified. Overall, 1584 SDH factors were identified resulting in an average of 2.12 factors per encounter. Of these identified factors, 493 received counseling and intervention strategies while only 108 diagnosis codes and 20 billing codes were added to patient charts. The most frequently identified factors were Educational Limitations, Language Barriers, and Family Care Demands. Table 4 provides a summary of SDH factors identified, organized greatest to least. The table also shows the number of services provided per SDH and the percentage of each SDH reported as billed or coded by the providers. Almost one third of the SDH were addressed through provider services (31.1 %) with very few documented with a diagnosis code (6.8 %) and only rarely could providers bill for their services addressing the SDH (1.2 %).Table 4 Summary of social factors identified, billed and coded Indicator Count Provided service Billing code Diagnosis code Educational limitations 234 104 (44.4 %) 2 (0.9 %) 5 (2.1 %) Language barrier 176 60 (34.1 %) 3 (1.7 %) 3 (1.7 %) Family care demands 144 67 (46.5 %) 5 (3.5 %) 34 (23.6 %) Poverty 141 24 (17.0 %) 1 (0.7 %) 5 (3.5 %) Unstable work schedule 132 21 (15.9 %) 1 (0.8 %) 2 (1.5 %) Transportation issues 122 51 (41.8 %) 0 (0.0 %) 0 (0.0 %) Near poverty 99 17 (17.2 %) 1 (1.0 %) 3 (3.0 %) Lack of insurancea 86 15 (17.4 %) 0 (0.0 %) 3 (3.5 %) Homelessnessa 68 3 (4.4 %) 0 (0.0 %) 13 (19.1 %) Food insecurity 67 40 (59.7 %) 1 (1.5 %) 29 (43.3 %) Other 67 26 (38.8 %) 2 (3.0 %) 2 (3.0 %) Immigrant status 66 9 (13.6 %) 0 (0.0 %) 1 (1.5 %) Cultural beliefs 60 24 (40.0 %) 3 (5.0 %) 2 (3.3 %) Unstable housinga 55 6 (10.9 %) 0 (0.0 %) 4 (7.3 %) Neighborhood safety 48 24 (50.0 %) 0 (0.0 %) 2 (4.2 %) Migrant status 19 2 (10.5 %) 0 (0.0 %) 0 (0.0 %) Total 1584 493 (31.1 %) 19 (1.2 %) 108 (6.8 %) Table 4 shows a summary of the SDH factors identified, billed and coded by providers aThese factors were not included in the California study card The results from the “other” category are detailed in Table 5. The factors listed in the “other” category include substance abuse and other issues frequently included in the social history. These also included behavioral health issues and poor living conditions.Table 5 Factors identified as “Other” by CHC location CHC location Categorized factors Count Uncategorized factors Total California Drug use (6) 25 9 34 Tobacco use (6) Poor living conditions (4) Behavioral health issues (3) Difficulty making appointments (2) Health literacy (1) Interrupted care - Separated parents (1) History of incarceration (1) Age (1) New York Behavioral health issues (12) 30 3 33 Poor social support (8) Work conditions (5) Physical disability (2) Environmental conditions (1) Drug use (1) Drug use in family (1) Illinois 0 0 0 Table 5 shows the number of factors identified by providers as “Other” by each CHC location. For each “other” identified, providers were asked to include descriptions which were then categorized into the above factors and tallied. In some cases, “other” factors were marked as identified, but did not receive a description by the provider and therefore could not be categorized Table 6 shows the distribution of SDH factors identified by CHC organization. There was an average of 2.12 SDH factors identified during each patient encounter. The most reported SDH factor overall was Educational Limitations. This factor was also the most reported at the California and New York sites. The most reported SDH factor in Illinois was Language Barrier.Table 6 Determinants by CHC location CHC location SDH count Encounters Average/Encounter Most reported California 629 267 2.36 Educational limitations New York 813 394 2.06 Educational limitations Illinois 142 86 1.65 Language barrier Total 1584 747 2.12 Educational limitations Table 6 shows the number of SDH factors reported by each CHC location and the average number of factors observed per patient encounter As detailed in Table 7, the California CHC network reported the highest rate of the use of codes to bill for the provision of services to address the SDH (10) and for specific SDH diagnoses (68). The SDH factor “Family Care Demands” was found to be the factor for which services rendered were coded most frequently with a billing code (5) and which received the most diagnosis codes (34). Overall, billing code documentation and diagnosis code documentation specific to the SDH were low across all CHC locations.Table 7 SDH factors coded and billed by CHC location CHC location Factors coded Factors billed Most frequent services billed Most frequent diagnosis code California 68 (10.8 %) 10 (1.6 %) Family care demands (3) Family care demands (29) New York 32 (3.9 %) 9 (1.1 %) Language barrier (2) Homelessness (13) Family care demands (2) Cultural barrier (2) Illinois 8 (5.6 %) 1 (0.7 %) Neighborhood safety (1) Neighborhood safety (2) Family care demands (2) Total 108 (6.8 %) 20 (1.3 %) Family care demands (5) Family care demands (34) Table 7 shows the SDH factors coded and billed by CHC site and also presents the SDH factors most frequently billed and coded at each site Discussion Based on the data gathered, all three CHC networks have patient populations that likely would benefit from systems to identify, address and formally account for SDH factors. Although all three CHC networks have enabling services in place to address the most frequently identified SDH, less than half (31 %) of the identified factors were reported to be addressed by providers and only rarely were providers able to code for their services provided to address the social factors with counseling, education or another intervention. Further, very few providers were able to enter diagnosis codes into the EHR for the specifically identified SDH. Although not every provider from the CHC organizations participated in the study, participants reported being familiar with the SDH concept, and they expressed the opinion that it is an important aspect to consider in treating patients. There is an obvious discrepancy between the number of identifiable SDH factors, provider ability to address the factors, documentation of the providers’ actions to address the factors using billing codes and documentation of the factors using diagnosis codes. While providers are identifying those SDH factors that they perceive to be impacting their patients’ health, they are not able to provide counseling or other interventions for all of the identified factors. This disparity could be related to their inability to code for SDH factors or bill for related time and services. With very few SDH factors documented with diagnosis codes and even fewer tied to reimbursement codes, it is no surprise providers could not address each in a clinic visit. The International Classification of Diseases, Version 10 (ICD-10) went into effect in the United States on October 1, 2015 and contains 71,924 procedure codes and 69,823 diagnosis codes [34]. Chapter XXI of ICD-10 contains a listing of Z-Codes categorized as “Factors influencing health status and contact with health services”. Included in this chapter of codes is a sub-category of codes (Z55- Z65) intended for “Persons with potential health hazards related to socioeconomic and psychosocial circumstances”. The 11 codes included in this group cover a range of problems including education and literacy, employment, economic circumstances and social environment. Further branching within these categories produces an additional 96 codes [35]. Table 8 presents each SDH factor used in this study along with an ICD-10 code which best corresponds and a description of why the code does or does not fit. SDH factors which did not have a corresponding ICD-10 code include Language Barriers, Insufficient/Lack of Insurance, Transportation Issues, Immigrant Status and Migrant status. Although most factors had a relevant corresponding code, these codes are not SDH-specific and do not fully cover the intended issues. The ICD-9 used the same Z-codes but in a different format, these were still in place for the CA and NY runs of this study. Looking towards the future, the ICD-10 codes are detailed here.Table 8 SDH factors and related ICD-10 codes SDH factor and definition Related ICD-10 code Review of code Educational limitations Z55.0 - Illiteracy and low-level literacy Other corresponding ICD-10 codes were related to children in school. Patient may be lacking in health literacy, but have appropriate level of reading literacy. The available code is too specific. Observed difficulty processing and understanding medical information. Can include difficulty reading, listening, asking questions or applying information. Language barriers None identified Primary language not English; inability to communicate freely and openly with provider. Poverty Z59.5 - Extreme poverty Relatively good match with the social factor. Income below poverty line; lack of basic needs such as nutrition, clothing, shelter. Near poverty Z59.6 - Low income Relatively good match with the social factor. Just enough money to meet basic needs but not enough for extras. Qualifies for sliding fee discounts at FQHC. Insufficient/Lack of insurance None identified Either no health insurance or has insurance which is not sufficient to cover medical expenses or doesn’t cover medications. Prohibits seeking care or follow through. Food insecurity Z59.4 - Lack of adequate food and safe drinking water ICD-10 code is a good match. It may be advantageous to have additional codes which address food availability separately from ability to afford food. Does not have reliable access to sufficient quantity of affordable, nutritious food. Does not know where next meal is coming from. Might live in food desert. Homeless Z59.0 - Homelessness Relatively good match with the social factor. Does not have permanent housing, may live on the streets, in a shelter, mission, abandoned building, vehicle or any unstable non-permanent situation. Unstable housing Z59.1 - Inadequate housing Relatively good match with the social factor. Lives in several places. Lacks fixed, regular and adequate residence due to loss of housing or economic need. No lease, ownership or occupancy agreement. Unsure if place of shelter will be open for extended period of time. Unstable work schedule Z56.3 - Stressful work schedule ICD-10 code is related, but does not cover a work schedule which may prevent patients from seeking regular medical care. Difficulty scheduling or keeping appointments due to variable work schedule; multiple jobs, varying start/stop times, long shifts or unsure when will work. Transportation issues None identified Hard to get to appointments due to lack of transportation. Does not own vehicle, can’t afford public transportation, lives far from public transportation or services unreliable. Family care demands/Issues Z63.2 - Inadequate family support ICD-10 code is related, but somewhat different; does not cover patients who are unable to care for themselves due to the demands of family care. Z63.6 - Dependent relative needing care at home Responsibilities at home caring for others (children, partner, parents, family) which prevent person from caring for themselves. Cultural beliefs Z60.3 - Acculturation difficulty ICD-10 code is somewhat related but doesn’t specifically cover issues pertaining to cultural beliefs which may prohibit a patient from seeking or following medical advice. Cultural background is not in concordance with Western Medicine. May believe Western Medicine can be detrimental or is the place of last resort. Beliefs may conflict with medical care - prohibit patient from seeking care or adhering to treatment plan. Immigrant status None identified Not born in US, now living here legally or illegally. Can have difficulty obtaining public assistance if ‘illegal’. May be child with legal status whose parents do not have legal status. Migrant status None identified Person is a migrant worker who relocates frequently due to work availability. Neighborhood safety Z59.2 - Discord with Neighbors, lodgers, landlord ICD-10 code is related but there remains a lack of codes which address safety and environmental issues. The available code does not clearly match the issue. Not feel safe going outside in neighborhood, threat of crime/violence. Under stress from environment. Children can’t play outside, can’t exercise, hard to get to appointments. Table 8 provides a listing of the SDH factors used by investigators for this study along with related ICD-10 codes. The last column provides a description of how well the codes are matched It is unclear whether the CHC providers are familiar with these coding options and whether they find them useful for SDH factors. However, due to the low rate of recorded codes, it appears these codes are under-utilized and may not be easily documented in an EHR. Tools or procedures to improve the documentation and monitoring of SDH factors need to be considered, such as SDH-specific screening tools or the addition of SDH factors to standard vital signs. Further, codes do not exist for many of the identified SDH and those that do exist are frequently pointing to different issues than the specific SDH providers observed in this study. Limitations This study was performed at three separate CHC organizations in California, New York and Illinois with a total of 43 providers who together documented 747 patient visits. This study was limited through its small size and lack of generalizability. Providers volunteered to participate and could represent CHC providers who are more inclined towards recognizing the SDH. Only a portion of overall patient visits over a defined time period were documented on the cards. Thus, this study represents only a snapshot of what is happening at the point of care. Given the study design with anonymous cards (for both the providers and patients) we can not determine the proportion of patients seen by any given provider for which a card was filled out. The average number of cards per provider was 17. This is less than the number of patients an average provider would see on 4 clinic days. Thus we cannot generalize the average number of SDH per patient to all CHC patients. Nevertheless, obvious discrepancies in SDH identified and accounted for are evident. Conclusions The current state of the health care system coupled with wide spread expansion of coverage provides CHCs with an excellent opportunity to influence change in the way health care is delivered and the focus that is placed on actions at the point of care. This card study is easily replicable, and could be performed in any ambulatory care practice setting to determine providers’ perceptions regarding what SDH issues the patient populations are facing and the ways in which providers are documenting and addressing these issues. PBRN leaders are encouraged to contact the principal investigator in order to replicate this study. It is suggested that the current infrastructure is deficient in that it lacks the “systems put in place to deal with illness”, as recommended by the WHO definition of SDH [2]. While data indicate awareness on the part of CHC providers regarding the SDH, their inability to bill for services rendered to address the SDH and inability to enter diagnosis codes for the specific SDH undermines the goals of the ACA. While there are diagnosis codes available, which are somewhat related to 10 of the SDH identified in this study, only 4 of the ICD-10 codes were deemed a good match with the SDH by the study team. Further work should explore why the available codes were not used. It is striking that these codes were not used by a population of providers who express familiarity with the SDH and who specifically work at CHCs where service is provided regardless of ability to pay. Future research should also explore provider interpretations of the present codes and suggestions for additional codes. Education programs to increase the use of SDH- related codes could be implemented and evaluated. Improving the US health care system and the public’s health must begin with attention to all factors impacting an individual’s or a group of individuals’ health. The IOM’s new framework for educating health professionals to address the social determinants of health is a good start [5]. The ACA expanded coverage to more Americans, can also help. However, coverage in and of itself does not address social barriers to care or social factors affecting health and well-being. Providers must be given the tools within the emerging systems to code, bill and be reimbursed as they seek to address the SDH negatively affecting their patients. CHCs and other health care organizations should seek to implement procedures to document and monitor SDH and the actions taken to address them. The IOM developed EHR screening tool is one available resource intended to allow providers an efficient way to screen and follow-up on identified SDH issues [10, 13]. While possibly related, The SDH factors included in the IOM tool vary from those used in this study. While the IOM screening tool includes behavioral and psychosocial factors not included in this study, the IOM factors are not comprehensive or specific to the detailed SDH factors identified by study providers. The results of this study suggest simple methods of SDH identification may be sufficient, as providers are able to check off identified factors after patient encounters. With minimal direction and the use of a simple tool such as the study card, providers were able to identify SDH through normal clinical practice and conversations with their patients. SDH factors could be included in EHRs along with standard vital signs and marked when identified during a patient encounter. The identified SDH could then be automatically linked with diagnosis codes. Further EHR solutions could automate the linkage to billing codes for actions taken to address the specific SDH. Allowing the potential to receive reimbursement for SDH specific procedures could potentially increase the rates at which providers address these issues. The implementation of any new tool or procedure comes with the potential for an increased burden on providers and health organizations. Therefore, attention needs to be paid to ensuring an efficient system. The SDH factors utilized in this study could also be assessed through patient self-report obtained during a normal clinical encounter. The use of a simple check-list could potentially serve as an effective mode of documentation. The results of the study indicate providers are already identifying SDH factors affecting their patients and spending time during clinical encounters to address these factors. With the addition of searchable codes and reimbursements, we envision improvements in the way these factors are addressed on both an individual and a population level. A growing number of providers, public health and policy experts are working towards expanding standard vital signs to include specific SDH [10, 13, 17]. Only when the SDH are incorporated into the EHR can they be quantified, reviewed and fully addressed by health care providers and health systems. Including the SDH in the EHR can enable more effective treatment of patients at the point of care and more effective population management. Knowledge is power, and in the EHR world, the knowledge must be available in a systematic and searchable mechanism. With the ability to directly tie observed SDH factors to health outcomes for individuals and populations, future funds and programs can more efficiently and more effectively promote true health. Additional files Additional file 1: Provider questionnaire. Description of file: File contains the questionnaire completed by participating providers. (DOCX 15 kb) Additional file 2: Study card. Description of file: File provides an example of the card used for the study. (DOCX 28 kb) Additional file 3: Card data by location. Description of data: Data is broken down by location to show each card that was completed. Each CHC site has two corresponding data sheets. The first sheet contains the raw data including totals which show how many SDH was recorded. The second sheet contains tables with counts for each SDH and how often it was billed and/or diagnosed. (XLSX 235 kb) Additional file 4: Survey data by location. Description of data: Coded survey responses for each provider are included by CHC location. For each question, the average score and mode are provided. A key is included which shows the answer for each coded score. (XLSX 14 kb) Abbreviations ACAPatient Protection and Affordable Care Act ATSU SOMAA.T. Still University of Health Sciences School of Osteopathic Medicine in Arizona ATSU SOMA PBRNA.T. Still University School of Osteopathic Medicine in Arizona Practice-Based Research Network CHCCommunity Health Center EHRElectronic Health Record ICD-10International Classification of Diseases, Volume 10 IOMInstitute of Medicine NACHCNational Association of Community Health Centers SDHSocial determinants of health WHOWorld Health Organization Acknowledgements The authors would like to acknowledge the providers who participated in this study. Collaborating Authors: Mokshasree Atluri, Zachary Blankenship, Laura Grady DO, Timothy Long MD, Amelia MacIntyre MS, Geraldine Malana DO, MPH, Perel Schneid DO and Norma Villanueva MD, MPH Funding This project is supported in part by the ATSU Internal Funding mechanism, Warner Fermaturo Grant, grant number 501–439. This project is also supported by an internal ATSU SOMA research grant. Availability of data and materials All datasets supporting the conclusions of this article are included within the article and in its additional files (Additional file 3 - Card Data by Location and Additional file 4 - Survey Data by Location). Authors’ contributions JHL was the principal investigator for this project. She was primarily responsible for study conception, design and data acquisition. She made substantial contributions to data analysis, interpretation of data and drafting the article. KW was a co-investigator for this project. She made substantial contributions to study design, data acquisition, data analysis, interpretation of data and drafting the article. IN was a co-investigator for this project. He made substantial contributions to study design, data acquisition and interpretation and critical revision of the article for important intellectual content. KRB was a co-investigator for this project. She made substantial contributions to study design, data acquisition and interpretation and critical revision of the article for important intellectual content. All authors have read and approved the final manuscript. The SDH Card Study Implementation team implemented the study design by distributing and collecting study materials, uploading data and coordinating visits with participants. Authors’ information JHL is a professor of Medicine and Public Health at ATSU SOMA, the Chair of the ATSU SOMA Department of Public Health and the Director of the ATSU SOMA PBRN. KW is a Data Specialist in the ATSU SOMA Department of Public Health. IN is the Director of Community Oriented Primary Care at ATSU SOMA. KRB is a consultant for ATSU SOMA. The SDH Card Study Implementation team is comprised of ATSU SOMA students, residents and Regional Directors of Medical Education. Competing interests The authors declare that they have no competing interests. Consent to publish Not applicable Ethics approval and consent to participate This project was reviewed and deemed exempt by the ATSU Arizona Institutional Review Board. All providers signed consent forms to participate. ==== Refs References 1. Irwin AI Scali E Action on the social determinants of health: a historical perspective Glob Public Health 2007 2 3 235 56 10.1080/17441690601106304 19283626 2. World Health Organization. Social Determinants of Health. http://www.who.int/topics/social_determinants/en/. Accessed 12 Apr 2016. 3. Jong-wook L Public health is a social issue Lancet 2005 365 1005 6 10.1016/S0140-6736(05)71115-6 15781079 4. CSDH Closing the gap in a generation: health equity through action on the social determinants of health. Final report of the Commission on Social Determinants of Health 2008 Geneva World Health Organization 5. National Academies of Sciences, Engineering and Medicine A framework for educating health professionals to address the social determinants of health 2016 Washington The National Academies Press 6. American Public Health Association. What Impacts our Health? https://www.apha.org/topics-and-issues/generation-public-health/what-impacts-our-health. Accessed 14 July 2016. 7. Kaiser Family Foundation. Beyond health care: The role of social determinants in promoting health and health equity. Issue Brief. http://kff.org/disparities-policy/issue-brief/beyond-health-care-the-role-of-social-determinants-in-promoting-health-and-health-equity/. Published November 2015. Accessed July 2016. 8. Healthy People 2020. Social determinants of health. https://www.healthypeople.gov/2020/topics-objectives/topic/social-determinants-of-health. Updated July 14, 2016. Accessed 14 July 2016. 9. Centers for Disease Control. Social determinants of health: know what affects health. Frequently asked questions. http://www.cdc.gov/socialdeterminants/faqs/index.htm. Updated October 19, 2015. Accessed 14 July 2016. 10. Adler NE Stead WW Patients in context - EHR capture of social and behavioral determinants of health N Engl J Med 2015 372 8 698 701 10.1056/NEJMp1413945 25693009 11. Seligman HK Laraia BA Kushel MB Food insecurity is associated with chronic disease among low-income NHANES participants J Nutr 2010 140 2 304 10 10.3945/jn.109.112573 20032485 12. Pelletier SG. Addressing social determinants of health: Tackling food insecurity in communities. Association of American Medical Colleges. https://www.aamc.org/newsroom/reporter/september2015/442218/social-determinants.html. Published September 2015. Accessed July 2016. 13. Institute of Medicine Capturing social and behavioral domains and measures in electronic health records: phase 2 2014 Washington The National Academies Press 14. Forde I Raine R Placing the individual within a social determinants approach to health inequity Lancet 2008 372 1694 6 10.1016/S0140-6736(08)61695-5 18994669 15. Kaiser Family Foundation. Disparities in health and health care: Five key questions and answers. Issue Brief. http://kff.org/disparities-policy/issue-brief/disparities-in-health-and-health-care-five-key-questions-and-answers/. Published November 30, 2012. Accessed 14 July 2016. 16. Milstein B Homer J Briss P Burton D Pechacek T Why behavioral and environmental interventions are needed to improve health at lower cost Health Aff 2011 30 5 823 32 10.1377/hlthaff.2010.1116 17. U.S. Department of Health and Human Resources. Read the law: the Affordable Care Act, section by section. http://www.hhs.gov/healthcare/about-the-law/read-the-law/index.html. Updated August 28, 2015. Accessed 12 Apr 2016. 18. National Association of Community Health Centers. America’s health centers (fact sheet #0811). http://allianceforhealthcare.org/files/pdf1.pdf. Published August, 2011. Accessed 12 Apr 2016. 19. Franks P Clancy C Gold M Health insurance and mortality. Evidence from a National Cohort JAMA 1993 27 6 737 41 10.1001/jama.1993.03510060083037 8336376 20. Galewitz P. Quality of community health centers varies widely. Kaiser Health News. http://khn.org/news/community-health-centers-quality-varies-widely/. Published October 31, 2012. Accessed 12 Apr 2016. 21. Health Resources and Services Administration. What is a health center. http://www.bphc.hrsa.gov/about/what-is-a-health-center/index.html. Accessed 12 Apr 2016. 22. Schroeder S We can do better - Improving the health of the American people N Engl J Med 2007 357 1221 8 10.1056/NEJMsa073350 17881753 23. McGinnis JM Williams-Russo P Knickman JR The case for more active policy attention to health promotion Health Aff 2002 21 2 78 93 10.1377/hlthaff.21.2.78 24. Institute for Alternative Futures. Community Health Centers Leveraging the Social Determinants of Health. Alexandria, VA. http://www.altfutures.org/pubs/leveragingSDH/IAF-CHCsLeveragingSDH.pdf. March 2012. Accessed 12 Apr 2016. 25. Kaiser Commission on Key Facts. Community Health Centers. Kaiser Family Foundation. Washington, DC. http://www.kff.org/uninsured/upload/7877.pdf. March 1, 2009. Accessed 12 Apr 2016. 26. Agency for Healthcare Research and Quality. A.T. Still University, School of Osteopathic Medicine in Arizona PBRN. US Department of Health and Human Resources. https://pbrn.ahrq.gov/pbrn-registry/t-still-university-school-osteopathic-medicine-arizona-pbrn. January 1, 2012. Accessed 12 Apr 2016. 27. Westfall JM Zittleman L Stanton E Card studies for observational research in practice Ann Fam Med 2011 9 1 63 8 10.1370/afm.1199 21242563 28. HRSA Health Center Program. 2014 Health Center Profile, Family Healthcare Network. US Department of Health and Human Resources. 2015. http://bphc.hrsa.gov/uds/datacenter.aspx?q=d&bid=093640&state=CA&year=2014. Accessed 12 Apr 2016. 29. Family Healthcare Network Who we serve 2016 30. NYU Lutheran, NYU Langone Health System. NYU Lutheran Family Health Centers. http://www.lutheranhealthcare.org/Main/LutheranFamilyHealthCenters.aspx. Accessed 12 Apr 2016. 31. HRSA Health Center Program. 2014 Health Center Profile, Sunset Park Health Council, Inc. US Department of Health and Human Resources. 2015. http://bphc.hrsa.gov/uds/datacenter.aspx?q=d&bid=0218870&state=NY&year=2014. Accessed 12 Apr 2016. 32. NearNorth Health Service Corporation About us 2015 33. HRSA Health Center Program. 2014 Health Center Profile, Near North Health Service Corporation. US Department of Health and Human Resources. 2015. http://bphc.hrsa.gov/uds/datacenter.aspx?q=d&bid=053280&state=IL&year=2014. Accessed 12 Apr 2016. 34. Centers for Disease Control and Prevention. International Classification of Diseases, (ICD-10-CM/PCS) Transition - Background. http://www.cdc.gov/nchs/icd/icd10cm_pcs_background.htm. Published November 6, 2015. Updated October 1, 2015. Accessed 12 Apr 2016. 35. World Health Organization International statistical classification of diseases and health related problem, 10th revision 1992 Geneva World Health Organization

PMC005xxxxxx/PMC5002328.txt

==== Front BMC Pregnancy ChildbirthBMC Pregnancy ChildbirthBMC Pregnancy and Childbirth1471-2393BioMed Central London 102210.1186/s12884-016-1022-9Research ArticleDeterminants of institutional birth among women in rural Nepal: a mixed-methods cross-sectional study Maru Sheela sheela@possiblehealth.org 1234Rajeev Sindhya sindhya.rajeev@gmail.com 5Pokhrel Richa richapokhrel@gmail.com 6Poudyal Agya agyapoudyal@gmail.com 7Mehta Pooja pooja.mehta@bmc.org 23Bista Deepak dbista@possiblehealth.org 1Borgatta Lynn borgatta@bu.edu 23Maru Duncan duncan@possiblehealth.org 189101 Bayalpata Hospital, Possible, Sanfebagar-10, Achham Nepal 2 Department of Obstetrics and Gynecology, Boston Medical Center, Boston, MA USA 3 Department of Obstetrics and Gynecology, Boston University School of Medicine, Boston, MA USA 4 Department of Medicine, Division of Women’s Health, Brigham and Women’s Hospital, Boston, MA USA 5 Bellevue Hospital Center, Ronald O. Perelman Department of Emergency Medicine, New York University School of Medicine, New York, NY USA 6 City of Berkeley, Berkeley, CA USA 7 Faculty of Health and Life Sciences, Department of Biological and Life Sciences, Oxford Brookes University, Oxford, United Kingdom 8 Department of Medicine, Division of Global Health Equity, Brigham and Women’s Hospital, Boston, MA USA 9 Department of Medicine, Division of General Pediatrics, Boston Children’s Hospital, Boston, MA USA 10 Department of Medicine, Harvard Medical School, Boston, MA USA 27 8 2016 27 8 2016 2016 16 1 2522 4 2015 8 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Encouraging institutional birth is an important component of reducing maternal mortality in low-resource settings. This study aims to identify and understand the determinants of persistently low institutional birth in rural Nepal, with the goal of informing future interventions to reduce high rates of maternal mortality. Methods Postpartum women giving birth in the catchment area population of a district-level hospital in the Far-Western Development Region of Nepal were invited to complete a cross-sectional survey in 2012 about their recent birth experience. Quantitative and qualitative methods were used to determine the institutional birth rate, social and demographic predictors of institutional birth, and barriers to institutional birth. Results The institutional birth rate for the hospital’s catchment area population was calculated to be 0.30 (54 home births, 23 facility births). Institutional birth was more likely as age decreased (ORs in the range of 0.20–0.28) and as income increased (ORs in the range of 1.38–1.45). Institutional birth among women who owned land was less likely (OR = 0.82 [0.71, 0.92]). Ninety percent of participants in the institutional birth group identified safety and good care as the most important factors determining location of birth, whereas 60 % of participants in the home birth group reported distance from hospital as a key determinant of location of birth. Qualitative analysis elucidated the importance of social support, financial resources, birth planning, awareness of services, perception of safety, and referral capacity in achieving an institutional birth. Conclusion Age, income, and land ownership, but not patient preference, were key predictors of institutional birth. Most women believed that birth at the hospital was safer regardless of where they gave birth. Future interventions to increase rates of institutional birth should address structural barriers including differences in socioeconomic status, social support, transportation resources, and birth preparedness. Electronic supplementary material The online version of this article (doi:10.1186/s12884-016-1022-9) contains supplementary material, which is available to authorized users. Keywords Institutional birth rateMaternal mortalitySkilled birth attendantsWomen’s healthGlobal healthImplementation researchNepalissue-copyright-statement© The Author(s) 2016 ==== Body Background Introduction The maternal mortality ratio in Nepal is estimated at 281 per 100,000; approximately half of these cases are attributed to postpartum hemorrhage [1]. Given that active management of the third stage of labor minimizes obstetric morbidity and mortality in low-resource settings, increasing the rate of births that take place in a medical facility is a key approach to reducing maternal mortality [2–4]. Despite efforts in Nepal to increase rates of institutional birth, 35 % of births nation-wide take place in a medical facility and in the impoverished Far-Western Development Region where this study takes place, that percentage is only 29 %. In Nepal, only 5 % of the estimated need for emergency obstetric care is met, and only 0.7 % of births are by cesarean section compared to the United Nations target of 5–15 % [5]. In a literature review of the determinants of birth service use in the Global South, maternal age, parity, education, household wealth, and urban residence were positively associated with service usage [6–9]. In a key study done by the Nepal Safe Motherhood Program in five districts in Nepal, 96 women were interviewed about their decision to give birth at an emergency obstetric care facility. When asked about why they chose a particular emergency obstetric care facility, availability of the full range of obstetric services was an important determinant (35 % of mothers). Of note, 2 % of the mothers reported making a decision as to where to give birth themselves, while 83 % of mothers reported that male members of the family made the decisions [10]. Other published works from Nepal confirm that family members influence the decision to give birth at home [11, 12]. Significance Specific barriers to safe institutional birth can be understood within the framework of the Three Delays Model: first, delay in the decision to seek care, second, delay in arrival at a health facility, and third, delay in the provision of appropriate care at the healthcare facility [13, 14]. The objective of this study is to understand which factors determine a woman’s decision to have an institutional birth in rural Nepal. Study findings will inform priorities in scale up to comprehensive emergency obstetric care, quality improvement strategies, and the creation of more responsive community-based and hospital-based antenatal, obstetric, and postpartum programming in rural Nepal. These findings may also be useful to the development of comprehensive emergency obstetric care programs in culturally similar, low-resource settings. Methods Study design Data for this study was obtained from the Pre-Surgical Expansion Survey (Additional file 1) administered for 3 months in 2012 in the area around Bayalpata Hospital. The sample was comprised of 1) postpartum women in the hospital’s catchment area population who had given birth at home, and 2) postpartum women who had given birth at the hospital or at a village clinic. Written informed consent, either by signature or thumbprint on a consent form, was obtained for both the quantitative survey and the qualitative interview. This is the first part of a larger study to explore how decision-making changes after expansion of obstetric services at a district hospital. The data from the second survey period (Post-Surgical Expansion Survey) will be compared to the Pre-Surgical Expansion Survey data and presented in a subsequent publication. Setting Bayalpata Hospital is located in Achham, a rural district in the Far-Western Development Region of Nepal. It is a district-level hospital run through a private-public partnership between the Ministry of Health of Nepal and the Nepali non-profit organization Possible. Services are provided to patients of the hospital free of charge. These services include birth services, laboratory diagnostics, an on-site pharmacy, and an operating theatre. The Community Health Program includes counseling and tracking of pregnant women by Community Health Workers (CHWs; known locally in Nepal as Female Community Health Volunteers). The CHWs’ coverage area, at the time of research, included seven village clusters (known locally in Nepal as Village Development Committees) around the hospital that contributed a large portion of patients to Bayalpata Hospital and defined a catchment area population of approximately 17,000 people for this study. At the time of the survey, the hospital’s ambulance service charged a distance-based fee, while the hospital provided Basic Emergency Obstetrics Care, with no capacity for cesarean sections or blood transfusion. Selection of participants Study parameters included a convenience sampling of nearly all women who gave birth in the home setting as well as an emergency obstetric care facility over a 3 month period. Taking the findings of the Nepal Safe Motherhood study into account, which found that only 2 % of the mothers make the decision regarding where to give birth themselves, the study team sought to include a subset of male family members. However, because the survey’s male sample size was too small to analyze trends, this group was ultimately excluded. Therefore the final inclusion criteria included only post-partum women in the catchment area population who had given birth in the previous 6 weeks. Power was calculated for a planned evaluation of the Pre- and Post- Surgical Expansion Survey data. Using an algorithm based on the Wald test, with power set at 0.80, statistical significance set at 0.05, and a baseline institutional birth rate of 20 %, the study was powered to detect a 150 % increase (i.e., to 30 %) in institutional births after the implementation of comprehensive emergency obstetric services. This power calculation required a total sample size of 200 subjects over both Pre- and Post- Surgical Expansion Surveys. For the Pre-Surgical Expansion survey, the study team stopped data collection at 3 months, when the target of 100 participants was nearly reached. All eligible women were approached for recruitment during this time period. Data collection and processing Bayalpata Hospital Pre-Surgical Expansion Survey data were used to examine attitudes and behaviors toward institutional birth before the establishment of the hospital’s comprehensive emergency obstetric care facility. The survey included demographic inquiries and questions to assess the decision-making process, accessibility of institutional birth, and acceptability of services. In addition, patient narratives of their birth stories and their recommendations for improving utilization of institutional services were collected. Community Health Worker Leaders (CHWLs), paid employees in Bayalpata Hospital’s Community Health Program managing the CHWs, administered the surveys to post-partum women in the catchment area population of the hospital who had given birth at home or at a village clinic in the prior 6 weeks. If more than one woman in a household was eligible for participation, all eligible women were approached for enrollment. Auxiliary Nurse Midwives (ANMs) administered surveys to post-partum women who had given birth at the hospital in the prior 6 weeks, inclusive of both uncomplicated and complicated births. An attempt was made to reach all post-partum women during the study period. Surveys were printed in Nepali and enumerators conducted the surveys in Nepali or in Achhami, the local dialect. Possible’s Director of Research was present for the majority of survey administration to ensure consistency. The CHWs and ANMs were also given weekly reminders to complete the survey forms. Outcome measures The primary outcome was location of birth. The goals were to describe and quantify factors associated with institutional birth and home birth, and to understand which factors are significant predictors of institutional birth. The following factors were examined: age, distance from hospital, household income, land ownership, race, literacy, parity, antenatal care, birth decision-making, and father’s presence. The qualitative analysis was based on the social contextual theoretical framework [15], and aimed to identify modifying and mediating factors in relation to the future intervention of comprehensive emergency obstetric care services. Quantitative analysis All data were collected on paper forms and input into Excel (Microsoft Corp., Richmond WA) by the staff of the Community Health Program and the Hospital Program at Bayalpata Hospital. For this study, survey data were analyzed using Stata 12.1 (Stata Corporation, College Station, TX) and JMP Pro version 10.0.0. Data collection tools were pilot tested for errors at Bayalpata Hospital prior to the survey. Each variable was examined by location of birth. For categorical variables, cell counts were examined. Each continuous variable was examined for normality of distribution. For normally distributed variables, means and standard deviations were calculated, for non-normal data, medians and inter-quartile ranges were calculated. For the logistic regression, a manual forward-selection method was used, starting with scientifically important covariates: age, parity, literacy, income, and distance from the hospital. A covariate in the model was kept if it was significant (p < 0.10) or if it altered the beta coefficient by more than 20 %. This process led to a final logistic regression model with five covariates: parity, literacy, monthly household income, land, and presence of father of baby. To examine the effect of these covariates on the outcome of location of birth, odds ratios with 95 % confidence intervals were used. To test the significance of the effects, the likelihood ratio tests were used. In addition to the regression analysis, descriptive statistics of perceptions of care, accessibility and acceptability of services, and major recommendations for improving institutional birth utilization were examined. Qualitative analysis A single open-ended question was posed at the start of each survey: “Tell me the story of your birth.” Responses were transcribed on the survey by the interviewer in Nepali. These responses were then translated into English for analysis. Analysis was based on the social contextual model [15]. This theoretical framework helped to illuminate pathways by which social contextual factors lead to differing health outcomes or health behaviors. Based on the model, these factors were then examined as they may or may not be affected by a proposed intervention. For this study, the outcome of interest was institutional birth and the proposed intervention was expansion of services at Bayalpata Hospital from basic to comprehensive emergency obstetric care, specifically providing blood banking and cesarean sections. Through immersion crystallization by two investigators, modifying and mediating factors were identified [16]. Modifying factors were thought to independently affect the outcome, and not be influenced by the intervention. Mediating factors were thought to be on the pathway between the intervention and the outcome. The two investigators identified and categorized these factors prior to the roll out of the intervention using data from Pre-Surgical Expansion Survey. These factors occurred on multiple levels: individual, interpersonal, organizational, community, and societal. Results Over a 3-month collection period in 2012, 98 subjects were enrolled in the Pre-Surgical Expansion Assessment study. Of these, 57 of the women gave birth at home and 39 gave birth at a health institution. Of the final total of enrolled women, 77 were from the catchment area population; the remaining 21 women were from neighboring communities outside the catchment area population, but who received healthcare services at Bayalpata Hospital. The reported institutional birth rate presented is restricted to the 77 women from the catchment area population, as an accurate rate could not be calculated to include neighboring communities. The remainder of the analysis included all 98 women in order to maximize power in looking at predictors of institutional birth. The distributions of characteristics by location of birth are presented in Table 1. Mean participant age and median distance from the hospital were similar between groups. For the institutional birth group, the median monthly household income was Nepali Rupees (NRs) 5000 (IQR 3000, 10000) and that for the home birth group was NRs 200 (IQR 0, 3250). Median monthly household income was much greater in the institutional birth group, however median amount of land was greater in the home birth group. Forty-six percent of home births were from the Dalit caste (a disadvantaged socio-ethnic group stratified by the Hindu caste system), versus only 37 % of institutional births. Seventy seven percent of home births were of multiparous women, compared to 61 % of institutional births. Percent of adequate antenatal care and self-birth-decision-making were similar between groups. In 76 % of institutional births, the father of the baby was present in the district at the time of post-partum interview, compared with just 53 % of home births.Table 1 Characteristics of the study samples: Comparison of institutional versus home births Characteristic Institutional Birth (n = 39, 40 %) Home Birth (n = 57, 60 %) P-Value Age, years, mean (SD) 24 (6) 25 (4) 0.61 Distance from district hospital, hours, median (IQR) 2 (1, 2.9) 2 (1.5, 2) 0.33 Monthly household income, Nepali Rupees, median (IQR) 5000 (3000, 10000) 200 (0, 3250) 0.02 Land, Ropania, median (IQR) 3 (1, 7.5) 5 (2, 12) 0.06 Caste, number (%)  Dalit 15 (37) 26 (46) 0.37 Literacy, number (%)  Literate 30 (73) 28 (49) 0.02 Parity, number (%)  Primiparity 16 (39) 13 (23) 0.08 Antenatal care, number (%)  Adequateb 30 (73) 39 (68) 0.61 Birth decision making, number (%)  Self alone 14 (34) 17 (30) 0.69 Presence of father of baby, number (%)  Present 31 (76) 30 (53) 0.02 aRopani is a commonly used unit in Nepal to measure land and is equal to 508.72 m2 bDefined as at least 4 antenatal care visits, per Nepali Government guidelines The logistic regression model (Table 2) explained 39 % of the variability in the data. Age, monthly household income, and land were significant predictors of institutional birth (p < 0.05). Because age and monthly household income have a non-linear relationship with institutional birth, squared terms were included in the final regression model. As age increased, the likelihood of institutional birth decreased. The odds ratio was higher in older age groups (Table 3). As income increased, the likelihood of an institutional birth also increased. However, beyond a certain point, additional income had less of an effect on birth location (Table 4). Income and land, both measures of socio-economic status, are expected to be related to birth location in the same direction. In the study data, they relate in opposite directions. A higher monthly household income predicts institutional birth, whereas a lower amount of land ownership predicts institutional birth.Table 2 Results of logistic regression: likelihood of institutional birth by predictors Covariatea OR [95 % CI] P-Value (likelihood ratio) Age 0.06 [0.01, 0.33] 0.0008 Age^2 1.06 [1.03, 1.11] 0.0003 Land, Ropani 0.82 [0.71, 0.92] 0.0001 Monthly household income, in 1000 Nepali Rupees (NR) 1.45 [1.16, 1.88] 0.0006 Monthly household income^2, in 1000 NR 0.99 [0.98, 0.99] 0.0021 Parity 0.69 [0.13, 3.89] 0.6640 Literacy 0.33 [0.07, 1.40] 0.1323 Presence of father of baby 0.40 [0.09, 1.73] 0.2215 Distance 1.09 [0.67, 1.73] 0.7218 aReference categories: for Parity, Nulliparous; for Literacy, Illiterate; and for Presence of father of baby, Absent Table 3 Odds Ratios for adjusted age as a predictor of institutional birth Percentile Age OR 25 21 0.20 50 24 0.23 75 27.25 0.28 Table 4 Odds Ratios for adjusted monthly household income as a predictor of institutional birth Percentile Monthly household income, in 1000 NR OR 25 0.0075 1.45 50 2.25 1.42 75 5.25 1.38 When asked about the most important factor in determining the location of birth, most women in the institutional birth group selected "safety/good care" (90 %), whereas most women in the home birth group selected "long distance" (60 %, Fig. 1). Giving birth at the hospital was considered to be safer by the majority of both the institutional birth group (98 %) and the home birth group (88 %).Fig. 1 Location of birth and most important factors The cost of travel to the birth location for the institutional birth group was NRs 831 (CI 486.5–1174.6). Overall, the most common participant recommendation for improving institutional service utilization was improving ambulance accessibility (72 %); only 37 % of the institutional birth sample arrived via an ambulance. In terms of acceptability of services, institutional birth satisfaction was at 93 % while home birth satisfaction was 32 %. All participants in the institutional birth group stated they would prefer an institutional birth in the future, while only one participant from the home birth group stated a preference for home birth in the future. The qualitative analysis examined women’s ‘birth stories’ for social contextual factors as they would relate to the proposed expansion of emergency obstetric services. Modifying factors and mediating factors were identified and fit into a conceptual model (Fig. 2). Modifying factors at the individual and interpersonal level included family (particularly mother-in-law) and partner support, access to financial resources and means of transport to an institutional setting, birth planning and preparedness, gendered work responsibilities, and positive or negative personal experiences during a prior birth.Fig. 2 Factors affecting institutional birth. *Note: The mediating factors were defined as likely being affected by the proposed intervention while the modifying were likely not going to be affected by the intervention The importance of family and partner support was identified in the stories of many women, as this quote exemplifies: “She wanted to deliver in the hospital but her in-laws did not listen. Her husband was not in Achham. She didn't have the courage to come to the hospital by herself so she delivered at home.” The factors of access to transportation, birth planning, and preparedness were also seen as community level factors due to the need for several community members to participate in the process of getting to an institution: “A [20–30] year old woman who was pregnant with [a] child was herding her goats during the day. She went into labor around 7 pm. The family was trying to gather people to bring her to the hospital but she had her baby an hour and half later.” A societal level modifying factor identified was the quality of services at the village clinics, which was often poor: “They wanted to go to the village clinic but no medical staff were there.” The government-provided financial incentive was another societal level modifying factor that motivated women to have an institutional birth. The birth stories also revealed social and contextual factors that would likely be affected by the proposed expansion of services. On an individual and interpersonal level, perceptions of safety and intention to present to an institution seemed important for achieving institutional birth, as this woman’s story exemplifies:“[A 40–50] year old woman who had [multiple] prior deliveries in her home, just had [a] delivery in the hospital. One year ago there was a maternal death in a nearby village while a woman was delivering at home. News of this death changed her mind about having her [next] baby at home. She wanted to come to the hospital, where there was a doctor and good medical staff to keep her safe.” Unlike this story of a woman who achieved institutional birth, there were several women who gave birth at home and mentioned that they had limited awareness of the hospital or services provided there. Referral to a higher level of care was an organizational level mediating factor. Three women were referred to an institution with comprehensive emergency obstetric care services, but delivered at home due to delays caused by referral. Discussion The results of this study are in line with prior literature, in that age, income, and land, were significantly related to institutional birth. However, age is inversely related to institutional birth, with younger women more likely to give birth in an institution. The direction of this relationship is contrary to findings in other settings [6]. This may be explained by circumstances in rural Nepal that are different than in other parts of the world, such as literacy. In this study, the average age of literate women is significantly younger (23) than that of illiterate women (27). Literacy may be a proxy for other characteristics, such as empowerment and independence, which could impact usage of institutional birth services [17]. Higher income predicted institutional birth in our analysis. The non-linear relationship, however, reveals that the amount of income matters less in the higher range for achieving institutional birth. Land has an opposite effect on institutional birth, as ownership of more land negatively predicts institutional birth. Perhaps this is due to the specific very rural context of the study as families with more land are unlikely to be earning other sources of cash revenue. In a remote area like Achham, cash is important to secure transportation. Perhaps those with less land had easier access to that cash. Further analysis of the current data, however, shows no significant association between land and income. This issue warrants further research in a larger sample of women or with qualitative methods. Findings suggest that the most important consideration for women is safety/good care among the institutional birth group and long distance/transportation in the home birth group. The majority of all women stated that the hospital was safer place to give birth and would prefer an institutional birth in the future. These findings indicate that underutilization of institutional birth services is unlikely to be the result of patient preference, and rather due to socioeconomic vulnerability to structural barriers. Although distance from the hospital was not a significant predictor of institutional birth in the final regression model, ‘long distance or transportation’ was the most important consideration reported by the home birth group. Other studies show distance to be a significant factor that affects choice of birthplace [18, 19]. In this study, the most common participant recommendation to increase institutional service utilization was to improve ambulance accessibility and currently only a minority of those who gave birth at an institution arrived via an ambulance. Thus, the delay in arriving at the hospital is an area that must be addressed if institutional birth rates are to improve. Improving transportation delays have been shown in other studies to reduce neonatal mortality and stillbirths, when integrated with community mobilization, education, and facility improvement [20]. The mediating social and contextual factors identified in the qualitative analysis are those that will most likely be impacted by the proposed expansion of emergency obstetric services: perceptions of safety, intent to give birth in the hospital, referrals, and awareness of services. On the other hand, modifying social and contextual factors, such as social support, gender equality, financial and transportation resources, and birth preparedness, are less likely to be impacted by the expansion of services. Thus, they should be the targets of interventions conducted alongside expansion of emergency obstetric services to improve institutional birth in this region. Limitations to this study include a small sample size that was made smaller by missing values. The power calculation was for the larger study, comparing Pre- and Post- Surgical Expansion Survey data, not for the present analysis of determining factors associated with location of birth. Thus, the study may be under-powered to detect all factors associated with institutional birth. The convenience sampling strategy employed may have biased the findings. Future research with a consecutive sampling strategy should be employed to address this issue. Additionally, the institutional birth rate among this sample (42 %) was higher than the estimated rate in the hospital catchment area population (30 %). The skew towards women who had an institutional birth brings bias to the sample. The use of community health workers and nurse midwives as enumerators also increases bias in self-reported perceptions of and preferences for institutional birth. The generalizability of the findings is limited due to the data being obtained from a single hospital and its catchment area population. Finally, due to resource limitations in our setting, qualitative interviews could not be audio-recorded, perhaps leading to error in transcription and translation of participant responses. Despite these limitations, this study makes a unique contribution to the small and growing literature on determinants of institutional birth in the Global South. Conclusions In exploring determinants of institutional birth in rural Nepal, we found that socioeconomic vulnerability to structural barriers are paramount in understanding who gives birth at home versus an institution and why. Contrary to data from the Nepal Demographic and Health Survey 2011, where 62 % of women who gave birth at home believed it was not necessary to give birth in a health facility [21], most women surveyed in our study, in rural Nepal in 2012, believed that it was safer to have an institutional birth. Despite this difference in thinking, the majority of women in our study still gave birth at home. Efforts to reduce maternal mortality must both make institutions safer by expanding to comprehensive emergency obstetric services, as well as target the underlying social and economic inequity and structural barriers that prevent women from achieving institutional birth. Additional file Additional file 1: Institutional Birth Survey_Pre-Surgical Expansion_Eng,Nep (PDF 257 kb) Abbreviations ANM(s)Auxiliary Nurse Midwives CHW(s)Community Health Worker(s) CHWL(s)Community Health Worker Leader(s) IQR(s)Interquartile Range(s) NRsNepali Rupees OR(s)Odds Ratio(s) SDStandard Deviation UNUnited Nations Acknowledgements We wish to express our appreciation to the Nepal Ministry of Health for their continued efforts to improve the institutional birth rate in Nepal; to the Nepal Health Research Council and the Brigham and Women’s Hospital Institutional Review Board for granting ethical approval; to Garrett Fitzmaurice for assisting with statistical calculations and manuscript preparations; to Scott Halliday for help in editing the manuscript; and to the staff and Community Health Workers of Bayalpata Hospital for their contributions during the implementation of the study. We also wish to thank all of our participants who generously shared their stories and experiences with us. Funding Dr. Sheela Maru received a Global Women’s Health Fellowship from the Mary Horrigan Connors Center for Women’s Health & Gender Biology at Brigham and Women’s Hospital and the Harvard Humanitarian Initiative. The authors did not receive any other sources of funding. The Mary Horrigan Connors Center for Women’s Health & Gender Biology at Brigham and Women’s Hospital and the Harvard Humanitarian Initiative played no role in the design, data collection, analysis, decision to publish, or publication of this study. Availability of data and materials De-identified quantitative data are available at the Healthcare System Design Group's (Possible's Implementation Research Team) website (http://hsdg.partners.org/data/). We do not provide full transcriptions of the qualitative data. These transcripts contain quotes and identifiable information that could compromise the identity of participants. Data may also be requested by emailing: research@possiblehealth.org. Authors’ contributions SM conceived the study and serves as guarantor for the work. The study grew out of discussions between SM, PM, LB, and DM. SM, AP, and RP developed the survey instruments. SM, SR, RP, and AP assisted with data collection. DB was responsible for data entry. SM, SR, and PM performed the data analysis. SM and SR drafted the manuscript. All authors have reviewed and approved the manuscript. Competing interests SM and DM work in partnership with and DB is employed by a nonprofit healthcare company (Possible) that delivers free healthcare in rural Nepal using funds from the Government of Nepal and other public, philanthropic, and private foundation sources. At the time of research, RP and AP were employed by Possible, while SR was an unpaid volunteer with Possible. SM and PM are employed at an academic medical center (Boston Medical Center) that receives public sector research funding, as well as revenue through private sector fee-for-service medical transactions and private foundation grants. SM and PM are faculty members at a private university (Boston University). SM is a research fellow at and DM is employed by an academic medical center (Brigham and Women’s Hospital) that receives public sector research funding, as well as revenue through private sector fee-for-service medical transactions and private foundation grants. SR is a resident at a public hospital (Bellevue Medical Center) that receives public sector funding, revenue through fee-for-service medical transactions, and public sector research grants, and is affiliated with a private university (New York University). RP is a government employee (City of Berkeley). AP is employed by a public university (Oxford Brookes University). DM is employed at a separate academic medical center (Boston Children’s Hospital) that receive public sector research funding, as well as revenue through private sector fee-for-service medical transactions and private foundation grants. DM is a non-voting member on the Board of Directors with Possible, a position for which he receives no compensation. DM is also a faculty member at a private university (Harvard Medical School). All authors have read and understood BMC Pregnancy and Childbirth’s policy on declaration of interests, and declare that we have no competing financial interests. The authors do, however, believe strongly that healthcare is a public good, not a private commodity. Consent for publication Not applicable. Ethics approval and consent to participate The study protocol was approved by the Institutional Review Board of both the Nepal Health Research Council (#98/2011) and Brigham and Women’s Hospital (2012P000856). Written informed consent, either by signature or thumbprint on a consent form, was obtained for the quantitative survey and the qualitative interview. ==== Refs References 1. Ministry of Health and Population Nepal New ERA Macro International Inc Nepal Demographic and Health Survey, 2006 2007 Kathmandu Ministry of Health and Population Nepal, New ERA, and Macro International Inc 2. International joint policy statement FIGO/ICM global initiative to prevent post-partum hemorrhage J Obstet Gynaecol Can 2004 26 12 1100 2 10.1016/S1701-2163(16)30440-6 15696639 3. Kayongo M Making EmOC a reality--CARE's experiences in areas of high maternal mortality in Africa Int J Gynaecol Obstet 2006 92 3 308 19 10.1016/j.ijgo.2005.12.003 16442536 4. Lawn JE Reducing intrapartum-related deaths and disability: can the health system deliver? Int J Gynaecol Obstet 2009 107 Suppl 1 S123 40 10.1016/j.ijgo.2009.07.021 19815205 5. Clapham S The evolution of a quality of care approach for improving emergency obstetric care in rural hospitals in Nepal Int J Gynaecol Obstet 2004 86 1 86 97 10.1016/j.ijgo.2004.03.013 15207689 6. Gabrysch S Campbell OM Still too far to walk: literature review of the determinants of delivery service use BMC Pregnancy Childbirth 2009 9 34 10.1186/1471-2393-9-34 19671156 7. Karkee R Binns CW Lee AH Determinants of facility delivery after implementation of safer mother programme in Nepal: a prospective cohort study BMC Pregnancy Childbirth 2013 13 193 10.1186/1471-2393-13-193 24139382 8. Gao Y Barriers to increasing hospital birth rates in rural Shanxi Province, China Reprod Health Matters 2010 18 36 35 45 10.1016/S0968-8080(10)36523-2 21111349 9. Garg R Study on delivery practices among rural women in Punjab Health Popul Perspect Issues 2010 33 1 23 33 10. Institute of Medicine Study On Utilization of Emergency Obstetric Care (EmOC) in selected districts of Nepal 2004 Kathmandu Institute of Medicine 11. Brunson J Confronting maternal mortality, controlling birth in Nepal: the gendered politics of receiving biomedical care at birth Soc Sci Med 2010 71 10 1719 27 10.1016/j.socscimed.2010.06.013 20713304 12. Shrestha SK Changing trends on the place of delivery: why do Nepali women give birth at home? Rep Health 2012 9 25 10.1186/1742-4755-9-25 13. Thaddeus S Maine D Too far to walk: maternal mortality in context Soc Sci Med 1994 38 8 1091 110 10.1016/0277-9536(94)90226-7 8042057 14. Morrison J Exploring the first delay: a qualitative study of home deliveries in Makwanpur district Nepal BMC Pregnancy Childbirth 2014 14 1 1 7 10.1186/1471-2393-14-89 24383788 15. Sorensen G Model for incorporating social context in health behavior interventions: applications for cancer prevention for working-class, multiethnic populations Prev Med 2003 37 3 188 97 10.1016/S0091-7435(03)00111-7 12914824 16. Crabtree BF Miller WL Doing qualitative research 1999 Thousand Oaks Sage Publications 17. Engel J Nepal's Story: Understanding improvements in maternal health 2013 London Overseas Development Institute 18. Mwaliko E “Not too far to walk”: the influence of distance on place of delivery in a western Kenya health demographic surveillance system BMC Health Serv Res 2014 14 212 10.1186/1472-6963-14-212 24884489 19. McKinnon B Distance to emergency obstetric services and early neonatal mortality in Ethiopia Trop Med Int Health 2014 19 7 780 90 10.1111/tmi.12323 24750556 20. Hussein J The effectiveness of emergency obstetric referral interventions in developing country settings: a systematic review PLoS Med 2012 9 7 10.1371/journal.pmed.1001264 22807658 21. Ministry of Health and Population New ERA ICF International Inc Nepal Demographic and Health Survey 2011 2012 Kathmandu Ministry of Health and Population, New ERA, ICF International Inc

PMC005xxxxxx/PMC5002329.txt

==== Front Cardiovasc DiabetolCardiovasc DiabetolCardiovascular Diabetology1475-2840BioMed Central London 44010.1186/s12933-016-0440-3ReviewGlucagon and heart in type 2 diabetes: new perspectives Ceriello Antonio +34 932275400aceriell@clinic.ub.es 12Genovese Stefano stefano.genovese@multimedica.it 2Mannucci Edoardo edoardo.mannucci@unifi.it 3Gronda Edoardo edoardo.gronda@multimedica.it 21 Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS) and Centro de Investigación Biomedica en Red de Diabetes y Enfermedades Metabólicas Asociadas (CIBERDEM), C/Rosselló, 149-153, 08036 Barcelona, Spain 2 Department of Cardiovascular and Metabolic Diseases, IRCCS Multimedica, Sesto San Giovanni, MI Italy 3 Diabetology, Careggi Hospital, University of Florence, Florence, Italy 27 8 2016 27 8 2016 2016 15 1 12328 6 2016 13 8 2016 © The Author(s) 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Increased levels of glucagon in type 2 diabetes are well known and, until now, have been considered deleterious. However, glucagon has an important role in the maintenance of both heart and kidney function. Moreover, in the past, glucagon has been therapeutically used for heart failure treatment. The new antidiabetic drugs, dipeptidyl peptidase-4 inhibitors and sodium-glucose co-transporter-2 inhibitors, are able to decrease and to increase glucagon levels, respectively, while contrasting data have been reported regarding the glucagon like peptide 1 receptors agonists. The cardiovascular outcome trials, requested by the FDA, raised some concerns about the possibility that the dipeptidyl peptidase-4 inhibitors can precipitate the heart failure, while, at least for empagliflozin, a positive effect has been shown in decreasing both cardiovascular death and heart failure. The recent LEADER Trial, showed a significant reduction of cardiovascular death with liraglutide, but a neutral effect on heart failure. A possible explanation of the results with the DPPIV inhibitors and empagliflozin might be related to their divergent effect on glucagon levels. Due to unclear effects of glucagon like peptide 1 receptor agonists on glucagon, the possible role of this hormone in the Leader trial remains unclear. Keywords T2DGlucagonHeart failureissue-copyright-statement© The Author(s) 2016 ==== Body Glucagon and glucose metabolism Glucagon was identified as a pancreatic contaminant at the time of the discovery of insulin; it received the name of glucagon (GLUCose AGONist substance) because it was thought to be a glucose agonist [1]. This “long-known opponent” [2] of insulin in glucose homeostasis is a 29-amino acid hormone secreted by the α-cells of the pancreas. Its secretion is strictly correlated to the blood glucose levels: low levels of blood glucose in the fasting state determine secretion of glucagon and inhibit insulin secretion. Consequently, glucagon secretion restores glucose levels through hepatic glycogenolysis and gluconeogenesis, along with inhibition of glycogenesis. Glucagon secretion is regulated by insulin and somatostatin, (as the main paracrine/endocrine inhibitors), glucose, glucagon-like peptide-1 (GLP-1), amylin, leptin, fatty acids, ketone bodies—all inhibiting glucagon secretion, glucose-dependent insulinotropic peptide (GIP), amino-acids (as l-arginine, leucine)—stimulating glucagon secretion, and by the autonomic nervous system. Meier et al. [3] demonstrated that also glucagon-like peptide-2 (GLP-2) stimulates glucagon secretion. Some of the drugs often used in patients with type 2 diabetes (T2D), such as furosemide or acetylsalicylic acid, may influence prostaglandins (PG) synthesis, mainly PGE, which in turn control glucagon release [4]. Also, stressful stimuli, as hypovolemia, stimulate glucagon secretion [5]. It has been known for a long time that large meals containing only proteins increase whereas meals rich in carbohydrates decrease, glucagon secretion [6]. Insulin modulates glucagon secretion while physiological levels of glucagon stimulate insulin secretion [1, 2, 7]. Glucagon receptors are mainly located in the liver (hepatocytes and Kupfer cells) and kidney. They may be found also, at a lesser degree, in the endocrine pancreas (β- and α-cells), heart, adipocytes, gastrointestinal (GI) tract, brain, adrenal glands, lymphoblasts, retina and placenta [1, 2, 8]. Intense physical exercise, hypercorticism and stimulation of the ventro-medial hypothalamus determine an increase of glucagon secretion [6]. Glucagon beyond glucose metabolism Glucagon has been used not only as an inotropic agent and vasodilator but also as an inhibitor of the smooth muscle activity of the GI tract (including as an aid for radiological examinations) [9]. Pleiotropic actions of glucagon Today, besides its actions on glucose metabolism, glucagon is known to have other relevant effects [2, 3, 8]:On lipid metabolism: decreased plasma cholesterol, total esterified fatty acids, decreased hepatic synthesis of triglycerides and apolipoproteins. In addition, glucagon determines lipolysis in white adipose tissue. Increased ketone-body production and fatty acid oxidation. Increased energy expenditure and thermogenesis by an increased oxygen consumption, blood flow and heat production in brown adipose tissue. Decreased food intake, with reduction of meal size, increased satiation (possibly mediated through ghrelin) and decreased gastric emptying. Regulation of secretion of other hormones, such as insulin, ghrelin, somatostatin, cortisol and growth hormone. Retinal function: loss of retinal function and of visual acuity and retina cells death have been described after the disruption of glucagon receptor (gcgr) gene and were directly correlated with the degree of hypoglycaemia in rodent models. The above-described effects make glucagon an attractive therapeutic target in obesity, eating and lipid disorders [3]. Glucagon and the heart Along with the key actions of glucagon in glucose homeostasis, since the 60s, (after Unger adapted the radioimmunoassay to measure glucagon), human and animal experiments have highlighted direct actions on the heart of physiological levels of glucagon [3, 7]. In humans it has been proved that acute glucagon administration exerts a positive action on cardiovascular performance either increasing cardiac index, either decreasing peripheral vascular resistances [10]. Moreover, the positive enhancement of cardiovascular performance is comparable to what has been observed in cat and dog, with persistence of action despite beta-receptor blockade with propranolol [11]. In the non-failing heart, glucagon determines a rise in heart rate, almost without changes in cardiac output and auricular pressure; in the failing heart, it increases heart rate and cardiac output, together with a dose-dependent increase in coronary blood flow and oxygen consumption [12]. These actions of glucagon are mediated through cyclic AMP (cAMP). Several factors affect glucagon actions on the heart. Among them, as seen above, the most important for our topic is heart failure (HF). In this context, the type of HF, its severity and chronicity are relevant [12, 13]. In fact, a greater severity of HF is associated with a smaller hemodynamic response to glucagon. In addition, glucagon determines a weaker response in chronic than in acute HF [12, 13]. The inotropic effects of glucagon are more potent in the ventricle than in the atrium, [7]. It does not increase irritability of the myocardium and it is active in the presence of digitalis and propranolol. Thus, attempts for a therapeutic use of glucagon have been made in resistant cardiac failure, myocardial infarction, hypotension following cardiac operations, intoxication with β-/calcium channel blockers and heart block [7, 14–16]. Glucagon also facilitates the atrio-ventricular conduction and its inotropic action is accompanied by an anti-arrhythmogenic effect (which, might be due partially to an increased insulin-mediated uptake of potassium (K+) by the myocardium following glucagon administration) [12, 14]. Glucagon has an important effect on sino atrial node rate [17] and the antiarrhythmogenic effect of glucagon has been several times reviewed [18–23]. Furthermore, glucagon is able to decrease histamine-induced cardiac injury during reperfusion [2] and restores the pressure of the coronary perfusion during ischemic vasodilation [24]. As an inotropic agent, glucagon increases the work of the heart and, consequently, it increases oxygen consumption, lipolysis and beta-oxidation of lipids [1]. It is noteworthy that both insulin and glucagon increase fuel availability in the heart. In animal studies, glucagon, similarly to insulin, increases glycolysis and glucose oxidation through phosphatidylinositol 3-kinase-dependent and adenylate cyclase- and cAMP-independent pathways [7]. As hyperglucagonemia determines an increased availability of substrate and improves cardiovascular (CV) performance (crucial in a physiological stress response), this hormone is considered today to be a stress hormone [5]. Glucagon and the kidney In the kidneys, glucagon at relatively high doses, induces vasodilation with a concomitant increase in renal plasma flow (RPF), glomerular filtration rate (GFR) and electrolyte excretion. These changes are more evident in patients with diabetes, possibly due to the modified insulin:glucagon ratio; the administration of insulin, which normalizes this ratio, brings GFR and RPF close to normal values. In respect to the electrolytes, glucagon is responsible of the increase of natriuresis in the fasting state. During starvation glucagon is increased and insulin is decreased; re-feeding with carbohydrates has an anti-natriuretic effect [12]. Glucagon determines initially a transient increase in plasma K+ levels (partially due to the hepatic glycogenolysis), followed by hypopotassemia determined by an increased uptake (muscles, liver) induced by insulin. Glucagon also increases the urinary excretion of calcium, phosphate and zinc [12]. The direct action of glucagon (and vasopressin) in protein-induced hyperfiltration is well established now [25]. Glucagon plays an important role in the excretion of nitrogen end products (increased urea synthesis in the liver and urea excretion in the kidney), while vasopressin concentrates these products in a hyperosmotic urine and producing water economy. In the absence of any of these two hormones, glomerular hyperfiltration is not possible [25]. In conclusion, glucagon is a relevant component of the close relationship between heart and kidney function, aiming to maintain the continuum pressure volume circulatory balance. Glucagon in type 2 diabetes Schematically, type 2 diabetes (T2D) is characterized by: β-cell failure, α-cells insulin resistance and decreased incretin effect.β-cell failure due to partial loss of β-cell mass and β-cell dysfunction, influenced by a genetic background and by chronic exposure to gluco- and lipotoxicity, amylin and advanced glycation end-products (AGEs). α-cells insulin resistance the so called by Unger and Orci, “paracrinopathy” and T2D “a bi-hormonal disorder”. In T2D, α-cells might be resistant to the inhibitory effect of insulin [26] or to other β-cell secretory products such as zinc or γ-aminobutyric acid [7]. Consequently, T2D is characterized by fasting hyperglucagonemia and impaired glucose-induced glucagon suppression in the post-prandial state (insulin/glucagon concentration inversely related in the post-prandial state). Mainly due to β-cell apoptosis, β/α-cell ratio is altered, contributing to a decreased insulin/glucagon ratio. Also, β-cell may de-differentiate to progenitor pluripotent cells that may release glucagon and somatostatin, thus further decreasing insulin/glucagon ratio [3, 7]. In this context, it is clear that glucagon is a key hormone worsening the metabolic consequences of insulin deficiency [27]. Type 2 diabetes is also characterized by a decreased incretin effect: in T2D, glucose-dependent insulinotropic peptide (GIP) and GLP-1 account only for <20 % of the postprandial insulin response. While GLP-1 action is relatively preserved, β-cell impairment determines a decrease in the insulinotropic action of GIP; the following hyperglycaemia further downregulates the GIP receptor in β-cells, aggravating the impairment of the incretin effect, creating a vicious cycle [1]. A decrease in incretin action may affect the crosstalk between AGEs-RAGE axis, favouring the appearance of complications [28]. It is worthy of interest that when hepatic glycogen content is low, glucagon response to insulin-induced hypoglycemia is reduced [29]. In this view, glucagon has been recently suggested as the key feature of type 2 diabetes [30]. New antidiabetes treatments and glucagon The increased glucagon/insulin ratio is a key factor in the pathogenesis of hyperglycaemia in T2D; the lack of glucagon suppression in the post-prandial state leads to post-prandial hyperglycaemia, whereas inappropriate levels of glucagon in the fasting state leads to increased hepatic glucose production and fasting hyperglycaemia. Consequently, addressing glucagon seemed an attractive treatment for T2D by either suppression of glucagon secretion or by blocking gcgr. Monoclonal glucagon antibodies, glucagon receptor antagonists (peptide and non-peptide) and molecules targeting the expression of gcgr have all been tested as potential treatments for T2D [7]. However, research on those agents encountered a number of major obstacles, most notably a limited efficacy, the risk of iatrogenic hypoglycaemia, and other safety issues related to lack of specificity of glucagon blockade, immunogenity and toxicity [31]. Interestingly, the reduction of glucagon levels should be one of the main mechanisms of action of some of the recent antidiabetes drugs (ADD) for T2D, such as GLP-1 receptor agonists (GLP-1 RA) and inhibitors of dipeptidyl peptidase-4 (DPP-4i) [31]. GLP-1RA increase insulin synthesis and secretion in a glucose-dependent manner and decrease glucagon levels probably; this latter effect could be exerted through somatostatin or neural regulation, as the α-cell do not show GLP-1 receptors [1, 31]. Additionally, GLP1-RA slow gastric empting, decrease appetite, induce satiety and potentially, inhibit β-cell apoptosis [32]. They have favorable effects on body weight and body composition (decreasing mainly the visceral fat) and plasma lipids. In animal models, long-term treatment with GLP1-RA also increase β-cell mass [33, 34]. However, it has been shown that glucagon and GLP-1 co-agonism has a significant greater efficacy than GLP-1 RA monotherapy on body weight, body composition, glucose and lipid metabolism, including the reversal of hepatic steatosis, making this combination therapy an attractive and promising treatment for obesity and metabolic syndrome [2]. Short-acting GLP-1 RA (exenatide and lixisenatide) lower mainly post-prandial glucose, partly by inhibiting gastric empting: conversely longer-acting molecules of the class (albiglutide, dulaglutide, exenatide LAR, liraglutide) target predominantly fasting plasma glucose through their insulinotropic and glucagonostatic actions [33, 34]. The durability of the glucagonostatic effect in the long-term treatment with GLP-1 RA has been investigated for liraglutide in the LIBRA trial. Quite surprisingly, chronic treatment with liraglutide in patients with early T2D has been associated with an increased post-challenge glucagonemia [35]. The real effects of GLP-1RA on glucagon probably remains to be further elucidated, considering that also Albiglutide does not affect glucagon secretion during hypoglycemia [36]. DPP-4i increase the duration of action of the endogenous GLP-1 through the inhibition of DPP-4 enzyme. Consequently, the GLP-1 levels are increased although the stimulation of the GLP-1 axis is not so potent as that determined by GLP1-RA. This action is sufficient to determine a moderate reduction of blood glucose, with no effects on body weight, no risk of severe hypoglycemia, and no gastrointestinal side effects. DPP-4i also induce an increase in circulating levels of GIP, which is another one of the substrates of DPP-4. The reduction in glucagon levels observed during treatment with DPP-4i results from the increase of both GLP-1 and GIP [34, 37]. Worth to mention, mainly in the context of this article, is that the incretins have beneficial effects on the CV system by decreasing blood pressure, improving left ventricular function and endothelium-dependent vasodilation and by increasing the endothelial progenitor cells [38]. SGLT-2i are the last approved class of ADD, which inhibit glucose reabsorption in the proximal renal tubule, independently of insulin secretion or action. Consequently, glucose is lost in the urine and glycaemia is decreased [34]. Regarding glucagon secretion, in opposition with GLP-1 RA and DPP-4i, SGLT-2i increase plasma glucagon. In the kidney, glucagon controls the rate of filtration, urea excretion and the reabsorption of water through direct and indirect mechanisms, even though its role in renal glucose output is unclear. In an animal model, long-term infusion of glucagon induced hypertension, hypertrophy, and increased proliferation of mesangial cells [7]. A non-selective inhibitor of glucose reabsorption in the proximal nephrons which targets both SGLT-1 and SGLT-2, phlorizin, was increasing glucagon levels, an effect confirmed by SGLT-2i [3]. The association between SGLT-2 inhibition, increased endogenous glucose production (EGP) and glucagon during euglycaemia has not been expected. It has been demonstrated that the SGLT-2 transporters are expressed not only in the proximal tubules but also in the α-cells, but not in β-cells. The inhibition of SGLT-2 directly stimulates glucagon secretion through the activation of KATP channel [39, 40]. Since the mechanism of action of SGLT-2i is independent of β-cells and different from that of the other existing classes of ADD, it has been suggested that SGLT-2i might be combined with other drug(s) and mainly with DPP4i; this combination therapy is well-tolerated and not accompanied by the most common seen side-effects: hypoglycaemia or weight gain. Also, DPP4i increase insulin secretion and decrease glucagon; so, potentially they block the compensatory EGP seen with SGLT-2i and increase their glucose-lowering capacity [38, 41, 42]. Cardiovascular outcome trials and glucagon After the CV safety concerns raised by some ADD either in development or marketed (e.g. muraglitazar, rosiglitazone) and by the results of some clinical trials (e.g. ACCORD), the US Food and Drug Administration (FDA) issued in 2008 the “FDA Guidance for Industry” focusing on the CV risk assessment of ADD. Consequently, post-approval, there is the requirement to perform a CV outcome trial (CVOT) [43, 44], which has as primary endpoint major adverse cardiac events (MACE) such as: CV death, non-fatal MI, and non-fatal stroke—the “canonical” CVD endpoints; some CV outcome trials include also heart failure, unstable angina requiring hospitalization, amputation, and revascularization procedures—the “hard” CVD endpoints [45]. It should be mentioned that the actual “FDA Guidance for Industry”, issued in 2008, is not requiring hospitalization for HF as a primary/secondary outcome parameter. In Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus (SAVOR)—Thrombolysis in Myocardial Infarction (TIMI) 53 (SAVOR-TIMI 53) trial the major secondary endpoint was a composite of CV death, myocardial infarction (MI), stroke, hospitalization for unstable angina, coronary revascularization or HF. A total unexpected finding was the increased hospitalization for HF in T2D subjects on saxagliptin versus placebo (HR 1.27, 95 % CI 1.07–1.51, p = 0.007) [46]. In the primary publication of the Examination of Cardiovascular Outcomes with Alogliptin versus Standard of Care (EXAMINE) trial, despite the fact that 28 % of T2D subjects had congestive HF at baseline, no information on HF outcomes has been reported [47, 48]. Unfortunately, hospitalization for HF was not planned as a stand-alone endpoint in the EXAMINE, so that a specific analysis on this outcome could not be reported in the primary publication. (Hospitalization for heart failure 106 out of 2701 events in the alogliptin arm compared to 89 out of 2679 events in the placebo arm: HR 1.19, 95 % CI 0.89–1.58). In a post hoc analysis [49], there was a trend (p = NS) for increased hospitalization for HF with alogliptin versus placebo. The fact that the trend toward increase of hospitalizations for HF in the overall cohort was not statistically significant is not surprising, considering the relatively small sample size of subjects in this trial. However, in the same study, an increased hospitalization for HF has been reported in patients without a previous history of HF treated with alogliptin, but not in patients with a previous history of HF [49]. Standl and Schnell show the limitation of this post hoc analysis (different composite endpoints) and point out the similarities between SAVOR TIMI 53 and EXAMINE (in both studies, in patients without history of HF and treated with DPP4i the rate of HF was increased) [50]. In addition, vildagliptin in ventricular dysfunction diabetes (VIVIDD) trial, showed an unanticipated rise in left ventricular end-diastolic volume, end-systolic volume, and cardiac stroke volume [50]. On the other hand, while an increase of HF has been also reported with sitagliptin use [51], no increase in the incidence of hospitalization for HF was observed with sitagliptin in the trial evaluating cardiovascular outcomes with sitagliptin (TECOS) trial [52, 53], or with incretin therapy in a recent observational study with a large cohort of T2D patients [54]. Noteworthy, the T2D population included in each of these outcome trials was different: if in TECOS, these subjects were with an established CVD, in SAVOR TIMI they had a history or were at risk for CVD and lastly, in EXAMINE they had an acute coronary syndrome within the previous 15–90 days [46, 47, 52]. This fact might (partly) explain the different results on the hospitalization for HF. Anyhow, recently the AHA stated that this could be a class effect [55]. Surprisingly, the EMPA-REG study with empagliflozin, a SGLT-2i, showed a CV protective effect of this compound [56]. Apart the osmotic diuresis, there are proposed many other mechanisms behind CV protection of SGLT2i: improved insulin sensitivity, decreased blood pressure and arterial stiffness, body weight and visceral adiposity, decreased inflammation and oxidative stress [40]. Both results of CVOT on hospitalization for HF in T2D unfavorable with saxagliptin in SAVOR-TIMI 53 and favorable ones with empagliflozin in EMPA-REG might be explained, at least in part, by their impact on glucagon levels: DPP-4i reduce, and SGLT-2i increase, glucagon levels [57]. The reduced risk of hospitalization for heart failure with empagliflozin [56], which is present in patients with and without baseline heart failure [58], might be partly explained by a direct enhancement of myocardial function, determined by the increased levels of glucagon, and by its natriuretic effect. In addition, the beneficial effect of glucagon on disturbances of cardiac rhythm could be partly responsible for the reduction of CV mortality with empagliflozin. Conversely, the reduction of glucagon levels during treatment with DPP4i could precipitate HF in individuals with unstable hemodynamic compensation. Indeed, in SAVOR-TIMI 53 the increased risk for hospitalization for HF with saxagliptin was observed in patients with the highest N-terminal pro B-type natriuretic peptide quartile and in those with chronic kidney disease [59], a trend confirmed also in the EXAMINE [49], while no data on N-terminal pro B-type natriuretic peptide are available for the TECOS. Furthermore, the protective effects on diabetic kidney disease of empagliflozin [60] could also be partly explained by the increase of glucagon, considering its effects on the kidney described before. It has been shown that GLP-1 RA have an insulinotropic and glucagonostatic effect [32, 34]. However, the data regarding GLP-1RA and glucagon need to be more clarified: recently, in the LIBRA trial it has been demonstrated that long-term treatment with liraglutide in early T2D has been associated with an increased post-challenge glucagonemia [35], while for Albiglutide the evidence is that it does not affect glucagon secretion during hypoglycemia [36]. Liraglutide reduced the cardiovascular mortality in the LEADER trial, but with a neutral effect on HF [61]. Both liraglutide and albiglutide were also not effective in improving HF in two recent trials in type 2 diabetes with advanced HF [62, 63]. At the state of the art, in our opinion, a possible involvement of glucagon in the results on the mortality of the LEADER trial cannot be either claimed or excluded. Conclusions The glucagon inotropic action in human heart is well represented in not failing heart, but progressively declines in the failing heart, becoming undetectable in severe heart failure condition [13]. Therefore, the plasma levels of glucagon may contribute to maintain the heart function when the HF is not severe, which is the case of people recruited in the studies we mentioned. This means that glucagon might be very important for the heart and cardiovascular system in type 2 diabetes. This is a paradox, considering that this hormone has been always considered harmful in diabetes. Future studies are needed to endorse or not the role of glucagon on heart, and may be also on kidney, in T2D. Abbreviations AGEsadvanced glycation end-products ADDantidiabetes drugs BNPbrain natriuretic peptides CVcardiovascular CVDcardiovascular disease CVOTCV outcome trial DPP-4dipeptidyl peptidase-4 DPP-4iinhibitors of dipeptidyl peptidase-4 EGPendogenous glucose production GIgastrointestinal GFRglomerular filtration rate GLP-1glucagon-like peptide-1 GLP-1 RAGLP-1 receptor agonists GLP-2glucagon-like peptide-2 Gcgrglucagon receptor GIPglucose-dependent insulinotropic peptide HFheart failure MACEmajor adverse cardiac events MImyocardial infarction PGprostaglandins RPFrenal plasma flow SGLT-2isodium-glucose co-transporter-2 inhibitors T2Dtype 2 diabetes Authors’ contributions AC made the search of the literature and drafted the manuscript. SG, EM, EG contributed to the discussion and drafted the manuscript. All authors read and approved the final manuscript. Competing interests AC received speaking and/or consulting fees and/or research grants from AstraZeneca, Bayer Healthcare, Boehringer Ingelheim, Danone, DOC Generici, Eli Lilly, Janssen, Mendor, Merck Sharp & Dohme, Mitsubishi, Novartis, Novo Nordisk, OM Pharma, Roche Diagnostics, Sanofi, Servier, Takeda and Unilever. SG received speaking and/or consulting fees and/or research grants from Abbott, AstraZeneca, Boehringer Ingelheim, Eli Lilly, Janssen, Lifescan, Merck Sharp & Dohme, Novartis, Novo Nordisk, Sanofi and Takeda. EM received speaking and/or consulting fees and/or research grants from Abbott, AstraZeneca, Boehringer Ingelheim, Eli Lilly, Janssen, Merck Sharp & Dohme, Novartis, Novo Nordisk, Sanofi, and Takeda. EG none. Search strategy and selection criteria References for this review were identified through searches of PubMed for articles published from January 1960 to August, 2016, by use of the terms “glucagon” in combination or not with the terms “heart” and “kidney”. Articles resulting from these searches and relevant references cited in those articles were reviewed. Articles published in English, French, Italian and German were included. ==== Refs References 1. Godoy-Matos AF The role of glucagon on type 2 diabetes at a glance Diabetol Metab Syndr 2014 6 91 10.1186/1758-5996-6-91 25177371 2. Habegger KM Heppner KM Geary N Bartness TJ DiMarchi R Tschöp MH The metabolic actions of glucagon revisited Nat Rev Endocrinol 2010 6 689 697 10.1038/nrendo.2010.187 20957001 3. Ahrén B Glucagon—early breakthroughs and recent discoveries Peptides 2015 67 74 81 10.1016/j.peptides.2015.03.011 25814364 4. Giugliano D Torella R D’Onofrio F Prostaglandins and the alpha-cell Prostaglandins Med 1981 6 283 297 10.1016/0161-4630(81)90152-X 7012876 5. Jones BJ Tan T Bloom SR Minireview: glucagon in stress and energy homeostasis Endocrinology 2012 153 1049 1054 10.1210/en.2011-1979 22294753 6. Sherwin R Felig P Glucagon physiology in health and disease Int Rev Physiol 1977 16 151 171 332652 7. Ali S Drucker DJ Benefits and limitations of reducing glucagon action for the treatment of type 2 diabetes Am J Physiol Endocrinol Metab 2009 296 E415 E421 10.1152/ajpendo.90887.2008 19116373 8. Charron MJ Vuguin PM Lack of glucagon receptor signaling and its implications beyond glucose homeostasis J Endocrinol 2015 224 R123 R130 10.1530/JOE-14-0614 25568163 9. Glucagon SmPC. http://www.accessdata.fda.gov/drugsatfda_docs/label/1998/20928lbl.pdf. Accessed 18 May 2016. 10. Parmley WW Glick G Sonnenblick EH Cardiovascular effects of glucagon in man N Engl J Med 1968 279 12 17 10.1056/NEJM196807042790103 4872564 11. Glick G Parmley WW Wechsler AS Sonnenblick EH Glucagon. Its enhancement of cardiac performance in the cat and dog and persistence of its inotropic action despite beta-receptor blockade with propranolol Circ Res 1968 22 789 799 10.1161/01.RES.22.6.789 5659816 12. Farah AE Glucagon and the circulation Pharmacol Rev 1983 35 181 217 6318231 13. Goldstein RE Skelton CL Levey GS Glancy DL Beiser GD Epstein SE Effects of chronic heart failure on the capacity of glucagon to enhance contractility and adenyl cyclase activity of human papillary muscles Circulation 1971 44 638 648 10.1161/01.CIR.44.4.638 5094143 14. [No authors listed]. Glucagon and the heart. BMJ. 1969;4:635. 15. Murtagh JG Binnion PF Lal S Hutchison KJ Fletcher E Haemodynamic effects of glucagon Br Heart J 1970 32 307 315 10.1136/hrt.32.3.307 5420074 16. Regan TJ Lehan PH Henneman DH Behar A Hellems HK Myocardial, metabolic and contractile response to glucagon and epinephrine J Lab Clin Med 1964 63 638 647 14155451 17. Furukawa Y Saegusa K Ogiwara Y Chiba S Different effectiveness of glucagon on the pacemaker activity and contractility in intact dog hearts and in isolated perfused right atria Jpn Heart J 1986 27 215 222 10.1536/ihj.27.215 3723799 18. Madan BR Effect of glucagon on ventricular arrhythmias after coronary artery occlusion and on ventricular automaticity in the dog Br J Pharmacol 1971 43 279 286 5158198 19. Singh J Bala S Kaur AH Garg KN Effect of glucagon on arrhythmias induced by coronary artery occlusion and ouabain in dogs Indian J Physiol Pharmacol 1980 24 329 334 7216388 20. Wilkerson RD Partlow DB Pruett JK Patterson CW A possible mechanism of the antiarrhythmic action of glucagon Eur J Pharmacol 1977 44 57 63 10.1016/0014-2999(77)90116-9 885159 21. Scholz H Inotropic drugs and their mechanisms of action J Am Coll Cardiol 1984 4 389 397 10.1016/S0735-1097(84)80231-4 6330195 22. Cohn KE Agmon J Gamble OW The effect of glucagon on arrhythmias due to digitalis toxicity Am J Cardiol 1970 25 683 689 10.1016/0002-9149(70)90619-3 4192693 23. Parmley WW Sonnenblick EH Glucagon: a new agent in cardiac therapy Am J Cardiol 1971 27 298 303 10.1016/0002-9149(71)90307-9 5543916 24. Rosic M Pantovic S Rosic G Tomic-Lucic A Labudovic T Zivkovic V Jakovljevic V Glucagon effects on ischemic vasodilatation in the isolated rat heart J Biomed Biotechnol 2010 2010 231832 10.1155/2010/231832 20339509 25. Bankir L Roussel R Bouby N Protein- and diabetes-induced glomerular hyperfiltration: role of glucagon, vasopressin, and urea Am J Physiol Renal Physiol 2015 309 F2 F23 10.1152/ajprenal.00614.2014 25925260 26. Unger RH Orci L The essential role of glucagon in the pathogenesis of diabetes mellitus Lancet 1975 1 14 16 10.1016/S0140-6736(75)92375-2 46337 27. Gerich JE Physiology of glucagon Int Rev Physiol 1981 24 243 275 7021445 28. Yamagishi S Fukami K Matsui T Crosstalk between advanced glycation end products (AGEs)-receptor RAGE axis and dipeptidyl peptidase-4-incretin system in diabetic vascular complications Cardiovasc Diabetol 2015 14 2 10.1186/s12933-015-0176-5 25582643 29. Winnick JJ Kraft G Gregory JM Edgerton DS Williams P Hajizadeh IA Kamal MZ Smith M Farmer B Scott M Neal D Donahue EP Allen E Cherrington AD Hepatic glycogen can regulate hypoglycemic counterregulation via a liver-brain axis J Clin Invest 2016 126 2236 2248 10.1172/JCI79895 27140398 30. Unger RH Cherrington AD Glucagonocentric restructuring of diabetes: a pathophysiologic and therapeutic makeover J Clin Invest 2012 122 4 12 10.1172/JCI60016 22214853 31. Moon JS Won KC Pancreatic α-cell dysfunction in type 2 diabetes: old kids on the block Diabetes Metab J 2015 39 1 9 10.4093/dmj.2015.39.1.1 25729706 32. Meier JJ GLP-1 receptor agonists for individualized treatment of type 2 diabetes mellitus Nat Rev Endocrinol 2012 8 728 742 10.1038/nrendo.2012.140 22945360 33. Retnakaran R Kramer CK Choi H Swaminathan B Zinman B Liraglutide and the preservation of pancreatic β-cell function in early type 2 diabetes: the LIBRA trial Diabetes Care 2014 37 3270 3278 10.2337/dc14-0893 25249651 34. Fonseca VA New developments in diabetes management: medications of the 21st century Clin Ther 2014 36 477 484 10.1016/j.clinthera.2014.01.018 24594068 35. Kramer CK Zinman B Choi H Connelly PW Retnakaran R The impact of chronic liraglutide therapy on glucagon secretion in type 2 diabetes: insight from the LIBRA trial J Clin Endocrinol Metab 2015 100 3702 3709 10.1210/jc.2015-2725 26230296 36. Hompesch M Jones-Leone A Carr MC Matthews J Zhi H Young M Morrow L Reinhardt RR Albiglutide does not impair the counter-regulatory hormone response to hypoglycaemia: a randomized, double-blind, placebo-controlled, stepped glucose clamp study in subjects with type 2 diabetes mellitus Diabetes Obes Metab 2015 17 82 90 10.1111/dom.12398 25263215 37. Ahrén B Landin-Olsson M Jansson PA Svensson M Holmes D Schweizer A Inhibition of dipeptidyl peptidase-4 reduces glycemia, sustains insulin levels, and reduces glucagon levels in type 2 diabetes J Clin Endocrinol Metab 2004 89 2078 2084 10.1210/jc.2003-031907 15126524 38. Sharma MD Potential for combination of dipeptidyl peptidase-4 inhibitors and sodium-glucose co-transporter-2 inhibitors for the treatment of type 2 diabetes Diabetes Obes Metab 2015 17 616 621 10.1111/dom.12451 25690671 39. Bonner C Kerr-Conte J Gmyr V Queniat G Moerman E Thévenet J Beaucamps C Delalleau N Popescu I Malaisse WJ Sener A Deprez B Abderrahmani A Staels B Pattou F Inhibition of the glucose transporter SGLT2 with dapagliflozin in pancreatic alpha cells triggers glucagon secretion Nat Med 2015 21 512 517 10.1038/nm.3828 25894829 40. Inzucchi SE Zinman B Wanner C Ferrari R Fitchett D Hantel S Espadero RM Woerle HJ Broedl UC Johansen OE SGLT-2 inhibitors and cardiovascular risk: proposed pathways and review of ongoing outcome trials Diab Vasc Dis Res 2015 12 90 100 10.1177/1479164114559852 25589482 41. Ferrannini E Muscelli E Frascerra S Baldi S Mari A Heise T Broedl UC Woerle HJ Metabolic response to sodium-glucose cotransporter 2 inhibition in type 2 diabetic patients J Clin Invest 2014 124 499 508 10.1172/JCI72227 24463454 42. Merovci A Solis-Herrera C Daniele G Eldor R Fiorentino TV Tripathy D Xiong J Perez Z Norton L Abdul-Ghani MA DeFronzo RA Dapagliflozin improves muscle insulin sensitivity but enhances endogenous glucose production J Clin Invest 2014 124 509 514 10.1172/JCI70704 24463448 43. FDA Guidance for Industry. Diabetes mellitus—evaluating cardiovascular risk in new antidiabetic therapies to treat type 2 diabetes. http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/ucm071627.pdf. Accessed 16 Mar 2016. 44. Schnell O Standl E Catrinoiu D Genovese S Lalic N Skra J Valensi P Ceriello A Report from the 1st cardiovascular outcome trial (CVOT) summit of the diabetes & cardiovascular disease (D&CVD) EASD study group Cardiovasc Diabetol 2016 15 33 10.1186/s12933-016-0357-x 26892706 45. Ferrannini E DeFronzo RA Impact of glucose-lowering drugs on cardiovascular disease in type 2 diabetes Eur Heart J 2015 36 2288 2296 10.1093/eurheartj/ehv239 26063450 46. Scirica BM Bhatt DL Braunwald E Steg PG Davidson J Hirshberg B Ohman P Frederich R Wiviott SD Hoffman EB Cavender MA Udell JA Desai NR Mosenzon O McGuire DK Ray KK Leiter LA Raz I SAVOR-TIMI 53 Steering Committee and Investigators Saxagliptin and cardiovascular outcomes in patients with type 2 diabetes mellitus N Engl J Med 2013 369 1317 1326 10.1056/NEJMoa1307684 23992601 47. White WB Cannon CP Heller SR Nissen SE Bergenstal RM Bakris GL Perez AT Fleck PR Mehta CR Kupfer S Wilson C Cushman WC Zannad F EXAMINE Investigators Alogliptin after acute coronary syndrome in patients with type 2 diabetes N Engl J Med 2013 369 1327 1335 10.1056/NEJMoa1305889 23992602 48. Standl E Saxagliptin, alogliptin, and cardiovascular outcomes N Engl J Med 2014 370 483 10.1056/NEJMc1313880 24476445 49. Zannad F Cannon CP Cushman WC Bakris GL Menon V Perez AT Fleck PR Mehta CR Kupfer S Wilson C Lam H White WB EXAMINE Investigators Heart failure and mortality outcomes in patients with type 2 diabetes taking alogliptin versus placebo in EXAMINE: a multicentre, randomised, double-blind trial Lancet 2015 385 2067 2076 10.1016/S0140-6736(14)62225-X 25765696 50. Standl E Schnell O DPP-4 inhibitors and risk of heart failure EXAMINEd Lancet 2015 385 2022 2024 10.1016/S0140-6736(15)60037-X 25765695 51. Weir DL McAlister FA Senthilselvan A Minhas-Sandhu JK Eurich DT Sitagliptin use in patients with diabetes and heart failure: a population-based retrospective cohort study JACC Heart Fail 2014 2 573 582 10.1016/j.jchf.2014.04.005 24998080 52. Green JB Bethel MA Armstrong PW Buse JB Engel SS Garg J Josse R Kaufman KD Koglin J Korn S Lachin JM McGuire DK Pencina MJ Standl E Stein PP Suryawanshi S Van de Werf F Peterson ED Holman RR TECOS Study Group Effect of sitagliptin on cardiovascular outcomes in type 2 diabetes N Engl J Med 2015 373 232 242 10.1056/NEJMoa1501352 26052984 53. McGuire DK Van de Werf F Armstrong PW Standl E Koglin J Green JB Bethel MA Cornel JH Lopes RD Halvorsen S Ambrosio G Buse JB Josse RG Lachin JM Pencina MJ Garg J Lokhnygina Y Holman RR Peterson ED Trial Evaluating Cardiovascular Outcomes With Sitagliptin (TECOS) Study Group Association between sitagliptin use and heart failure hospitalization and related outcomes in type 2 diabetes mellitus: secondary analysis of a randomized clinical trial JAMA Cardiol 2016 1 126 135 10.1001/jamacardio.2016.0103 27437883 54. Filion KB Azoulay L Platt RW Dahl M Dormuth CR Clemens KK Hu N Paterson JM Targownik L Turin TC Udell JA Ernst P CNODES Investigators A multicenter observational study of incretin-based drugs and heart failure N Engl J Med 2016 24 374 1145 1154 10.1056/NEJMoa1506115 27007958 55. Page RL, O'Bryant CL, Cheng D, Dow TJ, Ky B, Stein CM, Spencer AP, Trupp RJ, Lindenfeld J, American Heart Association Clinical Pharmacology and Heart Failure and Transplantation Committees of the Council on Clinical Cardiology, Council on Cardiovascular Surgery and Anesthesia, Council on Cardiovascular and Stroke Nursing, Council on Quality of Care and Outcomes Research. Drugs that may cause or exacerbate heart failure: a scientific statement from the american heart association. Circulation. 2016;134(6):e32–69. doi:10.1161/CIR.0000000000000426 56. Zinman B Wanner C Lachin JM Fitchett D Bluhmki E Hantel S Mattheus M Devins T Johansen OE Woerle HJ Broedl UC Inzucchi SE EMPA-REG OUTCOME Investigators Empagliflozin, cardiovascular outcomes, and mortality in type 2 diabetes N Engl J Med 2015 373 2117 2128 10.1056/NEJMoa1504720 26378978 57. Ceriello A Genovese S Mannucci E Gronda E Understanding EMPA-REG OUTCOME Lancet Diabetes Endocrinol 2015 3 929 930 10.1016/S2213-8587(15)00426-X 26590680 58. Fitchett D Zinman B Wanner C Lachin JM Hantel S Salsali A Johansen OE Woerle HJ Broedl UC Inzucchi SE EMPA-REG OUTCOME® trial investigators Heart failure outcomes with empagliflozin in patients with type 2 diabetes at high cardiovascular risk: results of the EMPA-REG OUTCOME® trial Eur Heart J 2016 37 1526 1534 10.1093/eurheartj/ehv728 26819227 59. Scirica BM Braunwald E Raz I Cavender MA Morrow DA Jarolim P Udell JA Mosenzon O Im K Umez-Eronini AA Pollack PS Hirshberg B Frederich R Lewis BS McGuire DK Davidson J Steg PG Bhatt DL SAVOR-TIMI 53 Steering Committee and Investigators Heart failure, saxagliptin, and diabetes mellitus: observations from the SAVOR-TIMI 53 randomized trial Circulation 2014 130 1579 1588 10.1161/CIRCULATIONAHA.114.010389 25189213 60. Wanner C Inzucchi SE Lachin JM Fitchett D von Eynatten M Mattheus M Johansen OE Woerle HJ Broedl UC Zinman B EMPA-REG OUTCOME Investigators Empagliflozin and progression of kidney disease in type 2 diabetes N Engl J Med 2016 375 323 334 10.1056/NEJMoa1515920 27299675 61. Marso SP Daniels GH Brown-Frandsen K Kristensen P Mann JF Nauck MA Nissen SE Pocock S Poulter NR Ravn LS Steinberg WM Stockner M Zinman B Bergenstal RM Buse JB LEADER Steering Committee on behalf of the LEADER Trial Investigators Liraglutide and cardiovascular outcomes in type 2 diabetes N Engl J Med 2016 375 311 322 10.1056/NEJMoa1603827 27295427 62. Margulies KB Hernandez AF Redfield MM Givertz MM Oliveira GH Cole R Mann DL Whellan DJ Kiernan MS Felker GM McNulty SE Anstrom KJ Shah MR Braunwald E Cappola TP NHLBI Heart Failure Clinical Research Network Effects of liraglutide on clinical stability among patients with advanced heart failure and reduced ejection fraction: a randomized clinical trial JAMA 2016 316 500 508 10.1001/jama.2016.10260 27483064 63. Lepore JJ Olson E Demopoulos L Haws T Fang Z Barbour AM Fossler M Davila-Roman VG Russell SD Gropler RJ Effects of the novel long-acting GLP-1 Agonist, albiglutide, on cardiac function, cardiac metabolism, and exercise capacity in patients with chronic heart failure and reduced ejection fraction JACC Heart Fail 2016 4 559 566 10.1016/j.jchf.2016.01.008 27039125

PMC005xxxxxx/PMC5002330.txt

==== Front J Transl MedJ Transl MedJournal of Translational Medicine1479-5876BioMed Central London 101110.1186/s12967-016-1011-9ResearchMicro-RNA-21 (biomarker) and global longitudinal strain (functional marker) in detection of myocardial fibrotic burden in severe aortic valve stenosis: a pilot study Fabiani Iacopo +39-050995237iacopofabiani@gmail.com 1Scatena Cristian cristian.scatena@email.it 2Mazzanti Chiara Maria c.mazzanti@fondazionepisascienza.org 3Conte Lorenzo dr.contelorenzo@yahoo.it 1Pugliese Nicola Riccardo n.r.pugliese88@gmail.com 1Franceschi Sara sara.franceschi@fondazionepisascienza.org 3Lessi Francesca f.lessi@fondazionepisascienza.org 3Menicagli Michele m.menicagli@fondazionepisascienza.org 3De Martino Andrea and.demartino@libero.it 4Pratali Stefano stefanopratali@virgilio.it 4Bortolotti Uberto uberto.bortolotti@med.unipi.it 4Naccarato Antonio Giuseppe salvatore.lacarrubba@gmail.com 2La Carrubba Salvatore giuseppe.naccarato@med.unipi.it 5Di Bello Vitantonio dibelvitantonio@gmail.com 11 Department of Surgical, Medical, Molecular Pathology and Critical Area, Cisanello Hospital, University of Pisa/A.O.U.P, Via Paradisa, 2, 56100 Pisa, PI Italy 2 Department of Translational Research and New Technologies in Medicine and Surgery, University of Pisa, 56100 Pisa, Italy 3 Fondazione Pisana per la Scienza, 56100 Pisa, Italy 4 Division of Cardiac Surgery, Department of Surgical, Medical, Molecular Pathology and Critical Area, Cisanello Hospital, University of Pisa/A.O.U.P, 56100 Pisa, Italy 5 Division of Internal Medicine, Villa Sofia Hospital, 90100 Palermo, Italy 27 8 2016 27 8 2016 2016 14 1 24828 6 2016 16 8 2016 © The Author(s) 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Aims Myocardial fibrosis (MF) is a deleterious consequence of aortic valve stenosis (AVS). Global longitudinal strain (GLS) is a novel left ventricular (LV) functional parameter potentially useful to non-invasively estimate MF. MicroRNAs (miRNAs) are non-coding small ribonucleic acids (RNA) modulating genes function, mainly through RNA degradation. miRNA-21 is a biomarker associated with MF in pressure overload. The aim of the present study was to find an integrated algorithm for detection of MF using a combined approach with both bio- and functional markers. Methods Thirty-six patients (75.2 ± 8 y.o.; 63 % Female) with severe AVS and preserved LV ejection fraction (EF), candidate to surgical aortic valve replacement (sAVR) were enrolled. Clinical, bio-humoral evaluation (including plasmatic miRNA-21 collected using specific tubes, PAXgene, for stabilization of peripheral RNA) and a complete echocardiographic study, including GLS and septal strain, were performed before sAVR. Twenty-eight of those patients underwent sAVR and, in 23 of them, an inter-ventricular septum biopsy was performed. Tissues were fixed in formalin and embedded in paraffin. Sections were stained with Hematoxylin and Eosin for histological evaluation and with histochemical Masson trichrome for collagen fibers. The different components were calculated and expressed as micrometers2. To evaluate tissue miRNA components, sections 2-μm thick were cut using a microtome blade for each slide. Regression analysis was performed to test association between dependent variable and various predictors included in the model. Results Despite a preserved EF (66 ± 11 %), patients presented altered myocardial deformation parameters (GLS −14,02 ± 3.8 %; septal longitudinal strain, SSL −9.63 ± 2.9 %; septal longitudinal strain rate, SL-Sr −0.58 ± 0.17 1/s; Septal Longitudinal early-diastolic strain rate, SL-SrE 0.62 ± 0.32 1/s). The extent of MF showed an inverse association with both GLS and septal longitudinal deformation indices (GLS: R2 = 0.30; p = 0.02; SSL: R2 = 0.36; p = 0.01; SL-Sr: R2 = 0.39; p < 0.001; SL-SrE: R2 = 0.35; p = 0.001). miRNA-21 was mainly expressed in fibrous tissue (p < 0.0001). A significant association between MF and plasmatic miRNA-21, alone and weighted for measures of structural (LVMi R2 = 0.50; p = 0.0005) and functional (SSL R2 = 0.35; p = 0.006) remodeling, was found. Conclusions In AVS, MF is associated with alterations of regional and global strain. Plasmatic miRNA-21 is directly related to MF and associated with LV structural and functional impairment. Keywords Aortic valveAortic stenosisAortic valve replacementTissue characterizationTissue and strain Doppler echocardiographyMyocardial strainissue-copyright-statement© The Author(s) 2016 ==== Body Background Aortic valve stenosis (AVS) is the most common valvular heart disease in Western Countries [1, 2]. In particular, AVS is a slowly progressive disease, associated with significant Left Ventricular (LV) pressure overload, which induces left ventricular hypertrophy (LVH) and secondary myocardial fibrosis (MF). MF is an early morphologic alteration in patients with AVS and a major determinant of LV functional impairment, ultimately leading to the development of heart failure. Accordingly, robust and repeatable measures of MF that may be applied in the clinical field are eagerly awaited. Nevertheless, the evaluation of patients with AVS is generally limited to the assessment of flow-dependent parameters (velocity; gradients) that reflect only the “valvular side” of the pathology, disregarding LV components of disease [3–5]. Beside the endo-myocardial biopsy (gold standard), not ethically feasible in a clinical setting, a series of biomarkers and cardiac imaging techniques have been lately proposed to combine tissue parameters with functional evaluation. In this respect, the evaluation of LV deformation by Speckle Tracking Echocardiography (2D-STE) has been shown to allow a better assessment of cardiac contractile function than traditional parameters [i.e. Ejection Fraction (EF)], giving the chance to assess the presence of subtle alterations of LV systolic performance. In particular, regional and global longitudinal strain (GLS), showed a better and earlier diagnostic power over EF, becoming reference indicators for the precocious assessment of sub-clinical LV functional impairment [6–12]. More recently, measures obtained through 2D-STE showed to correlate with the presence and extent of MF at cardiac magnetic resonance, as assessed through T1-mapping and late gadolinium enhancement as quantification techniques [13, 14]. MicroRNAs (miRNAs) are non-coding small ribonucleic acids (RNAs) that modulate the expression of target genes inducing mRNA degradation. The expression of miRNAs is associated with multiple pathological processes that affect also the cardiovascular system [15, 16]. A regulatory role for miRNA-21 has been evidenced in LV myocardial remodeling induced by hemodynamic stress [17, 18]. Recent reports indicate that the presence of circulating miRNAs may reflect specific cardiovascular pathologies and could be a useful biomarker for different cardiovascular diseases. Accordingly, the evaluation of both miRNA (as a biomarker) and GLS (as a functional marker), might allow an integrated assessment of the pathophysiological relationship between MF and adverse LV remodeling. With these considerations in mind, we aimed at assessing:the presence of MF (detected by endo-myocardial biopsy) both with gold standard histologic method and with an advanced laser micro-dissection methodology (tissue miRNA-21), in patients with severe AVS; the presence of a direct association between plasmatic and tissue pool of miRNA-21; the relationship between 2D-STE parameters, MF (endo-myocardial biopsy) and plasmatic/tissue miRNA-21 expression levels, in order to develop a non-invasive detection of myocardial fibrotic burden. Methods Study population Thirty-six consecutive patients with severe symptomatic AVS (Peak Trans-valvular Velocity > 4 m/s; Mean Gradient > 40 mmHg; Aortic Valve Area-AVA < 1 cm2; AVAi < 0.6 cm2/m2) and preserved EF, were prospectively evaluated in University of Pisa Hospital (A.O.U.P) for sAVR. Patients underwent laboratory analysis (for 36 patients brain natriuretic peptide, BNP pg/mL; high-sensitive assays troponin T, hs-TnT, ng/L; plasmatic miRNA-21 assay for 30 patients), and trans-thoracic echocardiography (M-Mode, 2D, Doppler, Tissue Doppler Imaging-TDI, 2D-STE). Twenty-eight patients were finally submitted to sAVR (three patients refused surgery; five patients decision for percutaneous procedures), 23 of whom (in 5 patients under-sampling/biopsy not performed) underwent intra-operatory basal inter-ventricular septum biopsy to evaluate MF (23 patients) and tissutal levels of expression of miRNA-21 (20 patients). All patients signed an informed consent, approved by local ethical committee, conform to the ethical guidelines of the 1975 Declaration of Helsinki. We excluded patients with at least one of the following: age < 18 y.o., significant major comorbidities (i.e.cancer; dialysis; cachexia), inability to sign consent, pregnancy, poor acoustic window, ischemic heart disease (including epicardial coronary artery disease > 50 %), associated valvular disease of moderate-severe degree, non-degenerative AVS, diskynetic septum (i.e. stimulator; Left Bundle Branch Block). Conventional echocardiography Transthoracic exams were performed with a dedicated machine (Vivid-7, General Electric Milwaukee, WI-USA). Patients were imaged in the left lateral decubitus position and data were acquired with a 4 MHz (M4S) transducer at a depth of 16 cm in the parasternal (long- and short-axis views) and apical views (two- and four-chamber and apical long-axis views). All parameters were derived according to current indications, and considered in relation to their established reference ranges [19, 20]. LV dimensions were calculated from the standard M-mode/2D images at the parasternal long-axis views and included LV diameters and end-diastolic thickness of the interventricular septum and posterior wall. Left ventricular mass was calculated and corrected by the body surface area to derive mass index (LVMi). The LV end-diastolic and end-systolic volumes were measured from the apical two- and four-chamber views, and EF was calculated using the Simpson’s rule. LV diastolic function was evaluated using early (E wave) and late (A wave) trans-mitral velocities, the E/A ratio, and the E wave deceleration time obtained from the spectral pulsed-wave Doppler recordings. In addition, TDI was performed, adjusting gain and frame rate to get an appropriate tissue characterization. The aortic valve area (AVA; indexed, AVAi) was calculated by the continuity equation, and the maximum pressure gradient across the restrictive orifice was estimated by the modified Bernoulli equation. Mean trans-aortic pressure gradient was calculated averaging the instantaneous gradients over the ejection period on the continuous-wave Doppler recordings. As a measurement of global left ventricular afterload, the valvulo-arterial impedance (ZVA) was calculated. Finally, color Doppler echocardiography was performed after optimizing gain and Nyquist limit in order to evaluate the presence of regurgitant valve disease. The severity of valvular regurgitation was determined on a qualitative scale (mild, moderate, and severe), according to the current guidelines [19, 20]. Speckle tracking echocardiography Assessment of LV GLS was performed using 2D-STE (frame rate 45–90 frame/s, fps). We limited the analysis to the global longitudinal component of strain (peak value-mid myocardium). Quantifications were performed using the available software (EchoPAC 10, General Electric), as described previously [21–23]. For this purpose, standard 2D grey-scale images of the LV were acquired at conventional apical two- and four-chamber and apical long-axis views. 2D-STE enables angle-independent myocardial deformation analysis by tracking frame-to-frame natural acoustic markers, or speckles, that appear equally distributed within the myocardial wall. Applying the strain Lagrangian formula, the percentage change in myocardial length relative to the initial length derives myocardial strain (expressed in percentage). The temporal derivation of myocardial strain results in strain rate and is a measure of the rate of deformation. The longitudinal deformation relates to motion from mitral annulus to the apex in the apical views and results in shortening (negative strain) and lengthening (positive strain). Using the dedicated application, the endocardial contour was manually traced at an end-systolic frame. The software then automatically traced a concentric region of interest (ROI), including the entire myocardial wall. The myocardial tracking was verified, and the ROI width was adjusted to optimize the tracking, if needed. Next, segmental strain analysis was performed dividing each LV image into six segments. Peak systolic parameters were calculated averaging the peak systolic values of the eighteen segments, derived from the six segments of the three apical views (two- and four-chamber and apical long-axis views). For dedicated septal analysis, a focused ROI (80–95/fps) was traced specifically for inter-ventricular septum. We assessed, septal longitudinal systolic strain (SSL), systolic (SL-Sr) and early-diastolic (SL-SrE) strain rates. We then averaged measure from anterior and inferior septum. Intra and inter-observer variability analysis for 2D-STE was evaluated by intra/inter-class correlation coefficient (ICC). Ten randomly selected patients were evaluated three-times by the same operator (same beats; consecutive days). The same measurements were repeated in the same day by a second clinician, blinded to previous results. All ICC resulted >0.80 (p < 0.05), showing good agreement. Invasive measurements In 22 patients, in addition to coronary angiography, standard left heart catheterization was performed before sAVR. Peak-to-peak gradient, invasive end-diastolic pressure (EDPi; fluid-filled catheter) and semi-quantitative aortic regurgitation were evaluated. Operating myocardial biopsy In 23 patients undergoing sAVR, concomitant intra-operatory basal left-side inter-ventricular septum biopsy was performed to assess MF, as previously described [24–26]. Briefly, tissues (30–80 mg) were fixed in 10 % formalin and embedded in paraffin. One Section (2 µm) was stained with Hematoxylin and Eosin for histological evaluation and one Section (5 µm) with histochemical Masson trichrome stain for collagen fibers. The different components of the myocardial biopsy were calculated by computer analysis (PALM MicroBeam, Carl Zeiss), and expressed as micrometers squared. In particular, the following parameters were analyzed:the overall myocardial area occupied by the myocytes and connective tissue (fibrotic area); myocyte area connective area connective/overall myocardial area (ratio % as MF grading) All measurements were made by two expert pathologists without knowledge of the clinical data (intra-class and inter-observer correlation coefficients on 5 random samples of 0.9 and 0.94, respectively). Origin of myocardial miRNA-21 Immunohistochemistry 4-micron sections were dewaxed in xylene and rehydrated through graded alcohols to water. Antigen retrieval was performed microwaving sections for 9 min in citrate/EDTA buffer (pH 7.8). Not specific peroxidase activity was blocked with 3 % hydrogen peroxidase for 15 min, and non-specific binding prevented by incubation with normal goat serum for 10 min. Afterwards, incubation with anti-vimentin mouse monoclonal antibody (clone VI10, Abcam, diluition 1:200) and anti-CD45/LCA rabbit monoclonal antibody (clone EP68, Cell Marque, diluition 1:100) was performed for 1 h at room temperature. A biotin conjugated goat derived secondary antibody was applied followed by the Vectastain Elite ABC kit (Vector Laboratories). Slide were incubated with diaminobenzidine tetrahydrochloride (DAB) and counterstained with hematoxylin. miRNA-21: plasmatic and tissue study Blood samples Peripheral blood samples were collected using specific tubes (PAXgene Blood RNA Tube) included in the commercial systems for collection and immediate stabilization of peripheral blood RNA (PreAnalytiX). Tissue samples For 20 biopsy specimens laser microdissection (LSMD) was performed. Sections 2-μm thick were cut from each case using a new microtome blade for each slide. The PALM MicroBeam laser micro-dissector from Carl Zeiss was used to select and collect cardiomyocytes and fibrotic cells to be studied separately. RNA isolation Blood RNA was purified using the commercial kit PAXgene Blood RNA Kit (PreAnalytiX). The quantity of extracted RNA was estimated with Qubit 2.0 Fluorometer (Life Technologies) by using 2ul of undiluted RNA solution. Yield ranged from 50 to 500 ng/ul of RNA. Microdissected samples were incubated at 55 °C overnight upside-down with 50 μl of lysis buffer and 10 μl of proteinase K. The day after the samples were loaded in Maxwell 16 Instrument (Promega) to extract RNA. Reverse transcription and analysis of miRNA profiles MicroRNAs were reverse transcribed from 6 µl of total extracted RNA sample using the miScript II RT Kit (QIAGEN). cDNA from micro dissected samples was pre-amplified prior to real-time PCR analysis of miRNA. miRNA expression analysis were performed in triplicate using 1 µl of diluted cDNA as a template for real-time PCR with the miScript SYBR Green PCR Kit (QIAGEN) and the miScript Primer Assays (SNORD61—assay code MS00033705, miRNA-21—assay code MS00009079 (QIAGEN)) according to manufacturer’s instructions on the CFX96 Real Time system c1000 thermal cycler (BIORAD). Data analysis Data analysis was performed using the Bio-Rad CFX Manager Software v3.1 and Microsoft Excel. miRNA 21 expression was calculated using SNORD61 expression level as reference and the relative normalized expression ∆∆Cq formula. Statistical analysis The data sets were assessed for normality with the Kolmogorov–Smirnov test. Continuous variables are described as mean ± standard deviation (SD). Otherwise as median (with minimum/maximum). Categorical data are reported as percentage. Plasmatic and tissue miRNA-21 levels of expression where measurable were treated as continuous variables. Continuous variables were compared using the Mann–Whitney U test when non-Gaussian. Univariate linear regression analysis was performed to test association between dependent variable and various potential predictors included in the model. We assessed also the association between Myocardial Fibrosis (Y) and Plasmatic miRNA-21 (X) weighted for left ventricular mass indexed for BSA and septal longitudinal strain, respectively. The threshold for statistical significance was p < 0.05. In order to preserve the statistical meaning of regression analysis (direct/inverse correlation/association), in the text we considered global longitudinal strain/systolic strain rate in absolute value. Using a c-statistic approach we derived the miRNA-21 value with the best combination of sensitivity and specificity for discrimination of patients with a significant amount of MF (more than 10 % of the specimen). The following statistic package was used: Medcalc 12.7 (Medcalc Software 2013, Belgium). Results The characteristics of the population regarding clinical, laboratory and echocardiographic parameters are shown in Tables 1, 2 and 3.Table 1 Population characteristics Mean SD General characteristics Age (year) 75.2 8.06 BSA (m2) 1.8 0.17 Log EUROSCORE (%) 5.9 4.17 EUROSCORE II (%) 2.2 1.13 SAP (mmHg) 139.1 19.00 DAP (mmHg) 71.3 10.03 HR (bpm) 73.5 11.92 n % Clinical characteristics Female sex 23 63 CHD (</=50 % epicardial coronary) 12 33 COPD 8 22 Anemia 14 38 Chronic kidney dis. 22 61 Diabetes Mell. 8 22 Arterial hypertension 31 86 Dyslipidemia 21 58 n % Drugs (admission) ACE-inhibitors 15 41 AT-II-inhibitors 10 27 Anti-aldosteronic 2 5 Diuretics 14 38 Calcium-antagonist 6 10 Laboratory data Mean/Median* SD/Min–Max** BNP (pg/mL) 250.9 220.4 GFR (mL/min/1.73 m2) 70.8 28.4 hsTnT (ng/L) 30.4 26.8 miRNA-21 (30 pts) 2.02* 0.02–11.26** ACE angiotensin converting enzyme, AT-II angiotensin 2 receptor, BNP brain natriuretic peptide, BSA body surface area, CHD coronary heart disease, COPD chronic obstructive pulmonary disease, DAP diastolic arterial pressure, GFR glomerular filtration rate, HR heart rate, SAP systolic arterial pressure Table 2 Echocardiographic and invasive data Mean SD Valvular parameters AVAi (cm2/m2) 0.45 0.09 Max gradient (mmHg) 80.2 16.76 Mean gradient (mmHg) 49.7 7.67 Peak-peak gradient (mmHg) 58.3 15.40 Velocity-ratio 0.18 0.04 Peak velocity (m/sec) 4.4 0.34 Non invasive haemodynamic data SVi (mL/m2) 38.5 16.08 CI (L/min/m2) 2.5 0.72 CO (L/min) 4.8 1.45 ZVA (mmHg/ml/m2) 5.9 1.26 Diastolic function parameters EDPi (mmHg) (invasive) 16.81 6.81 LAVi (mL/m2) 48.2 12.65 E/A 0.8 0.33 E/e’ Average 18.4 8.39 DT (msec) 247.2 93.25 Conventional systolic function parameters (with TDI) EF % 65.8 10.94 FS % 36.4 7.89 MAPSE (mm) 9.5 1.84 s’ l (cm/s) 6.4 1.50 s’s (cm/s) 5.6 1.49 Left and right ventricular echo parameters EDDi (cm/m2) 2.52 0.25 EDVi (mL/m2) 50.28 12.95 ESDi (cm) 1.71 0.30 ESVi (mL/m2) 17.47 8.55 LVMi (g/m2) 149.5 20.7 RWT 0.51 0.07 sPAP (mmHg) 30.8 6.46 TAPSE (cm) 1.8 0.25 AVAi indexed aortic valve area, CI cardiac index, CO cardiac output, DT deceleration time, E/A ratio of early to late diastolic mitral filling velocity, E/e′ ratio of early diastolic velocity (PW) to tissue proto-diastolic velocity (TDI), EDDi indexed left ventricular end-diastolic diameter, EDPi invasive left ventricular end-diastolic pressure, EDVi, indexed left ventricular end-diastolic volume, EF, ejection fraction, ESDi, indexed left ventricular end-systolic diameter, ESVi, indexed left ventricular end-systolic volume, FS, fractional shortening, LAVi, indexed left atrial volume, LVMi, indexed left ventricular mass, MAPSE, mitral annular plane systolic excursion, RWT, relative wall thickness, s’L, systolic velocity (TDI) lateral, s’S, systolic velocity (TDI) septal, sPAP, systolic pulmonary arterial pressure, SVi, indexed stroke volume, TAPSE, tricuspid annular plane systolic excursion, Z VA, valvulo-arterial impedance Table 3 Speckle tracking and tissue data Mean SD Speckle tracking GLS % −14.02 3.88 Mean/median* SD/min–max** Septal speckle tracking and tissue data (septum) SL-Sr (1/sec) −0.58 0.17 SL-SrE (1/sec) 0.62 0.32 SSL (%) −9.63 2.97 MF % (n.23) 18.45* 5.13–98.0** miRNA-21 myocardial expression/myocardial area (n.20) 0.416* 0.05–1.53** miRNA-21 fibrotic/fibrotic area (n.20) 4.041* 0.57–22.27** GLS global longitudinal strain, MF myocardial fibrosis, miRNA micro-RNA, SL-Sr septal systolic strain rate, SL-SrE septal early-diastolic strain rate, SSL septal longitudinal strain All patients (Tables 2 and 3) showed a significantly elevated left ventricular mass indexed (LVMi) for body surface area (BSA), with evidence of concentric LV hypertrophy. A variable degree of diastolic impairment was observed, with increased EDPi and left atrial dimensions. LVMi didn’t show a significant association with indices of AVS severity (AVAi; Max/mean gradients; peak-to-peak gradient; Velocity Ratio; ZVA). Even if conventional indices of global systolic function were preserved (EF, Fractional shortening), more sensitive parameters of longitudinal function (e.g. Mitral annular plane systolic excursion, MAPSE; TDI; Table 2) were reduced when compared to normal ranges. Speckle tracking analysis A significant impairment of global longitudinal deformation parameters (GLS) was observed (Table 3). In particular, GLS showed a direct relationship with indexed LV stroke volume (SVi) (R2 = 0.20; p = 0.006) and a significant inverse relationship with BNP levels (R2 = 0.47; p = 0.007). Modest inverse relationships between LVMi and AVAi (R2 = 0.16; p = 0.01), ZVA (R2 = 0.12; p = 0.03) and GLS (R2 = 0.23; p = 0.002), were also observed. Septal sub-analysis showed higher impairment of deformation indices and a significant direct relationship of SSL with stroke volume (R2 = 0.22; p = 0.003). Tissue analysis: myocardial fibrosis A variable amount of MF (with absence of inflammatory cells) was a common finding in patients who performed biopsy. To distinguish fibroblasts from inflammatory cells, immunohistochemistry was performed on myocardial fibrosis for vimentin and CD45. Figure 1 clearly showed that myocardial fibrosis was composed not only by collagen fibres, highlighted by Masson’s Trichrome, but also by fibroblasts (vimentin positive); on the contrary only few inflammatory cells (CD45 positive) were present.Fig. 1 Characterization of Fibrotic Areas (Immunohistochemistry). a Myocardial fibrosis is composed by collagen fibers with inter-dispersed vimentin positive fibroblasts (arrows). b Only rare CD45 positive inflammatory cells are encountered (arrow) (original magnification, ×10) MF % was associated (direct relationship) with EDPi (R2 = 0.31; p = 0.03) and showed an inverse relationship with SVi (R2 = 0.23; p = 0.02). Moreover, MF showed a significant inverse relationship with deformation indices (GLS: R2 = 0.30 and p = 0.02; SSL: R2 = 0.36 and p = 0.01; SL-Sr: R2 = 0.39 and p < 0.001; SL-SrE: R2 = 0.35 and p = 0.001) (in Fig. 2 two samples)Fig. 2 Tissue samples. Samples from intra-operatory biopsies (a low and b high myocardial fibrosis at basal interventricular septum level), showing myocardial fibrosis (Hematoxylin Eosin/Masson’s Trichrome). Region of interest (ROI) traced to derive longitudinal septal strain values (SSL %; c −11 %/d −8 %) are shown We did not find any relevant association between MF and other measures of LV systolic function (i.e. FE, MAPSE or TDI systolic velocities). Moreover, MF showed no association with afterload parameters, including AVAi, ZVA and gradients. No relationships were found between MF and the major clinical and demographical parameters, such as age, duration of the disease, gender, history of diabetes mellitus, dyslipidemia, arterial hypertension and obesity. Tissue and plasmatic miRNA-21 analysis While miRNA-21 was expressed both in myocytes and interstitial tissue, it resulted significantly more expressed in fibrous tissue (p < 0.0001; Fig. 3). Tissue miRNA-21 levels did not show an association with LVMi, body mass index (BMI), BSA or age. Conversely, interstitial miRNA-21 was inversely related to septal and global longitudinal deformation (SSL: R2 = 0.32 and p = 0.01; GLS: R2 = 0.34 and p = 0.008). Plasmatic miRNA-21 concentrations (n = 30) demonstrated a significant direct relationship with whole MF (R2 = 0.31; p = 0.001) and interstitial miRNA-21 compartment (R2 = 0.36; p = 0.001).Fig. 3 miRNA expression in tissue samples. Differential levels of expression of miRNA-21 in myocardial and interstitial tissue. The levels of expression (a) of miRNA-21 in the interstitial compart normalized for the area of fibrosis (21F) resulted higher (p < 0.0001) than in myocardial compart, normalized for the myocardial area of the specimen (21M). In (b) a picture from a specimen No relationships were found between plasmatic or tissue miRNA-21 and the major clinical and demographical parameters. Integrated speckle tracking and plasmatic miRNA-21 analysis A significant and strong positive relationship between MF and plasmatic miRNA-21 was found, also after weighting for cardiac remodeling (assessed as LVMi: R2 = 0.50; p = 0.0005) and LV function parameters (SSL R2 = 0.35; p = 0.006; Fig. 4).Fig. 4 Univariate regression (including 95 % confidence, prediction and line of equality) weighted for Left Ventricular Massi. Myocardial fibrosis in percentage (MF %) and plasmatic levels of micro-RNA-21 (miRNA-21): R2 = 0.50; p = 0.0005 Neither GLS nor SSL showed a significant diagnostic accuracy in MF evaluation using c-statistic approach. Conversely, patients with higher MF, showed a significant higher mean level of expression of plasmatic miRNA-21. (Figure 5; with table showing differences in clinical profile).Fig. 5 Differential levels of expression of plasmatic miRNA-21 in patients with significant MF (MF % > 10 %). The plasmatic levels of miRNA-21 in patients with high MF (Over Ten = 1) resulted higher respective to the low fibrosis group (Median 5.5043 vs. 0.8854; p = 0.03). The table shows principal differences in clinical profiles. AVAi indexed aortic valve area, BNP brain natriuretic peptide, DT deceleration time, E/A ratio of early to late diastolic mitral filling velocity (PW), EF ejection fraction, eGFR estimated glomerular filtration rate, GLS global longitudinal strain, LVMi indexed left ventricular mass, MF Myocardial Fibrosis, RWT relative wall thickness At ROC analysis, a plasmatic miRNA-21 value >2.4552 showed the best accuracy (Sensitivity 64.29 %; Specificity 100 %; AUC 0.81; p = 0.001) for discriminating patients with significant MF (described as equal or more than 10 % of the specimen). No gender-based differences were found in the study. Discussion The main findings of the present paper are:Patients with severe AVS show abnormalities of regional and global left ventricular myocardial strain, reflecting both pressure overload and geometric remodeling. These deformation abnormalities are related with the level of invasively measured MF (gold standard); The expression of textural miRNA-21 determined with laser micro-dissection may document its pathophysiological role in AVS. In particular, we focused on interdependence between textural miRNA-21 and fibrogenic stimulus induced by an abnormal left ventricular pressure overload; Circulating miRNA-21 (biomarker) levels are high in patients with severe AVS, reflecting the presence of significant myocardial fibrosis (defined as MF % higher or lower than 10 %). Deformation imaging and myocardial fibrosis in aortic stenosis Conditions of LV pressure overload determine a deep remodeling of the extracellular matrix, with the secondary deposition of MF. In the clinical setting, MF is known to be a deleterious consequence of AVS, contributing to systo-diastolic alterations and arrythmogenicity, affecting patients’ prognosis and quality of life after AVR. [27] Meanwhile, we should not forget that in AVS, the paradigm “Pressure overload—LV remodeling—Myocardial hypertrophy—interstitial and later replacement fibrosis” remains still not so definite: indeed, there is a wide variation, independent from the stage of the disease (especially if we consider only Valve Area). Thus, some patients with severe AVS have normal ventricular structure and no/mild fibrosis (10–30 %) while patients with only moderate AVS may have extensive hypertrophy and large amount of fibrosis [13, 25–27]. In our study population, despite a preserved EF, we found a significant amount of MF at endo-myocardial biopsy, confirming the insensitivity of EF in revealing subtle myocardial textural alterations. On the contrary, as previously reported, myocardial deformation parameters, assessed by 2D-STE, were altered (in a context of preserved EF) and inversely related to global afterload and remodeling parameters (LVMi) [12, 28, 29]. Similarly, we found a significant association between GLS and stroke volume, an important index for AVS re-classification and management [4]. Moreover, myocardial deformation indices showed a significant inverse association with tissue MF, offering the potential appeal of a non-invasive, cost-effective (respective to MRI) tool for the detection of MF and for better AVS risk stratification. In attempting to estimate MF, many non-invasive imaging modalities showed good correlation with tissue data [30, 31]. To our knowledge, while previous reports have shown a correlation between longitudinal echo parameters (MAPSE; strain TDI) [25, 26] or reflectivity indices (IBS) [32] and MF in AVS patients, the possible relationship between the presence and extent of MF and novel, more sensitive, echocardiographic parameters (i.e. 2D-STE) has never been addressed before. In our study, we did not find a significant association between MAPSE or TDI (systolic velocities) and textural parameters. These results are in line with recent studies conducted in patients with Hypertrophic Cardiomyopathy that underwent septal miectomy. In fact, deformation parameters showed a strict correlation with myocardial fibrosis, while there was no association between MF and conventional echo parameters, including TDI [33, 34]. In addition, recent T1-mapping MRI techniques, found a significant correlation between the signal and MF and between signal intensity and deformation indices (GLS). [35, 36] microRNA and aortic valve stenosis Recent evidences showed a key role for miRNAs, including miRNA-21, in cardio-vascular pathophysiology. However, only few recent papers evaluated their involvement in human heart, considering plasmatic and (seldom) tissue pools [15]. Several previous findings have underlined the fibrogenic potential of miRNA-21 in hearts with superimposed pressure overload, mediating mRNA for fibrillar proteins. miRNA-21 already showed a pathophysiological role in AVS, with plasmatic levels resulting higher respective to controls and correlating with mean valvular gradient. [37] To our knowledge, this is the first paper evaluating tissue miRNA expression in biopsies of patients (in vivo) with AVS with LSMD, a more precise method respective to in situ hybridation (FISH). In line with previous findings, tissue expression of miRNA was higher for interstitial compartment than myocardial tissue [37], with no relationship with myocardial mass. The reason is probably because LVM reflects both myocardial and fibrous tissue compartment, while miRNA-21 is mainly limited to fibrous compartment, potentially (as we may argue from immunohistochemical findings) derived from fibroblasts. Interestingly, miRNA-21 levels of expression in fibrous tissue, in line with their absolute over-expression at that level, showed a significant inverse association with deformation indices. Most important, for plasmatic levels of miRNA-21, which already resulted elevated in previous cohorts of patients with AVS respective to controls [37–39], we found a direct association with MF and interstitial miRNA-21 levels. This finding was stronger if weighted for LVMi (a gross indicator of “whole remodeling”) value. Thus, after validation in larger cohorts, plasmatic levels of miRNA-21 could be used as a reliable biomarker of myocardial fibrosis [40]. Recently, miRNA plasmatic levels confirmed their strong fibrogenic implications in other similar contexts [41]. This may also open, in perspective, to myocardial fibrosis inhibition targeting, as very recently shown by Gupta et al. in animal models of acute allograft cardiac transplantation. [42] Putting together the puzzle: dual-step functional and textural analysis We propose a complementary role of echocardiographic speckle tracking and plasmatic miRNA-21 analysis: the identification of the remodeling process (at macroscopic and tissue level) combined with a refined functional approach. In particular, the evaluation of plasmatic miRNA together with GLS (by summing functional, structural and textural parameters) could help in better stratifying those patients that currently fall in a diagnostic “gray zone” of AVS severity. Then, we can speculate a potential clinical implication in terms of clinical practice/safety (i.e. myocardial biopsy) and cost reduction (e.g. if we consider during the follow up other expensive imaging procedures, such as magnetic resonance imaging). Present results underscore the tight relationship between valve and myocardium (in our opinion the “main actor” of this complex pathophysiological process), suggesting that only a combined evaluation of both variables may allow a complete evaluation of patients with AVS [7]. Limitations The present work was designed as a pilot study. The main limitation of the study is its small sample size and, at present, the absence of a follow-up. Analysis of plasmatic miRNA-21 is promising, but must be validated in larger studies, as its prognostic role and remodeling implications. To define a significant amount of MF evaluated with myocardial biopsy, we arbitrarily decided the cut-off of 10 %, according to the literature [13, 24]. So far, above all due to the limited cases collected, we didn’t have the objective to derive cut-off for plasmatic/tissue values of miRNA-21. Finally, this study was not designed to identify risk-factors associated with MF in patients with AVS. Therefore, it is possible that the small number of patients fails in showing significant differences between subjects with similar age and cardio-vascular risk profiles (diabetes, hypertension etc.). Anyway, consistent with previous and larger observations [25–27], the degree of hypertrophy/MF is not strictly associated with cardio-vascular risk profile. Genetic factors and gender are likely to play an important role in modulating myocardial response. This may explain the large inter-individual variability in remodeling and fibrosis observed in the setting of AVS. Conclusions In patients with severe AVS, myocardial fibrosis was associated with significant alterations of both plasmatic and textural miRNA-21 (biomarker) levels, as well as with impairment of regional and global longitudinal strain (functional marker). This combined bio-humoral and functional evaluation could allow a better definition of the remodeling process that takes place in AVS, possibly further improving risk stratification of patients. Prospective studies in larger populations of patients with AVS, are needed to better analyze the effective prognostic value of this imaging and bio-humoral integrated approach, in order to shift the clinical focus also on myocardium, beside valvular apparatus. Abbreviations 2D-STE2D-speckle tracking echocardiography AVAiindexed aortic valve area AVSaortic valve stenosis BMIbody mass index BNPbrain natriuretic peptide BSAbody surface area cDNAcopied DNA CIcardiac index DTdeceleration time E/Aratio of early to late diastolic mitral filling velocity (PW) EDPiinvasive left ventricular end-diastolic pressure EFejection fraction fpsframe per second FISHfluorescence in situ hybridation GLSglobal longitudinal strain hs-TnThigh sensitivity assay troponin-T IBSintegrated back-scatter ICCintra/inter-class correlation coefficient LAViindexed left atrial volume LSMDlaser micro-dissection LVleft ventricular LVHleft ventricular hypertrophy LVMiindexed left ventricular mass MAPSEmitral annular plane systolic excursion MFmyocardial fibrosis miRNAmicroRNA (ribonucleic acids, RNA) MRImagnetic resonance imaging NYHANew York Heart association PCRpolymerase chain reaction ROIregion of interest RWTrelative wall thickness s’Lsystolic velocity (TDI) lateral s’Ssystolic velocity (TDI) septal sAVRsurgical aortic valve replacement SNORDsmall nucleolar RNA SSLseptal longitudinal strain SL-Srseptal longitudinal strain rate SL-SrEseptal early-diastolic longitudinal strain rate SViindexed stroke volume TDItissue Doppler imaging ZVAvalvulo-arterial impedance Authors’ contributions IF1,2,3, CS1,2,3, CMM1,2,3, LC1,2,3, NRP1,2,3, SF3, FL3, MM2,3, ADM3, SP3, UB2,3, AGN1,2,3, SLC1 and VDB1,2,3. Each author has contributed significantly to the submitted work, as indicated with the numbers below: (1) conception, design and interpretation of data; (2) drafting of the manuscript and revising it critically for important intellectual content; (3) final approval of the manuscript submitted. All authors read and approved the final manuscript Acknowledgements Not applicable Competing interests The authors declare that they have no competing interests. Availability of data and material Fully available under request (excel file) Ethics approval and consent to participate Study approved by the local ethical commitee (A.O.U.P); each patient signed a written informed consent. ==== Refs References 1. Carabello BA Paulus WJ Aortic stenosis Lancet 2009 373 956 966 10.1016/S0140-6736(09)60211-7 19232707 2. Vahanian A Alfieri O Andreotti F Antunes MJ Barón-Esquivias G Baumgartner H Guidelines on the management of valvular heart disease (version 2012) Eur Heart J 2012 33 19 2451 2496 10.1093/eurheartj/ehs109 22922415 3. Pibarot P Dumesnil JG Assessment of aortic stenosis severity: check the valve but don’t forget the arteries! Heart 2007 93 780 782 10.1136/hrt.2006.111914 17569800 4. Lancellotti P Magne J Donal E Davin L O’Connor K Rosca M Clinical outcome in asymptomatic severe aortic stenosis: insights from the new proposed aortic stenosis grading classification J Am Coll Cardiol 2012 59 235 243 10.1016/j.jacc.2011.08.072 22240128 5. Eleid MF Sorajja P Michelena HI Malouf JF Scott CG Pellikka PA Survival by stroke volume index in patients with low-gradient normal EF severe aortic stenosis Heart 2015 101 23 29 10.1136/heartjnl-2014-306151 25217490 6. Dumesnil JG Shoucri RM Laurenceau JL Turcot J A mathematical model of the dynamic geometry of the intact left ventricle and its application to clinical data Circulation 1979 59 1024 1034 10.1161/01.CIR.59.5.1024 428084 7. Pibarot P Dumesnil JG Aortic stenosis: look globally, think globally JACC Cardiovasc Imaging 2009 2 400 403 10.1016/j.jcmg.2009.01.004 19580720 8. Mor-Avi V Lang RM Badano LP Belohlavek M Cardim NM Derumeaux G Current and evolving echocardiographic techniques for the quantitative evaluation of cardiac mechanics Eur J Echocardiogr 2011 12 3 167 205 10.1093/ejechocard/jer021 21385887 9. Voigt JU Pedrizzetti G Lysyansky P Marwick TH Houle H Baumann R Definitions for a common standard for 2D speckle tracking echocardiography: consensus document of the EACVI/ASE/industry task force to standardize deformation imaging Eur Heart J Cardiovasc Imaging 2015 16 1 11 10.1093/ehjci/jeu184 25525063 10. Lancellotti P Moonen M Magne J O’Connor K Cosyns B Attena E Prognostic effect of long-axis left ventricular dysfunction and B-type natriuretic peptide levels in asymptomatic aortic stenosis Am J Cardiol 2010 105 383 388 10.1016/j.amjcard.2009.09.043 20102953 11. Lafitte S Perlant M Reant P Serri K Douard H DeMaria A Impact of impaired myocardial deformations on exercise tolerance and prognosis in patients with asymptomatic aortic stenosis Eur J Echocardiogr 2009 10 414 419 10.1093/ejechocard/jen299 18996958 12. Delgado V Tops LF van Bommel RJ van der Kley F Marsan NA Klautz RJ Strain analysis in patients with severe aortic stenosis and preserved left ventricular ejection fraction undergoing surgical valve replacement Eur Heart J 2009 30 3037 3047 10.1093/eurheartj/ehp351 19726436 13. Dweck MR Joshi S Murigu T Alpendurada F Jabbour A Melina G Midwall fibrosis is an independent predictor of mortality in patients with aortic stenosis J Am Coll Cardiol 2011 58 1271 1279 10.1016/j.jacc.2011.03.064 21903062 14. Chin CW Semple S Malley T White AC Mirsadraee S Weale PJ Optimization and comparison of myocardial T1 techniques at 3T in patients with aortic stenosis Eur Heart J Cardiovasc Imaging 2014 15 556 565 10.1093/ehjci/jet245 24282220 15. Small EM Frost RJ Olson EN MicroRNAs add a new dimension to cardiovascular disease Circulation 2010 121 1022 1032 10.1161/CIRCULATIONAHA.109.889048 20194875 16. Thum T Condorelli G Long noncoding RNAs and MicroRNAs in cardiovascular pathophysiology Circ Res 2015 116 751 762 10.1161/CIRCRESAHA.116.303549 25677521 17. Thum T Gross C Fiedler J Fischer T Kissler S Bussen M MicroRNA-21 contributes to myocardial disease by stimulating MAP kinase signalling in fibroblasts Nature 2008 456 980 984 10.1038/nature07511 19043405 18. Roncarati R Viviani Anselmi C Losi MA Papa L Cavarretta E Da Costa Martins P Circulating miR-29a, among other up-regulated microRNAs, is the only biomarker for both hypertrophy and fibrosis in patients with hypertrophic cardiomyopathy J Am Coll Cardiol 2014 63 920 927 10.1016/j.jacc.2013.09.041 24161319 19. Lang RM Badano LP Mor-Avi V Afilalo J Armstrong A Ernande L Recommendations for cardiac chamber quantification by echocardiography in adults Eur Heart J Cardiovasc Imaging 2015 16 233 271 10.1093/ehjci/jev014 25712077 20. Baumgartner H Hung J Bermejo J Chambers JB Evangelista A Griffin BP Echocardiographic assessment of valve stenosis: EAE/ASE recommendations for clinical practice Eur J Echocardiogr 2009 10 1 25 10.1093/ejechocard/jen303 19065003 21. Perk G Tunick PA Kronzon I Non-doppler two-dimensional strain imaging by echocardiography–from technical considerations to clinical applications J Am Soc Echocardiogr 2007 20 234 243 10.1016/j.echo.2006.08.023 17336748 22. Marwick TH Leano RL Brown J Sun JP Hoffmann R Lysyansky P Myocardial strain measurement with 2-dimensional speckle-tracking echocardiography: definition of normal range J Am Coll Cardiol Imaging 2009 2 80 84 10.1016/j.jcmg.2007.12.007 23. Yingchoncharoen T Agarwal S Popovic ZB Marwick TH Normal ranges of left ventricular strain: a meta-analysis J Am Soc Echocardiogr 2013 26 185 191 10.1016/j.echo.2012.10.008 23218891 24. Tanaka M Fujiwara H Onodera T Wu DJ Hamashima Y Kawai C Quantitative analysis of myocardial fibrosis in normals, hypertensive hearts, and hypertrophic cardiomyopathy Br Heart J 1986 55 575 581 10.1136/hrt.55.6.575 3718796 25. Weidemann F Herrmann S Stork S Niemann M Frantz S Lange V Impact of myocardial fibrosis in patients with symptomatic severe aortic stenosis Circulation 2009 120 577 584 10.1161/CIRCULATIONAHA.108.847772 19652094 26. Herrmann S Stork S Niemann M Lange V Strotmann JM Frantz S Low-gradient aortic valve stenosis myocardial fibrosis and its influence on function and outcome J Am Coll Cardiol 2011 58 402 412 10.1016/j.jacc.2011.02.059 21757118 27. Hein S Arnon E Kostin S Schonburg M Elsasser A Polyakova V Progression from compensated hypertrophy to failure in the pressure-overloaded human heart: structural deterioration and compensatory mechanisms Circulation 2003 107 984 991 10.1161/01.CIR.0000051865.66123.B7 12600911 28. Ng AC Delgado V Bertini M Antoni ML van Bommel RJ van Rijnsoever EP Alterations in multidirectional myocardial functions in patients with aortic stenosis and preserved ejection fraction: a two-dimensional speckle tracking analysis Eur Heart J 2011 32 1542 1550 10.1093/eurheartj/ehr084 21447510 29. Attias D Macron L Dreyfus J Monin JL Brochet E Lepage L Relationship between longitudinal strain and symptomatic status in aortic stenosis J Am Soc Echocardiogr 2013 26 868 874 10.1016/j.echo.2013.05.004 23768690 30. Jellis C Martin J Narula J Marwick TH Assessment of nonischemic myocardial fibrosis J Am Coll Cardiol 2010 56 89 97 10.1016/j.jacc.2010.02.047 20620723 31. Di Bello V Cucco C Giannini C Delle Donne MG Myocardial tissue characterization and aortic stenosis J Am Soc Echocardiogr 2010 23 1067 1070 10.1016/j.echo.2010.08.025 20868956 32. Di Bello V Giorgi D Viacava P Enrica T Nardi C Palagi C Severe aortic stenosis and myocardial function: diagnostic and prognostic usefulness of ultrasonic integrated backscatter analysis Circulation 2004 110 849 855 10.1161/01.CIR.0000138930.12773.41 15302789 33. Almaas VM Haugaa KH Strom EH Scott H Smith HJ Dahl CP Noninvasive assessment of myocardial fibrosis in patients with obstructive hypertrophic cardiomyopathy Heart 2014 100 631 638 10.1136/heartjnl-2013-304923 24368281 34. Kobayashi T Popovic Z Bhonsale A Smedira NG Tan C Rodriguez ER Association between septal strain rate and histopathology in symptomatic hypertrophic cardiomyopathy patients undergoing septal myectomy Am Heart J 2013 166 503 511 10.1016/j.ahj.2013.06.011 24016500 35. Lee SP Lee W Lee JM Park EA Kim HK Kim YJ Assessment of diffuse myocardial fibrosis by using MR imaging in asymptomatic patients with aortic stenosis Radiology 2015 274 359 369 10.1148/radiol.14141120 25251584 36. Hoffmann R Altiok E Friedman Z Becker M Frick M Myocardial deformation imaging by two-dimensional speckle-tracking echocardiography in comparison to late gadolinium enhancement cardiac magnetic resonance for analysis of myocardial fibrosis in severe aortic stenosis Am J Cardiol 2014 114 1083 1088 10.1016/j.amjcard.2014.07.018 25212549 37. Villar AV Garcia R Merino D Llano M Cobo M Montalvo C Myocardial and circulating levels of microRNA-21 reflect left ventricular fibrosis in aortic stenosis patients Int J Cardiol 2013 167 2875 2881 10.1016/j.ijcard.2012.07.021 22882958 38. Derda AA Thum S Lorenzen JM Bavendiek U Heineke J Keyser B Blood-based microRNA signatures differentiate various forms of cardiac hypertrophy Int J Cardiol 2015 1 196 115 122 10.1016/j.ijcard.2015.05.185 26086795 39. Coffey S Williams MJ Phillips LV Jones GT Circulating microRNA profiling. Needs further refinement before clinical use in patients with aortic stenosis J Am Heart Assoc 2015 4 8 002150 10.1161/JAHA.115.002150 40. López B González A Ravassa S Beaumont J Moreno MU San José G Circulating biomarkers of myocardial fibrosis: the need for a reappraisal J Am Coll Cardiol 2015 65 22 2449 2456 10.1016/j.jacc.2015.04.026 26046739 41. Fang L Ellims AH Moore XL White DA Taylor AJ Chin-Dusting J Circulating microRNAs as biomarkers for diffuse myocardial fibrosis in patients with hypertrophic cardiomyopathy J Transl Med 2015 24 13 314 10.1186/s12967-015-0672-0 26404540 42. Gupta SK Itagaki R Zheng X Batkai S Thum S Ahmad F miR-21 promotes fibrosis in an acute cardiac allograft transplantation model Cardiovasc Res 2016 110 2 215 226 10.1093/cvr/cvw030 26865549

PMC005xxxxxx/PMC5002331.txt

==== Front BMC NeurolBMC NeurolBMC Neurology1471-2377BioMed Central London 67210.1186/s12883-016-0672-6Research ArticleThe value of magnetic resonance imaging as a biomarker for amyotrophic lateral sclerosis: a systematic review http://orcid.org/0000-0002-1291-8703Grolez G. guillaume.grolez2@chru-lille.fr 12Moreau C. caroline.moreau@chru-lille.fr 12Danel-Brunaud V. veronique.danel@chru-lille.fr 12Delmaire C. christine.chd@gmail.com 23Lopes R. renaud.lopes@gmail.com 23Pradat P. F. pierre-francois.pradat@psl.aphp.fr 45El Mendili M. M. mounir.elmendili@gmail.com 4Defebvre L. luc.defebvre@chru-lille.fr 12Devos D. david.devos@chru-lille.fr 1261 Department of Movement Disorders and Neurology, Lille University Hospital, Faculty of Medicine, University of Lille, Lille, France 2 INSERM U1171, Lille University Hospital, Faculty of Medicine, University of Lille, Lille, France 3 Department of Neuroradiology, Lille University Hospital, Faculty of Medicine, University of Lille, Lille, France 4 Laboratoire d’Imagerie Biomédicale, Sorbonne Universités, UPMC Univ Paris 06, CNRS, INSERM, Paris, France 5 Département des Maladies du Système Nerveux, Groupe Hospitalier Pitié-Salpêtrière, APHP, Paris, France 6 Department of Medical Pharmacology, Lille University Hospital, Faculty of Medicine, University of Lille, Lille, France 27 8 2016 27 8 2016 2016 16 1 1554 3 2016 10 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Amyotrophic lateral sclerosis (ALS) is a fatal, rapidly progressive neurodegenerative disease that mainly affects the motor system. A number of potentially neuroprotective and neurorestorative disease-modifying drugs are currently in clinical development. At present, the evaluation of a drug’s clinical efficacy in ALS is based on the ALS Functional Rating Scale Revised, motor tests and survival. However, these endpoints are general, variable and late-stage measures of the ALS disease process and thus require the long-term assessment of large cohorts. Hence, there is a need for more sensitive radiological biomarkers. Various sequences for magnetic resonance imaging (MRI) of the brain and spinal cord have may have value as surrogate biomarkers for use in future clinical trials. Here, we review the MRI findings in ALS, their clinical correlations, and their limitations and potential role as biomarkers. Methods The PubMed database was screened to identify studies using MRI in ALS. We included general MRI studies with a control group and an ALS group and longitudinal studies even if a control group was lacking. Results A total of 116 studies were analysed with MRI data and clinical correlations. The most disease-sensitive MRI patterns are in motor regions but the brain is more broadly affected. Conclusion Despite the existing MRI biomarkers, there is a need for large cohorts with long term MRI and clinical follow-up. MRI assessment could be improved by standardized MRI protocols with multicentre studies. Keywords Amyotrophic lateral sclerosisMagnetic resonance imagingMorphometryDiffusion tensor imagingMagnetic resonance spectroscopySpinal cordBiomarkersissue-copyright-statement© The Author(s) 2016 ==== Body Background Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease that mainly affects the motor system. At present, the only drug to have produced an increase in patient survival in a controlled clinical trial is riluzole [1]. The condition is always fatal; the median survival time after onset is 36 [2], although there are great individual variations [3]. Although ALS is a clinically recognizable condition (in terms of the pattern of progressive upper and lower motor neuron degeneration), the clinical presentation and progression are heterogeneous [4]. Moreover, there are now strong reasons for considering a broader aetiological and pathogenic spectrum [5]. This clinical heterogeneity is observed in other neurodegenerative diseases, such as Parkinson’s disease (PD). Nevertheless, it has been found that all PD patients display degeneration and iron overload of the substantia nigra, which might become a surrogate biomarker [6]. In ALS, iron overload in the motor cortex was confirmed [7] after being suspected in mouse models [8]. Ongoing drug trials in the field of neurodegenerative disease are mainly seeking to establish neuroprotection and neurorestoration. Since neuroprotection can never be unambiguously demonstrated in a given patient, the concept of disease-modifying drugs (with a slowing of the disease progression) has arisen. At present, the evaluation of a drug’s clinical efficacy in ALS is based on the ALS Functional Rating Scale Revised, motor tests, survival or a combination of these measures (such as the Combined Assessment of Function and Survival (CAFS)) [9, 10]. However, these endpoints are general, variable and late-stage measures of the ALS disease process and thus require the long-term assessment of large cohorts. These assessments are risky and expensive, which considerably limits the number of trials being conducted. A sensitive surrogate biomarker might help to (i) reduce the sample size in pilot studies, (ii) better define the sample size required in Phase III clinical trials (by comparison with clinical scales) and (iii) define endophenotypes for separate assessment in clinical trials. Biomarkers are typically divided into “wet biomarkers” and “dry biomarkers”. Wet biomarkers are biological substances that are measured in a body fluid (such as whole blood, serum, plasma, saliva or cerebrospinal fluid (CSF)). “Dry biomarkers” are based on functional scales, task performance, electrophysiology or imaging. Magnetic resonance imaging is a widely available and non-invasive technique, which makes it an appropriate candidate for biomarkers. Although early reports stated that ALS starts in the spinal cord, the initial site of neurodegeneration has not been unambiguously identified; disease may start in the spinal cord, in the motor cortex or at both sites simultaneously [11, 12]. The development of radiological biomarkers in ALS is facilitated by the hypothesis that the pathological changes in ALS radiate out from the initial spinal and brain areas in which the most severe upper and lower motor neuron dysfunction respectively occur (from symptom onset onwards) with variability in the predominance of upper and lower motor neuron dysfunction [13]. The neurodegeneration then progresses along the motor neurons of the ventral horn of the spinal cord and into the corresponding areas of the motor cortex [13]. Despite this modelization, the pattern of onset and progression remains complex [13]. Hence, MRI of the spinal cord should be an early, sensitive marker of subtle degeneration-related changes. However, MRI of the spinal cord is complicated by a number of technical difficulties (relative to MRI of the brain). Firstly, the magnetic field is inhomogeneous because the spinal cord is close to bone, soft tissues and air. Secondly, the spine has a small cross-sectional dimension, which thus requires high-resolution data acquisition. Thirdly, the spine is subject to motion of the CSF [14] and physiological motion due to breathing [15]. Many different MRI techniques and sequences – including voxel-based morphometry (VBM), diffusion tensor imaging (DTI), magnetic resonance spectroscopy (MRS), iron-sensitive sequences (T2*, R2* and SWI) and functional MRI (fMRI) - are available for studying ALS-related changes in the brain or spinal cord. Hence, we reviewed the correlations between MRI findings and clinical scores in ALS, and sought to establish whether particular sequences may have value as surrogate biomarkers in this disease. Methods We searched the PubMed database with the following keywords: “ALS AND MRI OR amyotrophic lateral sclerosis AND magnetic resonance imaging”, “ALS AND VBM OR amyotrophic lateral sclerosis AND magnetic resonance imaging ”, “ALS AND MR spectroscopy OR amyotrophic lateral sclerosis AND magnetic resonance spectroscopy”, “amyotrophic lateral sclerosis AND diffusion tensor imaging OR ALS AND DTI”, “amyotrophic lateral sclerosis AND functional magnetic resonance imaging OR ALS AND fMRI”. There was no limitation on the publication date but only publications in French or English were considered. We included (i) general MRI studies with a control group and an ALS group and (ii) longitudinal follow-up studies (even if a control group was lacking). We excluded therapeutic trials articles, reviews, studies focused only on particular subgroups of ALS patients such as cognitive impairment, genetic mutations. Results A total of 1336 articles were identified using database searching, 651 were recorded after duplicates removal. 405 were excluded (not ALS, not MRI or, therapeutic clinical trials, not English or French, reviews) and 130 articles were finally excluded (studies focused only on particulars subgroups of patients, other MRI sequences or lacking of a control group). We analysed 116 articles. Brain imaging Structural magnetic resonance imaging (Table 1) Table 1 The main results of volumetric studies Publication Numbers of patients/controls Main results Main clinical correlations Abdulla et al., [39, 80] 58/29 Low volume in the right hippocampus Verbal memory test performance is correlated with the left hippocampal volume. Agosta et al., [21] 25/18 GM volume reduction in right precentral, left inferior frontal cortex and temporal superior cortex. Negative with ALSFRS-R and DD. Agosta et al., [35] 44/26 Reduction in cortical thickness in the precentral, frontal, limbic, parietal, temporal and occipital lobes. DPR is correlated with the mean cortical thickeness of left sensorimotor cortex. Bede et al., [37] 39/44 GM volume reduction in the thalamus, caudate nucleus, hippocampus, left putamen. Not available. Canu et al., [23] 23/24 GM volume reduction in the precentral, right opercular and angular cortex, WM volume reduction in the frontal lobe (especially the subcortical motor areas) and the temporal lobe. Not available. Cerami et al., [36] 14/20 Low GM density in the anterior cingulate and right inferior frontal gyri. Not available. Chang et al., [22] 20 (10 ALS and 10 ALS-FTD)/22 GM volume reduction in the precentral, frontal and temporal cortex and left posterior thalamus. Not available. Chapman et al., [40] 25/22 No intergroup difference in the corpus callosum area Not available. Devine et al., [27] 30/17 GM volume reduction in the left precentral cortex, left frontal gyrus, medial frontal gyri and anterior cingulate gyri. Increased GM volume reduction in the dominant (left) motor cortex irrespective to the side of limb onset. Asymmetric GM reduction of the left somatosensory cortex and the temporal gyri in in right limb onset ALS Ellis et al., [26] 16 (8 bulbar onset and 8 limb onset)/8 GM volume reduction in the superior, middle and medial frontal cortex. WM volume reduction in the right frontal cortex. Compared to limb onset ALS, bulbar onset ALS showed GM volume reduction in the brainstem, the cerebellum and in the fusiform gyri (BA 34) and WM volume reduction along the left corticospinal tract. Grosskreutz et al., [18] 17/17 GM volume reduction in the precentral, frontal and parietal cortex. No WM atrophy. ALSFRS-R correlated positively with GM volume reduction in the right medial frontal gyrus (BA 10). Kassubek et al., [32] 22/22 GM volume reduction in the right primary motor cortex and the left medial gyrus and bilateral inferior temporal gyrus. Negative with ALSFRS-R and DD. Meadcroft et al., 2014 8/6 Post mortem study; loss of signal difference between WM and GM (mainly in the primary motor cortex). Not available. Mezzapesa et al., [19] 16/9 GM volume reduction in the precentral, frontal and parietal cortex. Not available. Mezzapesa et al., [20] 29/20 GM volume reduction in the precentral, frontal, right temporal and right occipital cortex. Different patterns of volume reduction in ALS group, high UMN burden, spinal onset and faster progression compared to controls. Sach et al., [33] 15/12 No difference -study not designed to voxel-based morphometry. Not available. Sage et al., [34] 28/26 No difference -study not designed to voxel-based morphometry. Not available. Schuster et al., [25] 93 (60 ALS, 17 with dominant UMN, 16 with dominant LMN)/67 Correlations between the site of onset and upper vs. lower motor neuron involvement, with focal reductions in cortical thickness in the precentral cortex. Different patterns of volume reduction in ALS subgroups: bulbar and spinal UMN signs, only spinal UMN signs, classical ALS, UMN ALS-variants, LMN ALS-variants or sites of onset compared to controls. GM reduction is most important in the left precentral cortex in bulbar onset vs. limb onset. No correlations of MRI with DPR, DD and ALSFRS-R. Thivard et al., [24] 15/25 GM volume reduction in the precentral, frontal and temporal cortex (especially the hippocampus), the parietal and occipital cortex, the thalamus and the cerebellum. No difference in WM. Not available. Verstraete et al., [30] 12/12 GM reduction in the right and left precentral cortex. Not available. Walhout et al., [29] 153 (112 ALS, 19 UMN, 19 LMN)/60 GM reduction in the right and (predominantly) left precentral cortex and the right paracentral cortex. Bulbar and arms ALSFRS-R subscores are correlated with cortical thickness of corresponding precentral cortex areas of the motor homunculus. DPR is correlated with the thickness of the right inferior temporal cortex, the postcentral cortex and the right paracentral cortex. Westenberg et al., 2014 112/60 GM volume reduction in the hippocampus and left subiculum. Larger ventricles are correlated to a lower ALSFRS-R score. Smaller basal ganglia, smaller limbic structures and larger ventricles are associated with shorter survival. Zhang et al., [28] 43/43 GM reduction in the left precentral cortex, left supplemental area and left postcentral gyrus. Reduction in neocortex volume, GM and WM volume and brain parenchymal fraction. GM volume reduction is most important in the motor cortex of the contralateral hemisphere of the limb of onset. ALSFRS is positively correlated with GM density in the left postcentral cortex and DPR is negatively correlated with GM density in the right precentral cortex Zhu et al., [31, 87] 22/22 GM reduction in the right and left precentral cortex. Negative with DD, DPR and all other clinical parameters. GM grey matter, WM white matter, BA Brodman area, DD disease duration, DPR Disease progression rate, UMN upper motor neuron Three methods based on a three-dimensional, T1-weighted sequence have been applied to ALS: VBM, volume analysis and the measurement of cortical thickness. As a volume-based analysis, VBM involves the automatic segmentation of grey matter and inter-subject comparison of the local grey matter density [16, 17]. After all the data have been transformed in the same stereotaxic space, the images are partitioned and corrected for the separation between white matter, grey matter and CSF as a function of voxel intensity [17]. Measurement of cortical thickness is a surface-based technique analysis in which the white matter and pial surfaces are automatically extracted by applying image analysis packages such as Freesurfer (http://surfer.nmr.mgh.harvard.edu/). Most of these techniques have evidenced atrophy in the precentral gyri [18–32]. However, some DTI studies used VBM to show that ALS vs. control differences in the fractional anisotropy (FA) of the precentral gyri were not due to atrophy [33, 34]. In other studies, cortical atrophy was not limited to the precentral gyri but extended to other regions of the frontal lobe [18–24, 26–28, 32, 35] - especially in apathetic patients with involvement of the cingulate cortex [27, 36]. Other regions involved include several parts of the temporal cortex [19–24, 32, 35], the hippocampus [24, 37–39], the parietal cortex (mostly in the post-central cortex [18, 21, 23, 28]) and the insula [24]. Occipital grey matter atrophy [19, 20, 24, 35] and cerebellar atrophy were less common [24]. When considering the basal ganglia, a low volume of the thalamus [22, 24, 37] and the caudate nucleus [37] have been observed. Few studies have evidenced a relative reduction in the volume of white matter (mainly in the frontal and temporal lobes) [23]. There were no marked differences in the corpus callosum [40]. Whole-brain analyses have evidenced low volumes in the neocortex, grey matter, white matter [28] and brain parenchyma [28, 32]. A longitudinal study showed (i) a volume reduction in the left subiculum, the right hippocampal area CA4 and the right dentatus gyrus and (ii) an enlargement of the ventricles in 39 of the 112 ALS patients [38]. Clinical correlations Specific MRI abnormalities have been correlated with clinical scores, the disease stage (as estimated with the original or revised ALS Functional Rating Scale (ALSFRS and ALSFRS-R, respectively), the speed of disease progression (i.e. the change in ALSFRS or ALSFRS-R score, divided by the disease duration or the time interval between two clinical assessments), the disease duration and the survival time. Lower ALSFRS or ALSFRS-R scores were associated with larger ventricles and greater volume loss in the basal ganglia [38], and volume changes in several parts of the frontal lobe (such as Brodman area 10) [18]. In another study, a higher ALSFRS-R progression rate was associated with MRI changes in the left sensorimotor area; cortical thickness was lower in patients with predominantly upper motor neuron impairment (especially in the precentral gyrus and the left paracentral lobule), whereas patients with predominantly lower motor neuron impairment showed less difference relative to controls [20]. Grey matter density was lower in the cingulate and right inferior frontal gyrus in patients with impaired emotional empathy [36]. Poor survival in ALS is correlated with volume loss in the basal ganglia and limbic structures [38]. The initial side of limb weakness in ALS patients was not correlated with contralateral brain cortical loss of volume [27]. Although atrophy was predominant in the left sensorimotor cortex [27, 28], it was not related to the side of onset [27]. A subgroup analysis (by side of onset) found more widespread atrophy in the contralateral brain cortex than in the group of ALS patients as a whole [28]. Precentral cortical thickness was lower in patients with upper motor neuron clinical features than in those with lower motor neuron clinical features and in the ALS group as a whole [29]. In a study with a post mortem T1 sequence and histological assessment, an abnormally small difference in intensity between cortical grey matter and subcortical white matter was observed not only in the motor cortex (including the precentral gyrus, supplemental motor area and premotor cortex) but also in somatosensory areas and the primary visual cortex. These signal abnormalities were linked to neuron loss and an elevated number of astrocytes [41]. Diffusion tensor imaging (Table 2) Table 2 The main results of DTI studies Publication Numbers of patients/controls Fractional anisotropy Mean diffusivity or apparent diffusion coefficient Radial and axial diffusivity Main clinical correlations Abe et al., [64] 7/11 Low in parts of the right frontal lobe and in the left subcortical precentral area Not available Agosta et al., [21] 25/18 Low in parts of CSTs, the CC, parts of the frontal lobe and parts of the temporal lobe No difference Negative Bede et al., [75] 27/42 Low along CSTs, the CC, cerebellum, brainstem, occipital lobe and opercular and insular regions RD was low in the cerebellum, brainstem, occipital lobe and opercular and insular regions Not available Canu et al., [23] 23/14 Low in subcortical parts of CSTs and other parts of the frontal lobe Elevated in parts of the CSTs, left postcentral gyrus, left insula, parts of the left temporal lobe, right angular cortex, parts of the frontal lobe, CC, parts of the occipital lobe and parts of the cerebellum Correlation between MD in bilateral orbitofrontal region and DD Chapman et al., [67] 21/21 Low in the CC, corona radiata and PLIC RD was low in the CC, corona radiate and PLIC. AD was elevated in the left corona radiata and the internal and external capsules Correlation between FA in parts of CC and ALSFRS-R and negative correlation with DD, correlation between RD in parts of CC and DD Ciccarelli et al., [45] 26/41 Low along CSTs, the CC, anterior limb of the internal capsule, external capsule, parts of frontal lobe WM and postcentral gyri Correlation between DPR and FA in left cerebral peduncle, right PLIC, right corona radiate right WM adjascent to the precentral gyrus and CC Ding et al., [74] 10/10 Elevated in the PLIC Not available Ellis et al., [51] 22/20 Low along CSTs Elevated along CSTs Correlation between mean diffusivity in CST and DD and between FA along CSTs and ALS severity scale and spasticity scales. Filippini et al., [56] 24/24 Low along CSTs and in the CC Elevated RD in WM linked to the primary motor and premotor cortex and the CC Inverse correlation between FA and UMN score along CSTs, correlation between FA and ALSFRS-R and DD along CSTs Foerster et al., [55] 29/30 Low along CSTs Correlation between FA and ALSFRS-R along CSTs Iwata et al., [49] 31/31 Low along CSTs Negative Iwata et al., [50] 18/19 Low along CSTs and in the motor part of the CC Elevated along CSTs Correlation between DD and FA along CSTs. Inverse correlation between FA in CSTs and global and localized UMN impairment score and between FA along CST and UMN rapidity index. Kassubek et al., [12, 73] 111/74 Low along CSTs Not available Keil et al., [61] 24/24 Low in parts of CSTs, supplementary motor area, CC, parts of the frontal lobe and the parahippocampal area Elevated in motor areas, parts of CSTs, parts of the frontal and temporal lobe, and the postcentral gyrus Correlation between FA along CSTs and ALSFRS-R. Correlation between executives functions (sfs36 score) and FA in cerebellum. At 6 months, negative correlation between FA along CSTs and DD correlation between FA along CSTs and frontal WM and ALSFRS-R. Correlation between FA in CST at brainstem level and executives functions, negative correlation between ADC in the cerebellum and parahippocampal gyri and executives functions (sfs36) Keller et al., [68] 33/30 Low in the corona radiata and CC Negative Liu et al., 2014 19/13 Low along CSTs Correlation between FA along left CST and ALSFRS-R Menke et al., [77] 21/0 (follow-up study) Progressive reduction in the PLIC Correlation between FA at baseline and DPR Metwalli et al., [69] 12/19 Low along CSTs, the CC Elevated along CSTs, the CC AD and RD were elevated along CSTs, RD was elevated in the CC, parts of frontal and parietal lobe Negative Muller et al., 2011 19/19 Low in parts of CSTs, the parahippocampal area, insula and brainstem Correlation between FA in parts of CSTs and ALSFRS-R Nickerson et al., [46] 2/0 (follow-up study) A linear reduction along CSTs during one year follow-up Not available Poujois et al., [65] 19/21 Low in CSTs from the left corona radiate to the precentral gyrus and in both cerebral peduncles Muscular strength is lower on the right side corresponding to the lower FA in the left CST Prell et al., [60] 17/17 Low in parts of CSTs and the cingulate gyrus Elevated parts of CSTs, parts of frontal lobe, cingulate gyrus, parahippocampal region, CC, cerebellum. Correlation between FA in internal capsule and contralateral strength of the lower limb. Different patterns in FA and ADC of bulbar and limb onset compared to controls. Prudlo et al., [70] 22/21 Low throughout the CSTs, the anterior limb of the internal capsule, thalamic radiations, the CC, association fibres and the middle cerebellar peduncle Correlation between FA in many voxels of a whole DTI brain analyses with ALSFRS-R Pyra et al., [59] 14/14 Low in left precentral gyrus Correlation between spasticity and ADC in contralateral precentral gyrus Rajagopalan et al., [63] 47/10 Low in the left subcortical motor area and right PLIC AD was low in the PLIC. RD was elevated in the PLIC of patients with T2 hyperintensities in the CSTs Not available Rosskopf et al., [57] 100/93 Low along CSTs Correlation between FA along CSTs and ALSFRS-R Sach et al., [33] 15/12 Low in parts of CSTs, premotor areas, CC and right thalamus Low FA in patients without UMN signs in parts of CST, CC and right thalamus. No correlation available with clinical score Sage et al., [34] 28/26 Low along CSTs and the right postcentral gyrus Elevated in parts of CSTs Correlation between FA and ALSFRS in several parts of CSTs and in prefrontal lobe Sage et al., [52] 28/26 Low in parts of CSTs, parts of the frontal lobe, insula, hippocampus, cerebral peduncles and CC Elevated along CSTs, hippocampus, insula, parts of the temporal and frontal lobe and CC Correlation between FA along CSTs, in prefrontal area and ALSFRS. Negative correlation between MD along CST, hippocampus, cerebellum, parietal and temporal lobe and ALSFRS. Sarica et al., [53] 14/14 Low in right CSTs and left anterior thalamic radiations Elevated in right CSTs, cingulum and left anterior thalamic radiations RD elevated in right CSTs and left anterior thalamic radiations. AD elevated in the right cingulum Negative (p ≤ 0.05) Schirimrigt et al., 2007 10/20 Low along CSTs (nb: study with a technical objective) Correlation between FA along CSTs and DD Stagg et al., [47] 13/14 Low along CSTs Elevated along CSTs Negative Tang et al., [58] 69/23 Low along CSTs, in frontal WM and the genu of the CC Elevated in the centrum semi-ovale and frontal and parietal WM Not available Thivard et al., [24] 15/25 Low along CSTs, premotor cortex, right thalamus, insula, parts of parietal lobe Elevated in the motor cortex, premotor cortex, insula, hippocampus, and right superior temporal gyrus Correlation between FA along CST, insula, premotor cortex, cingulum, precuneus, CC and ALSFRS-R, negative correlation between FA in CC and centrum semiovale and DD Verstraete et al., [30] 12/12 Low in the rostral part of CSTs and the CC Not available Wang et al., [71] 16/17 Low CST volume in DTI Negative Yin et al., [48] 8/12 Low along CSTs Not available Zhang et al., [62] 17/19 Low in parts of CSTs Elevated in parts of the CSTs Correlation between FA in right superior CST and ALSFRS-R and motor subscore of ALSFRS-R FA fractional anisotropy, MD mean diffusivity, ADC apparent diffusion coefficient, RD radial diffusivity, AD axial diffusivity, CST corticospinal tract, PLIC posterior limb of the internal capsule, CC corpus callosum, DD disease duration, DPR disease progression rate, UMN upper motor neuron, LMN lower motor neuron DTI (also known as magnetic resonance tractography) is based on the random diffusion (Brownian motion) of molecules. In a spherical volume, the diffusion of water has no main direction and its diffusion in the three directions (λ1, λ2 and λ3) is equally likely. However, this is no longer the case in an ellipsoid or cylindrical volume and the diffusion is anisotropic. λ1 is the volume’s main axis (axial diffusivity, AD) and λ2 and λ3 are the minor axes (radial diffusivity, RD). By analysing the diffusion of water in three directions, four parameters can be defined: FA, RD, AD and mean diffusivity (MD, which is the average of diffusion in the λ1, λ2 and λ3 axis, also referred to as the apparent diffusion coefficient (ADC)). The first three parameters describe the spatial variation of water movement and are related to the orientation of the studied structures. In contrast, MD corresponds to the mean displacement of the water molecules within the volume. The main application of DTI is the analysis of the anatomical bundles that compose the white matter of the brain and the spinal cord [42–44]. Whereas DTI corresponds to the analysis of the above-mentioned parameters in a defined volume, tractography corresponds to the analysis of a complete white matter tract. Fractional anisotropy in ALS Low FA along entire corticospinal tracts (CSTs) is the most common observation in patients with ALS [34, 45–58]. However, some studies have only reported low FA in one or several parts of a CST: the subcortical white matter of the precentral gyrus [23, 24, 30, 33, 59–65], the corona radiata [24, 60–62, 66–68], the posterior limb of the internal capsule (PLIC) [24, 33, 60, 62, 63], the cerebral peduncles [21, 59, 61] and the pons [61]. Lastly, only one study failed to observe a difference in FA along a CST in ALS patients by region of interest approach whereas tract based spatial statistics (TBSS) showed decreased FA in corona radiate and corpus callosum [68]. Low FA has also been observed in other regions: the frontal lobe (excluding the CST) [21, 24, 33, 34, 45, 52, 56, 57, 61, 64], the cingulum [60], the corpus callosum [21, 33, 45, 50, 52, 58, 61, 67–69], the parietal lobe [24, 34, 45], the temporal lobe [21], parahippocampal areas [61], the hippocampus [34], the insula [34, 52], the cerebellum [61] and the thalamus [24, 33, 53]. In a whole-brain comparison, many other small areas of ALS-modified white matter tracts were observed [70]. A tractography study revealed that the CST volumes of ALS patients are lower than in controls [71]. A DTI study using a tract of interest-based fibre tracking found that the radiological results mirrored the neuropathological staging system [72], suggesting that the disease process extended in the following sequence: the corticospinal tract (stage 1), the corticorubral and corticopontine tracts (stage 2), the corticostriatal tract (stage 3) and the proximal portion of the perforant pathway (stage 4) [73]. The researchers suggested that this MRI-based staging system could be used as a surrogate marker of disease progression in clinical studies of ALS. Diffusivity in ALS The most common observation in patients with ALS is a relative increase in MD, reflecting changes in white matter tracts. The MD was elevated along entire CSTs [51, 53, 69] and in parts of CSTs, including subcortical white matter in the precentral gyrus [23, 34, 58, 60, 61], the corona radiata [34, 60], the PLIC [23, 34, 60, 74], the cerebral peduncles and the pons [34]. Diffusivity was also elevated in several parts of the frontal lobe [23, 24, 52, 58, 60, 61], the temporal lobe [23, 24, 52, 61], the hippocampal region [24, 34, 52], the parahippocampal region [60], the parietal lobe [23, 58, 61], the insula [23, 24, 52], the cingulum [53, 60], the occipital lobe [23], the cerebellum [23, 60] (especially in C9ORF72-positive patients) [75], the corpus callosum [23, 52, 60, 69] and the thalamus [53]. Conversely, two studies did not detect any differences in diffusivity in ALS patients vs. controls [21, 63]. Radial and axial diffusivity in ALS Several studies have found elevated RD and AD values in the right CST, the right cingulum and the left anterior thalamic radiations (on the basis of changes in FA or MD) [53]. In the PLIC, AD is low [63] and RD is elevated [67] - especially in patients presenting with T2 hyperintensities in CSTs [63]. RD is elevated in white matter connected to the primary motor cortex, the premotor cortex [56] and the corpus callosum [56, 67, 69]. Lastly, RD and AD are elevated in the cerebellum of C9ORF72-positive patients [75]. Longitudinal studies In longitudinal analyses, low FA is observed in the right superior CST [62], precentral subcortical regions, mesencephalic CSTs and parts of cerebellum [61]. FA was found to fall in a progressive, linear manner along the CSTs [34, 46, 76], and the diffusivity of the external and internal capsule was elevated [61]. AD is elevated in CSTs [77]. However, a recent longitudinal study found that white matter defects (as assessed by DTI) progressed much slowly that in the grey matter defects - notably in the basal ganglia [78]. Clinical correlations Many studies have found that disease progression (as defined by the change over time in the ALSFRS or ALSFRS-R score) is correlated with the changes in FA in CSTs [24, 34, 45, 61, 62, 77, 79]. Furthermore, the disease duration is correlated with FA in entire CSTs or parts of CSTs [24, 50, 54, 56, 61, 77]. Conversely, the disease duration was negatively correlated with FA in the cerebellum [61], the subcortical white matter of insula, the ventrolateral premotor cortex, the cingulum, the precuneus and the splenium of the corpus callosum [24]. Lower-limb muscle strength was correlated with the contralateral FA in CSTs [60]. Diffusivity values in CSTs and in the parietal lobe, temporal lobe and cerebellum were also correlated with disease progression [52, 79]. Diffusivity in CSTs was correlated with disease duration [79], and contralateral diffusivity in CSTs was correlated with spasticity [59]. Whereas both limb-onset patients and bulbar-onset patients differed from controls in terms of FA along CSTs, the two patient groups had similar values [60]. The CSTs were more impaired in bulbar-onset patients than in limb-onset patients when considering FA, MD and RD but not AD [80]. Tractography revealed a large number of differences between controls and patients defined as definite/probable according to the El Escorial criteria, whereas there were fewer differences between controls and a possible/suspected group [79]. Neurophysiological correlations FA in CSTs was related to the central conduction time in a transcranial magnetic stimulation study [33, 49] (Iwata et al., [49]; Sach et al., [33]). Diagnostic accuracy An individual patient data meta-analysis of CST studies concluded that DTI lacks sufficient diagnostic discrimination [81]. The pooled sensitivity and specificity were only 0.68 and 0.73, respectively. Researchers have also suggested that a multimodal approach (combining DTI with methods such as MRS) may be more promising [55]. Magnetic resonance spectroscopy (Table 3) Table 3 main results of magnetic resonance spectroscopy studies Publication Numbers of patients/controls Main results Main clinical correlations Bowen et al., [94] 18/12 Cho and ins were elevated in the MC. NAA and Cr were correlated in left MC Correlation between ins in MC and UMN disability, negative correlation between Naa in MC and UMN disability, higher Cho in sever UMN disability group Cervo et al., [90] 84/28 NAA/(Cho + Cr) was low in the MC Negative Foerster et al., [95] 10/9 Low gamma aminobutyric acid in the MC but not in WM Negative Foerster et al., [55] 29/30 NAA and gamma aminobutyric acid were low and ins was elevated in the left MC Negative correlation between gamma aminobutyric acid in MC and DD, correlation between Naa peak in MC and ALSFRS-R Govind et al., [92] 38/70 NAA was low and Cho was elevated in most parts of CST, and Cho/NAA was elevated in all parts of the CSTs Negative correlation between Cho/Naa in the left entire CST and forced vital capacity, negative correlation between Cho/Naa in the left CST and right and left finger tap rate, negative correlation between Cho/Naa in left MC and semiovale centrum and right finger tape or forced vital capacity Han et al., [85] 15/15 NAA/Cr peak was low in the MC and PLIC, Glu/Cr and Glu + Gln/Cr peaks were low in the MC and PLIC Negative correlation between Glu + Gln/Cr with Norris score Kalra et al., [88, 89] 63/18 NAA/Cho and NAA/Cr was low in the MC, and Cho/Cr was elevated in the MC Relation between decreased Naa/Cho in the MC and reduced survival. Kalra et al., [88, 89] 17/15 NAA/Ins, NAA/Cr and NAA/Cho were low in the MC, and Ins/Cr was elevated in the MC Negative (p ≤ 0.05) Liu et al., [87] 19/13 NAA/Cr was low in the MC Negative Lombardo et al., [79] 32/19 NAA/Cr was low and ins/Cr and Cho/Cr were elevated in the MC. Abnormalities were correlated with the El Escorial score Pohl et al., [91] 70/48 NAA, Pcr + Cr, NAA/Cho and NAA/(Pcr + Cr) were low in the MC. At 12 months, NAA/Cho was low and Cho/(Pcr + Cr) was elevated Not available Pyra et al., [59] 14/14 NAA/Cho and NAA/Cr were low in the MC and corona radiata, and Cho/Cr was elevated in the MC Correlation between Naa/Cho peak in MC and DD, negative correlation between Naa/Cho in MC and corona radiata and DPR Rooney et al., [86] 10/9 NAA/(Cho + Cr) was low in the MC and CST but not in other regions Correlation between Naa/(Cho + Cr) in the MC and maximum finger tape rate Stagg et al., [47] 13/14 NAA was low along the CSTs Correlation between Naa peak along CSTs and ALSFRS-R Unrath et al., [96] 8/0 Progressive decrease over time in the NAA peak throughout the MC Progressive decrease of Naa/(Cr + Cho) in the less affected hemisphere. Correlation between Naa and the more or less affected side (ALSFRS subscore). Correlation between Naa/(Cr + Cho) and the less affected side (ALSFRS subscore) Verma et al., [93] 21/10 NAA/Cho was low in the right lingual gyrus, parts of the occipital lobe, left supramarginal gyrus and left caudate. NAA/Cr was low in the right MC, left frontal inferior operculum, right cuneus, parts of the occipital lobe, left caudate and left Heschl gyrus Not available MC motor cortex, CST corticospinal tract, PLIC posterior limb of the internal capsule, WM white matter, NAA N-acetylaspartate, Cho choline, Gln glutamine, Glu glutamate, PCr phosphocreatine, Cr creatine or creatine + Pcr (the distinction is not relevant for interpretation), DD disease duration, DPR disease progression rate MRS is a non-invasive means of assessing brain metabolism, based on the chemical properties of molecular structures. Results are expressed as peaks. The main brain metabolites monitored with MRS include N-acetyl-aspartate (NAA), choline (Cho, a cell membrane marker), creatine (Cr) and phosphocreatine (Pcr). The Cr + Pcr peak is a marker of energy metabolism; the level of Cr is assumed to be stable and is used to calculate the metabolite ratio. At a field strength of 1.5 T, the peaks for glutamine (Gln), glutamate (Glu, an excitatory neurotransmitter) and gamma aminobutyric acid (GABA, an inhibitory neurotransmitter) overlap to form a complex referred to as “Glx”; higher field strengths are required for more accurate quantification of this metabolites. The simple sugar myo-inositol (Ins) is absent from neurons but present in glial cells. Hence, elevated brain levels of Ins are associated with glial proliferation, whereas low levels are associated with glial destruction [82–84]. The main results for the CST concern the ratio between the NAA peak on one hand and the Cho and Cr + Pcr peaks on the other. In ALS patients, low levels of NAA have been observed in the precentral corticosubcortical region [85–91], in the PLIC [85] and in entire CSTs [55, 59, 86, 92]. Glu and Glu + Gln were found to be elevated in the precentral corticosubcortical region and in the PLIC [85], whereas Cho was found to be elevated in the precentral region [88, 93, 94] and along the CSTs [47, 59, 92]. Levels of the inhibitory neurotransmitter GABA were low in the precentral cortex [55, 95], and levels of Ins were low in the precentral corticosubcortical region [89]. Whereas the above-mentioned studies used defined voxels for MRS acquisition, whole-brain spectroscopy has revealed low NAA/Cho or NAA/Cr ratios in many other regions [93]. Follow-up studies have shown a decrease in the NAA peak in the precentral cortex (with no differences for other metabolites) [96], a decrease in the NAA/Cho peak and an increase in the Cho/Cr peak [91]. Clinical correlations Patients with definite or probable ALS (according to the El Escorial criteria) had low levels of NAA and high levels of Ins and Cho in the precentral cortex. In patients with possible or suspected ALS, only Cho and (on the left side) Ins were elevated. In another study of suspected ALS patients, only right-side Ins and right-side Cho were elevated [79]. Low NAA and high Cho and Ins levels are associated with clinical disease severity [94]. Lower NAA levels in the precentral gyrus, disease duration and disease progression were intercorrelated [59], and a low NAA/Cho ratio was associated with poor survival [88]. Diagnostic accuracy In a recent study, the combination of MRS and DTI yielded a sensitivity of 0.93 and a specificity of 0.85, whereas the DTI data alone gave values of 0.86 and 0.70, respectively [55]. A criterion combining the MRS data, hypointensity in the precentral gyri and hyperintensity in the CSTs yielded a sensitivity of 0.78 and a specificity of 0.82; when each of the three parameters was considered alone, the sensitivity and specificity values did not exceed 0.71 and 0.75, respectively [90]. Iron imaging Iron overload appears to be involved in the physiopathology of ALS [8, 97]. However, very few studies have analysed the iron content in the brain of ALS patients. A variety of techniques are available, including relaxometry (T2, T2* or R2*) and the susceptibility-weighted imaging [98, 99] (SWI, which is probably more sensitive than relaxometry [100]). Multi-echo T2* and R2* sequences are sensitive to magnetic field inhomogeneities. The value of R2* is measured by voxel-by-voxel modelling of the exponential decrease in signal [101, 102]. In post-mortem studies, R2* values have been found to be correlated with the iron content [103]. Susceptibility-weighted imaging and quantitative susceptibility mapping (QSM) SWI is derived from T2* sequences and can be described as a flow-compensated, high-spatial-resolution T2* weighted sequence. The main advantage over conventional T2* sequence is better contrast [101]. QSM is based on the analysis of phase information. The main sources of phase shifts in biological tissues are the iron, calcium, lipid and myelin contents [104]. The value of QSM in ALS was confirmed in a post-mortem study, in which the iron content of grey matter was assessed more accurately than that of white matter [105]. The initial studies in the field noted a shorter T2 time in the motor cortex of most ALS patients (vs. controls) [106], although this was not confirmed a few years later [107]. Later studies found a correlation being T2* shortening and iron deposition in the microglia of the motor cortex (according to a post-mortem examination) [7, 106, 108]. Iron overload in the motor cortex was also described in SWI studies [109] and qualitative studies [100]; in the later study, there were no differences between ALS patients and controls in terms of T2 or T2*. A few quantitative studies have looked at white matter or deep grey matter [109, 110]. There was a trend towards iron accumulation in CST white matter in a relaxometry study [110]. Iron overload was also present in the red nucleus, substantia nigra, globus pallidus, putamen [109] and caudate nucleus [110]. The use of SWI sequences revealed widespread iron overload and myelin defects in the white matter of all brain lobes and corpus callosum [111]. A retrospective study of iron deposition in motor cortex found that quantitative susceptibility mapping was more accurate than T2*, T2 or FLAIR sequences [112]. A follow-up study has highlighted the progression of a T2 hypointense area in the motor cortex after 6 months, which was strongly and negative correlated with the ALSFRS score [108]. Lastly, changes in SWI sequences in the corpus callosum are correlated with the ALSFRS-R score [111]. Imaging of the spinal cord The few studies to have analysed the spinal cord revealed atrophy in the cervical and upper thoracic regions [113–115]. A DTI study reported low FA, elevated RD and a high magnetization transfer ratio in lateral segments of the cervical spinal cord [114]. Atrophy of the cervical and upper thoracic spinal cord progressed over time, whereas the DTI changes were relatively stable [115]. This observation agrees with the results of a longitudinal brain MRI study showing that grey matter defects (but not white matter defects) progressed over time [78]. Sensory pathways were also involved, with a decreased FA and an elevated MD and elevated RD in the posterior part of the cervical spinal cord [115, 116]. MRS of the cervical spinal cord revealed low NAA/Cr + Pcr and NAA/Ins peaks and an elevated Ins/Cr peak [117]. Clinical correlations Cervical and upper thoracic spinal cord atrophy was correlated with upper arm muscle strength, as measured by manual testing [114]. FA in the cervical spinal cord was correlated with the ALSFRS score. The rate of atrophy in a follow-up study was correlated with the rates of changes of an arm ALSFRS-R score and an arm strength score (in a manual test) [115]. The NAA/Cr + Pcr and NAA/Ins peaks were positively correlated with the ALSFRS-R score and forced vital capacity but negatively correlated with disease progression [117]. Functional MRI Brain fMRI is a recent technique based on blood flow and blood-oxygen-level dependent (BOLD) contrast, which in turn are based on the neurovascular consequences of neuronal activation [118]. The paramagnetic properties of deoxyhaemoglobin lead to signal reduction in gradient echo sequences [101]. Brain fMRI sequences are acquired in the resting state or during performance of a motor, auditory, cognitive, or visual task [119–121]. The resting state sequences are quite similar to T2* sequences [101]. In the resting state, functional connectivity is based on fluctuations of the BOLD signal. In general, seven “default” networks can be identified: default mode, executive control, visual salience, sensorimotor, dorsal attention and auditory networks [118, 122]. Functional connectivity between specific brain regions and/or networks and the rest of the brain can then be analysed [118]. Resting-state analysis can also provide information about local activation (by analysing the amplitude of low-frequency fluctuations (ALFF), for example). The resting state Resting-state studies have evidenced the widespread reorganization of functional connectivity in ALS patients. This reorganization can be observed as either elevated functional connectivity or low functional connectivity. Functional connectivity was found to be low in several parts of the frontal lobe (the right orbitofrontal cortex, the left inferior frontal cortex [123] and especially the motor cortex [124]) but was elevated in the parietal lobe (the left precuneus and the right angular gyrus [123]) and the left frontoparietal network [123]. This agrees with the elevated sensorimotor connectivity of the left sensorimotor cortex with several regions of the right hemisphere (mainly the cingulate cortex, parahippocampal gyrus and cerebellum crus II) [125]. When compared with alterations of the CSTs (according to FA measurements), there were also changes in right sensorimotor connectivity; the connectivity was elevated if the CST was normal in a DTI analysis but was low if the CST was abnormal [125]. A network-by-network analysis revealed low activity in the default mode network [126] and elevated activity in the sensorimotor network (both of which include parts of the frontal lobe [126, 127], although not the same ones). Thalamic connectivity is also elevated [127]. Resting-state activity (as measured by the ALFF) was elevated in the frontal lobe (the left anterior cingulate, right superior frontal cortex [31] left medial frontal gyrus and right inferior frontal gyrus [128]), the temporal lobe (the right parahippocampal cortex and the left inferior temporal cortex) and the occipital lobe (the middle occipital cortex) [31] but low in the right fusiform gyrus, the visual cortex and right postcentral gyrus [128]. The resting-state connectivity in the frontal cortex is correlated with disease progression and duration [128]. The connectivity between the left sensorimotor cortex and the right parahippocampal gyri and fourth cerebellar lobule is correlated with the ALSFRS-R [125]. Connectivity of the dorsal part of the precentral gyri was correlated with hand strength [124]. Activity in the right parahippocampal gyri was correlated with disease progression, whereas activity in the left anterior cingulate and left temporal gyrus was correlated with cognitive parameters [31]. Activation in fMRI In motor tasks, activation of the motor cortex was elevated [129–131] and spread widely across the supplementary motor area, the premotor cortex [132] and the parietal somatosensory cortex [129, 132]. Inhibition was low [130]. Activation of the ipsilateral sensorimotor cortex was possibly linked to a compensatory mechanism in ALS [133]. A visual task was associated with poor activation of the secondary visual cortex and strong activation of associative areas, whereas an auditory task produced delayed activation in the secondary auditory cortex, and somatosensory stimulation produced prolonged activation of the right inferior frontal gyrus and right posterior insular cortex [134]. In a modified “go”/“no-go” task, activation of the primary motor cortex, other motor areas, the frontal lobe, the insula and the cerebellum was found to decrease over a three-month period [135]. Using another type of stimulus, activation of the hippocampus also fell after 3 months [135]. In a theory-of-mind task designed to evaluation the mirror neuron system, ALS patients displayed abnormally strong activation of the right anterior cortical areas [136]. A link between the severity of motor weakness and the pattern of activation has never been observed, although the ALSFRS-R score was correlated with the signal change in the sensorimotor cortex in one study [129]. When considering the clinical phenotype, only a bulbar-onset subgroup showed low activation of the motor cortex during tongue movements [131]. Shifts in cerebellar or hippocampal activation were correlated with the ALSFRS-R score. Discussion When considering the data as a whole, the most disease-sensitive MRI patterns are located in motor regions (and especially along the CST from the cortex to the spinal cord). However, ALS also affects the brain more broadly; other parts of the frontal lobe, the temporal lobe, the hippocampus, the parietal lobe, the cingulum and the insula are commonly involved, whereas the occipital lobe is less frequently involved. For most MRI parameters, patient vs. control differences are mostly apparent in the left (dominant) hemisphere. The various MRI patterns appear to reflect disease severity, progression and duration. With fMRI, abnormally high levels of activation in the motor cortex are probably related to the disease-related loss of inhibition. Excessive activation appears to be associated with compensation for the loss of function caused by neuron loss. Limitations of neuroimaging as a biomarker for ALS The low average number of participants represents the main limitation: very few published studies have featured more than 30 subjects per group (i.e. ALS and control groups). Together with the differences in the proportions of the various patient phenotypes, this explains the high variability of the results - particularly outside the motor cortex and the CSTs [137]. Only a small proportion of morphometric studies included a whole-brain analysis. The proportion is higher for DTI studies and low (only one study) for MRS. The selection of regions of interest may over-emphasize changes in the CSTs and the motor system and under-emphasize changes outside these areas. Very few studies had a longitudinal design, which is nevertheless required for the full interpretation of abnormalities. The main difficulty in setting up longitudinal studies relates to the rapid disease progression and the appearance of respiratory failure, which prevents patients from lying in the MRI for any length of time. In studies longer than 6 months, the drop-out rate is usually over 50 %. fMRI studies have highlighted both elevated activity and low activity in several parts of the brain. The main physiopathological explanations relate to a physiological compensation response or elevated activity due to the loss of inhibitory mechanisms [127]. Spinal cord imaging is a rarely applied imaging modality in ALS, despite the pivotal involvement of the anterior horn of the spinal cord in ALS [137]. Spinal cord imaging appears to be most commonly used in studies of multiple sclerosis, since all studies in this disease context found differences (relative to controls) in the cervical spinal cord. To the best of our knowledge, no studies of the whole spinal cord have been published – perhaps as a result of technical obstacles. The biomarker concept Currently, the diagnosis of ALS is based on the revised El Escorial criteria (also known as the Airlie House criteria), which were established in 1998 and published in 2000 [138]. Although diagnosis during the course of the disease has become relatively easy, the onset of ALS can be masked by clinical overlap with many other diseases [139]. The ideal biomarker is not only sensitive and specific (for various disease phenotypes) early in the disease process but must also easy to access [140]. In ALS, the typical time interval of 12 months between the initial symptoms and disease diagnosis makes this challenge tougher still [140–142]. The best way of determining early biomarkers for ALS may be to monitor cohorts of presymptomatic patients identified by genetic studies [143]. Although wet biomarkers based on blood or CSF analysis are relatively accessible, there are many candidates and study-to-study reproducibility appears to be quite low [5, 140]. Perspectives There is currently a need for a robust clinical-radiological description of ALS using MRI as a function of the disease progression, the various endophenotypes and left/right asymmetry (in terms of upper and lower motor neuron impairment, bulbar involvement, and the cognitive profile ranging from frontotemporal dementia, apathy and dysexecutive syndrome to mild attention disorders). Despite the few number of studies, spinal cord could be a promising region of interest, although this approach is currently limited by technical post-processing difficulties and a sizeable number of large scale and longitudinal studies involving the spinal cord is needed to build stronger MRI biomarkers of the spinal cord. It may be of value to combine several techniques (e.g. atrophy mapping and measurements of the iron content in the grey matter). The results of MRI of the spinal cord are well correlated with the clinical impairment and might be a good surrogate marker. A comprehensive analysis of MRI parameters in very large cohorts of ALS patients might reveal a radiological surrogate marker for use in clinical trials. Several points should be considered for large scale studies: in 2010, the first Neuroimaging Symposium in ALS (NiALS) had the aim of establishing consensus about the various applications of MRI to the study of ALS and the possibility of multicentre collaboration. Consensus criteria for the main MRI sequences (VBM, DTI, fMRI and spectroscopy) and the main clinical dataset were established [144]. The aim of this consensus was to lead large multicentre and longitudinal studies in imaging research or therapeutic trials [144]. <In our days, studies might better assess the most recent discovers in ALS such as genetic parameters (especially C9ORF72 expansion), the knowledge in cognitive and behavioural impairments even if the patients do not fulfil criteria for frontotemporal dementia [145]. There is still a need for longitudinal studies especially in presymptomatic and early phase of the disease [145]. Following this multicentre aim, a very recent study showed in 253 patients and 189 controls from eight international ALS specialised centre more widespread white matter tract change pooling the analyses than in single-centre analysis and reached estimation of neuropathological changes [146]. Conclusion The MRI biomarkers appear to be well correlated with disease severity, duration and progression and are more sensitive in the brain and spinal cord motor regions. The built of MRI biomarkers is limited by the clinical various phenotypes but also by the lack of large and longitudinal studies. To date large cohorts with multicentre studies using a standardized MRI protocol are needed. Abbreviations ADAxial diffusivity ADCApparent diffusion coefficient ALFFAmplitude of low frequency fluctuations ALSAmyotrophic lateral sclerosis ALSFRSAmyotrophic lateral sclerosis functional rating scale ALSFRS-RAmyotrophic lateral sclerosis functional rating scale revised CAFSCombined assessment of function and survival ChoCholine CrCreatine CSFCerebro spinal fluid CST(s)Cortico-spinal tract(s) DTIDiffusion tensor imaging FAFractional anisotropy fMRIFunctional magnetic resonance imaging GABAGamma aminobutyric acid GlnGlutamine GluGlutamate GlxGlutamine and glutamate Insmyo-inositol MDMean diffusivity MRIMagnetic resonance imaging MRSMagnetic resonance spectroscopy NAAN-acetyl-aspartate NiALSNeurimaging symposium in amyotrophic lateral sclerosis PcrPhosphocreatine PDParkinson’s disease PLICPosterior limb of internal capsule QSMQuantitative susceptibility mapping RDRadial diffusivity SWISusceptibility weighted imaging VBMVoxel-based morphometry The authors thank Dr. David Fraser (Biotech Communication, Damery, France) for editing the article. Funding Not applicable. Availability of data and materials Data and the findings are fully available without restriction upon request to the authors. Authors’ contribution GG, CM, VDB, CD, LD and DD: conception of the study. GG and DD: writing the first draft and the final draft. CM, VDB, CD, RL, PFP, MMEM and LD: review and critical coment. All authors have read and approved the final version of the manuscript. Competing interests The authors declare that they have no competing interests about this manuscript. Consent for publication Not applicable as this manuscript is a systematic review and not a clinical trial. Ethics approval and consent to participate Not applicable as this manuscript is a systematic review and not a clinical trial. ==== Refs References 1. Bensimon G Lacomblez L Meininger V A controlled trial of riluzole in amyotrophic lateral sclerosis. ALS/Riluzole Study Group N Engl J Med 1994 330 9 585 591 10.1056/NEJM199403033300901 8302340 2. Soriani M-H Desnuelle C Épidémiologie de la SLA Rev Neurol (Paris) 2009 165 8–9 627 640 10.1016/j.neurol.2009.04.004 19524991 3. Beghi E Logroscino G Chiò A Hardiman O Mitchell D Swingler R The epidemiology of ALS and the role of population-based registries Biochim Biophys Acta (BBA) - Mol Basis Dis 2006 1762 11–12 1150 1157 10.1016/j.bbadis.2006.09.008 4. Mitchell JD Borasio GD Amyotrophic lateral sclerosis Lancet 2007 369 9578 2031 2041 10.1016/S0140-6736(07)60944-1 17574095 5. Turner MR Bowser R Bruijn L Dupuis L Ludolph A Mcgrath M Mechanisms, models and biomarkers in amyotrophic lateral sclerosis Amyotroph Lateral Scler Front Degener 2013 14 S1 19 32 10.3109/21678421.2013.778554 6. Devos D Moreau C Devedjian JC Kluza J Petrault M Laloux C Targeting Chelatable Iron as a Therapeutic Modality in Parkinson’s Disease Antioxid Redox Signal 2014 21 2 195 210 10.1089/ars.2013.5593 24251381 7. Kwan JY, Jeong SY, Van Gelderen P, Deng H-X, Quezado MM, Danielian LE, et al. Iron Accumulation in Deep Cortical Layers Accounts for MRI Signal Abnormalities in ALS: Correlating 7 Tesla MRI and Pathology. Ashizawa T, editor. PLoS ONE. 2012 Apr 17;7(4):e35241. 8. Jeong SY Rathore KI Schulz K Ponka P Arosio P David S Dysregulation of Iron Homeostasis in the CNS Contributes to Disease Progression in a Mouse Model of Amyotrophic Lateral Sclerosis J Neurosci 2009 29 3 610 619 10.1523/JNEUROSCI.5443-08.2009 19158288 9. Kollewe K Mauss U Krampfl K Petri S Dengler R Mohammadi B ALSFRS-R score and its ratio: A useful predictor for ALS-progression J Neurol Sci 2008 275 1–2 69 73 10.1016/j.jns.2008.07.016 18721928 10. Rudnicki SA Berry JD Ingersoll E Archibald D Cudkowicz ME Kerr DA Dexpramipexole effects on functional decline and survival in subjects with amyotrophic lateral sclerosis in a Phase II study: Subgroup analysis of demographic and clinical characteristics Amyotroph Lateral Scler Front Degener 2013 14 1 44 51 10.3109/17482968.2012.723723 11. Braak H Brettschneider J Ludolph AC Lee VM Trojanowski JQ Tredici KD Amyotrophic lateral sclerosis—a model of corticofugal axonal spread Nat Rev Neurol 2013 9 12 708 714 10.1038/nrneurol.2013.221 24217521 12. The ALS Schwaben Register Group Rosenbohm A Kassubek J Weydt P Marroquin N Volk AE Can lesions to the motor cortex induce amyotrophic lateral sclerosis? J Neurol 2014 261 2 283 290 10.1007/s00415-013-7185-7 24253481 13. Ravits JM La Spada AR ALS motor phenotype heterogeneity, focality, and spread Deconstructing motor neuron degeneration Neurology 2009 73 10 805 811 10.1212/WNL.0b013e3181b6bbbd 19738176 14. Stroman PW Wheeler-Kingshott C Bacon M Schwab JM Bosma R Brooks J The current state-of-the-art of spinal cord imaging: Methods Neuroimage 2014 84 1070 1081 10.1016/j.neuroimage.2013.04.124 23685159 15. Verma T Cohen-Adad J Effect of respiration on the B 0 field in the human spinal cord at 3T: Effect of Respiration on B 0 Field in Human Spinal Cord Magn Reson Med 2014 72 6 1629 1636 10.1002/mrm.25075 24390654 16. Bandettini PA What’s New in Neuroimaging Methods? Ann N Y Acad Sci 2009 1156 1 260 293 10.1111/j.1749-6632.2009.04420.x 19338512 17. Ashburner J Friston KJ Voxel-based morphometry--the methods Neuroimage 2000 11 6 Pt 1 805 821 10.1006/nimg.2000.0582 10860804 18. Grosskreutz J Kaufmann J Frädrich J Dengler R Heinze H-J Peschel T Widespread sensorimotor and frontal cortical atrophy in amyotrophic lateral sclerosis BMC Neurol 2006 6 1 17 10.1186/1471-2377-6-17 16638121 19. Mezzapesa DM Ceccarelli A Dicuonzo F Carella A De Caro MF Lopez M Whole-brain and regional brain atrophy in amyotrophic lateral sclerosis AJNR Am J Neuroradiol 2007 28 2 255 259 17296989 20. Mezzapesa DM D’Errico E Tortelli R Distaso E Cortese R Tursi M Cortical thinning and clinical heterogeneity in amyotrophic lateral sclerosis PLoS One 2013 8 11 e80748 10.1371/journal.pone.0080748 24278317 21. Agosta F Pagani E Rocca MA Caputo D Perini M Salvi F Voxel-based morphometry study of brain volumetry and diffusivity in amyotrophic lateral sclerosis patients with mild disability Hum Brain Mapp 2007 28 12 1430 1438 10.1002/hbm.20364 17370339 22. Chang JL Lomen-Hoerth C Murphy J Henry RG Kramer JH Miller BL A voxel-based morphometry study of patterns of brain atrophy in ALS and ALS/FTLD Neurology 2005 65 1 75 80 10.1212/01.wnl.0000167602.38643.29 16009889 23. Canu E Agosta F Riva N Sala S Prelle A Caputo D The topography of brain microstructural damage in amyotrophic lateral sclerosis assessed using diffusion tensor MR imaging AJNR Am J Neuroradiol 2011 32 7 1307 1314 10.3174/ajnr.A2469 21680655 24. Thivard L Pradat P-F Lehéricy S Lacomblez L Dormont D Chiras J Diffusion tensor imaging and voxel based morphometry study in amyotrophic lateral sclerosis: relationships with motor disability J Neurol Neurosurg Psychiatry 2007 78 8 889 892 10.1136/jnnp.2006.101758 17635981 25. Schuster C Kasper E Machts J Bittner D Kaufmann J Benecke R Focal thinning of the motor cortex mirrors clinical features of amyotrophic lateral sclerosis and their phenotypes: a neuroimaging study J Neurol 2013 260 11 2856 2864 10.1007/s00415-013-7083-z 23996101 26. Ellis CM Suckling J Amaro E Bullmore ET Simmons A Williams SC Volumetric analysis reveals corticospinal tract degeneration and extramotor involvement in ALS Neurology 2001 57 9 1571 1578 10.1212/WNL.57.9.1571 11706094 27. Devine MS Pannek K Coulthard A Mccombe PA Rose SE Henderson RD Exposing asymmetric gray matter vulnerability in amyotrophic lateral sclerosis NeuroImage Clin 2015 7 782 787 10.1016/j.nicl.2015.03.006 25844330 28. Zhang Q Mao C Jin J Niu C Bai L Dang J Side of Limb-Onset Predicts Laterality of Gray Matter Loss in Amyotrophic Lateral Sclerosis BioMed Res Int 2014 2014 1 11 29. Walhout R Westeneng H-J Verstraete E Hendrikse J Veldink JH van den Heuvel MP Cortical thickness in ALS: towards a marker for upper motor neuron involvement J Neurol Neurosurg Psychiatry 2015 86 3 288 294 10.1136/jnnp-2013-306839 25121571 30. Verstraete E, van den Heuvel MP, Veldink JH, Blanken N, Mandl RC, Hulshoff Pol HE, et al. Motor Network Degeneration in Amyotrophic Lateral Sclerosis: A Structural and Functional Connectivity Study. Zhan W, editor. PLoS ONE. 2010 Oct 27;5(10):e13664. 31. Zhu W, Fu X, Cui F, Yang F, Ren Y, Zhang X, et al. ALFF Value in Right Parahippocampal Gyrus Acts as a Potential Marker Monitoring Amyotrophic Lateral Sclerosis Progression: a Neuropsychological, Voxel-Based Morphometry, and Resting-State Functional MRI Study. J Mol Neurosci [Internet]. 2015 Jun 4 [cited 2015 Jul 29]; Available from: http://link.springer.com/10.1007/s12031-015-0583-9 32. Kassubek J Unrath A Huppertz H Lulé D Ethofer T Sperfeld A Global brain atrophy and corticospinal tract alterations in ALS, as investigated by voxel‐based morphometry of 3‐D MRI Amyotroph Lateral Scler 2005 6 4 213 220 10.1080/14660820510038538 33. Sach M Winkler G Glauche V Liepert J Heimbach B Koch MA Diffusion tensor MRI of early upper motor neuron involvement in amyotrophic lateral sclerosis Brain J Neurol 2004 127 Pt 2 340 350 10.1093/brain/awh041 34. Sage CA Peeters RR Görner A Robberecht W Sunaert S Quantitative diffusion tensor imaging in amyotrophic lateral sclerosis Neuroimage 2007 34 2 486 499 10.1016/j.neuroimage.2006.09.025 17097892 35. Agosta F Valsasina P Riva N Copetti M Messina MJ Prelle A The cortical signature of amyotrophic lateral sclerosis PLoS One 2012 7 8 e42816 10.1371/journal.pone.0042816 22880116 36. Cerami C Dodich A Canessa N Crespi C Iannaccone S Corbo M Emotional empathy in amyotrophic lateral sclerosis: a behavioural and voxel-based morphometry study Amyotroph Lateral Scler Front Degener 2014 15 1–2 21 29 10.3109/21678421.2013.785568 37. Bede P Bokde ALW Byrne S Elamin M McLaughlin RL Kenna K Multiparametric MRI study of ALS stratified for the C9orf72 genotype Neurology 2013 81 4 361 369 10.1212/WNL.0b013e31829c5eee 23771489 38. Westeneng H-J, Verstraete E, Walhout R, Schmidt R, Hendrikse J, Veldink JH, et al. Subcortical structures in amyotrophic lateral sclerosis. Neurobiol Aging [Internet]. 2014 Sep [cited 2014 Nov 6]; Available from: http://linkinghub.elsevier.com/retrieve/pii/S0197458014005971 39. Abdulla S Machts J Kaufmann J Patrick K Kollewe K Dengler R Hippocampal degeneration in patients with amyotrophic lateral sclerosis Neurobiol Aging 2014 35 11 2639 2645 10.1016/j.neurobiolaging.2014.05.035 25004891 40. Chapman MC Jelsone-Swain L Fling BW Johnson TD Gruis K Welsh RC Corpus callosum area in amyotrophic lateral sclerosis Amyotroph Lateral Scler 2012 13 6 589 591 10.3109/17482968.2012.708935 22873559 41. Meadowcroft MD Mutic NJ Bigler DC Wang J Simmons Z Connor JR Histological-MRI correlation in the primary motor cortex of patients with amyotrophic lateral sclerosis: MRI and Histological Analysis of the PMC in ALS J Magn Reson Imaging 2015 41 3 665 675 10.1002/jmri.24582 24615949 42. Le Bihan D Mangin JF Poupon C Clark CA Pappata S Molko N Diffusion tensor imaging: concepts and applications J Magn Reson Imaging JMRI 2001 13 4 534 546 10.1002/jmri.1076 11276097 43. Habas C Basic principles of diffusion tensor MR tractography J Radiol 2004 85 3 281 286 10.1016/S0221-0363(04)97579-8 15192519 44. Yamada K Sakai K Akazawa K Yuen S Nishimura T MR tractography: a review of its clinical applications Magn Reson Med Sci MRMS Off J Jpn Soc Magn Reson Med 2009 8 4 165 174 45. Ciccarelli O Behrens TE Johansen-Berg H Talbot K Orrell RW Howard RS Investigation of white matter pathology in ALS and PLS using tract-based spatial statistics Hum Brain Mapp 2009 30 2 615 624 10.1002/hbm.20527 18172851 46. Nickerson JP Koski CJ Boyer AC Burbank HN Tandan R Filippi CG Linear longitudinal decline in fractional anisotropy in patients with amyotrophic lateral sclerosis: preliminary results Klin Neuroradiol 2009 19 2 129 134 10.1007/s00062-009-8040-1 19636503 47. Stagg CJ Knight S Talbot K Jenkinson M Maudsley AA Turner MR Whole-brain magnetic resonance spectroscopic imaging measures are related to disability in ALS Neurology 2013 80 7 610 615 10.1212/WNL.0b013e318281ccec 23325907 48. Yin H Cheng SHT Zhang J Ma L Gao Y Li D Corticospinal tract degeneration in amyotrophic lateral sclerosis: a diffusion tensor imaging and fibre tractography study Ann Acad Med Singapore 2008 37 5 411 415 18536829 49. Iwata NK Aoki S Okabe S Arai N Terao Y Kwak S Evaluation of corticospinal tracts in ALS with diffusion tensor MRI and brainstem stimulation Neurology 2008 70 7 528 532 10.1212/01.wnl.0000299186.72374.19 18268244 50. Iwata NK Kwan JY Danielian LE Butman JA Tovar-Moll F Bayat E White matter alterations differ in primary lateral sclerosis and amyotrophic lateral sclerosis Brain 2011 134 9 2642 2655 10.1093/brain/awr178 21798965 51. Ellis CM Simmons A Jones DK Bland J Dawson JM Horsfield MA Diffusion tensor MRI assesses corticospinal tract damage in ALS Neurology 1999 53 5 1051 1058 10.1212/WNL.53.5.1051 10496265 52. Sage CA Van Hecke W Peeters R Sijbers J Robberecht W Parizel P Quantitative diffusion tensor imaging in amyotrophic lateral sclerosis: revisited Hum Brain Mapp 2009 30 11 3657 3675 10.1002/hbm.20794 19404990 53. Sarica A Cerasa A Vasta R Perrotta P Valentino P Mangone G Tractography in amyotrophic lateral sclerosis using a novel probabilistic tool: a study with tract-based reconstruction compared to voxel-based approach J Neurosci Methods 2014 224 79 87 10.1016/j.jneumeth.2013.12.014 24406465 54. Schimrigk SK Bellenberg B Schlüter M Stieltjes B Drescher R Rexilius J Diffusion tensor imaging-based fractional anisotropy quantification in the corticospinal tract of patients with amyotrophic lateral sclerosis using a probabilistic mixture model AJNR Am J Neuroradiol 2007 28 4 724 730 17416829 55. Foerster BR Carlos RC Dwamena BA Callaghan BC Petrou M Edden RAE Multimodal MRI as a diagnostic biomarker for amyotrophic lateral sclerosis Ann Clin Transl Neurol 2014 1 2 107 114 10.1002/acn3.30 25356389 56. Filippini N Douaud G MacKay CE Knight S Talbot K Turner MR Corpus callosum involvement is a consistent feature of amyotrophic lateral sclerosis Neurology 2010 75 18 1645 1652 10.1212/WNL.0b013e3181fb84d1 21041787 57. Rosskopf J Müller H-P Dreyhaupt J Gorges M Ludolph AC Kassubek J Ex post facto assessment of diffusion tensor imaging metrics from different MRI protocols: Preparing for multicentre studies in ALS Amyotroph Lateral Scler Front Degener 2015 9 1 10 58. Tang M Chen X Zhou Q Liu B Liu Y Liu S Quantitative assessment of amyotrophic lateral sclerosis with diffusion tensor imaging in 3.0 T magnetic resonance Int J Clin Exp Med 2015 8 5 8295 8303 26221413 59. Pyra T Hui B Hanstock C Concha L Wong JCT Beaulieu C Combined structural and neurochemical evaluation of the corticospinal tract in amyotrophic lateral sclerosis Amyotroph Lateral Scler 2010 11 1–2 157 165 10.3109/17482960902756473 19242831 60. Prell T Peschel T Hartung V Kaufmann J Klauschies R Bodammer N Diffusion tensor imaging patterns differ in bulbar and limb onset amyotrophic lateral sclerosis Clin Neurol Neurosurg 2013 115 8 1281 1287 10.1016/j.clineuro.2012.11.031 23266262 61. Keil C Prell T Peschel T Hartung V Dengler R Grosskreutz J Longitudinal diffusion tensor imaging in amyotrophic lateral sclerosis BMC Neurosci 2012 13 1 141 10.1186/1471-2202-13-141 23134591 62. Zhang Y Schuff N Woolley SC Chiang GC Boreta L Laxamana J Progression of white matter degeneration in amyotrophic lateral sclerosis: A diffusion tensor imaging study Amyotroph Lateral Scler 2011 12 6 421 429 10.3109/17482968.2011.593036 21745124 63. Rajagopalan V Allexandre D Yue GH Pioro EP Diffusion Tensor Imaging Evaluation of Corticospinal Tract Hyperintensity in Upper Motor Neuron-Predominant ALS Patients J Aging Res 2011 2011 481745 10.4061/2011/481745 22132329 64. Abe O Yamada H Masutani Y Aoki S Kunimatsu A Yamasue H Amyotrophic lateral sclerosis: diffusion tensor tractography and voxel-based analysis NMR Biomed 2004 17 6 411 416 10.1002/nbm.907 15386625 65. Poujois A Schneider FC Faillenot I Camdessanché J-P Vandenberghe N Thomas-Antérion C Brain plasticity in the motor network is correlated with disease progression in amyotrophic lateral sclerosis: Functional Cerebral Reorganization, Disease Progression, and ALS Hum Brain Mapp 2013 34 10 2391 2401 10.1002/hbm.22070 22461315 66. Müller H-P Lulé D Unrath A Ludolph AC Riecker A Kassubek J Complementary image analysis of diffusion tensor imaging and 3-dimensional t1-weighted imaging: white matter analysis in amyotrophic lateral sclerosis J Neuroimaging Off J Am Soc Neuroimaging 2011 21 1 24 33 10.1111/j.1552-6569.2009.00447.x 67. Chapman MC Jelsone-Swain L Johnson TD Gruis KL Welsh RC Diffusion tensor MRI of the corpus callosum in amyotrophic lateral sclerosis: DTI of the Corpus Callosum in ALS J Magn Reson Imaging 2014 39 3 641 647 10.1002/jmri.24218 23843179 68. Keller J Vymazal J Ridzoň P Rusina R Kulišt’ák P Malíková H Quantitative brain MR imaging in amyotrophic lateral sclerosis Magma N Y N 2011 24 2 67 76 10.1007/s10334-010-0237-4 69. Metwalli NS Benatar M Nair G Usher S Hu X Carew JD Utility of axial and radial diffusivity from diffusion tensor MRI as markers of neurodegeneration in amyotrophic lateral sclerosis Brain Res 2010 1348 156 164 10.1016/j.brainres.2010.05.067 20513367 70. Prudlo J Bißbort C Glass A Grossmann A Hauenstein K Benecke R White matter pathology in ALS and lower motor neuron ALS variants: a diffusion tensor imaging study using tract-based spatial statistics J Neurol 2012 259 9 1848 1859 10.1007/s00415-012-6420-y 22349938 71. Wang S Poptani H Bilello M Wu X Woo JH Elman LB Diffusion tensor imaging in amyotrophic lateral sclerosis: volumetric analysis of the corticospinal tract Am J Neuroradiol 2006 27 6 1234 1238 16775271 72. Brettschneider J Del Tredici K Toledo JB Robinson JL Irwin DJ Grossman M Stages of pTDP-43 pathology in amyotrophic lateral sclerosis: ALS Stages Ann Neurol 2013 74 1 20 38 10.1002/ana.23937 23686809 73. Kassubek J Muller H-P Del Tredici K Brettschneider J Pinkhardt EH Lule D Diffusion tensor imaging analysis of sequential spreading of disease in amyotrophic lateral sclerosis confirms patterns of TDP-43 pathology Brain 2014 137 6 1733 1740 10.1093/brain/awu090 24736303 74. Ding X-Q Kollewe K Blum K Körner S Kehbel S Dengler R Value of quantitative analysis of routine clinical MRI sequences in ALS Amyotroph Lateral Scler 2011 12 6 406 413 10.3109/17482968.2011.597402 21812629 75. Bede P Elamin M Byrne S McLaughlin RL Kenna K Vajda A Patterns of cerebral and cerebellar white matter degeneration in ALS J Neurol Neurosurg Psychiatry 2015 86 4 468 470 10.1136/jnnp-2014-308172 25053771 76. Jacob S Finsterbusch J Weishaupt JH Khorram-Sefat D Frahm J Ehrenreich H Diffusion tensor imaging for long-term follow-up of corticospinal tract degeneration in amyotrophic lateral sclerosis Neuroradiology 2003 45 9 598 600 10.1007/s00234-003-1014-0 12904924 77. Menke RAL Abraham I Thiel CS Filippini N Knight S Talbot K Fractional Anisotropy in the Posterior Limb of the Internal Capsule and Prognosis in Amyotrophic Lateral Sclerosis Arch Neurol 2012 69 11 1493 10.1001/archneurol.2012.1122 22910997 78. Menke RAL Korner S Filippini N Douaud G Knight S Talbot K Widespread grey matter pathology dominates the longitudinal cerebral MRI and clinical landscape of amyotrophic lateral sclerosis Brain 2014 137 9 2546 2555 10.1093/brain/awu162 24951638 79. Lombardo F Frijia F Bongioanni P Canapicchi R Minichilli F Bianchi F Diffusion tensor MRI and MR spectroscopy in long lasting upper motor neuron involvement in amyotrophic lateral sclerosis Arch Ital Biol 2009 147 3 69 82 20014653 80. Cardenas-Blanco A Machts J Acosta-Cabronero J Kaufmann J Abdulla S Kollewe K Central white matter degeneration in bulbar- and limb-onset amyotrophic lateral sclerosis J Neurol 2014 261 10 1961 1967 10.1007/s00415-014-7434-4 25059391 81. Foerster BR Dwamena BA Petrou M Carlos RC Callaghan BC Churchill CL Diagnostic accuracy of diffusion tensor imaging in amyotrophic lateral sclerosis: a systematic review and individual patient data meta-analysis Acad Radiol 2013 20 9 1099 1106 10.1016/j.acra.2013.03.017 23931423 82. Galanaud D Nicoli F Confort-Gouny S Le Fur Y Dormont D Girard N Brain magnetic resonance spectroscopy J Radiol 2007 88 3 Pt 2 483 496 10.1016/S0221-0363(07)89848-9 17457259 83. Jissendi Tchofo P Balériaux D Brain (1)H-MR spectroscopy in clinical neuroimaging at 3T J Neuroradiol 2009 36 1 24 40 10.1016/j.neurad.2008.04.001 18539330 84. Soares DP Law M Magnetic resonance spectroscopy of the brain: review of metabolites and clinical applications Clin Radiol 2009 64 1 12 21 10.1016/j.crad.2008.07.002 19070693 85. Han J Ma L Study of the features of proton MR spectroscopy ((1)H-MRS) on amyotrophic lateral sclerosis J Magn Reson Imaging JMRI 2010 31 2 305 308 10.1002/jmri.22053 20099342 86. Rooney WD Miller RG Gelinas D Schuff N Maudsley AA Weiner MW Decreased N-acetylaspartate in motor cortex and corticospinal tract in ALS Neurology 1998 50 6 1800 1805 10.1212/WNL.50.6.1800 9633731 87. Liu C Jiang R Yi X Zhu W Bu B Role of diffusion tensor imaging or magnetic resonance spectroscopy in the diagnosis and disability assessment of amyotrophic lateral sclerosis J Neurol Sci 2015 348 1–2 206 210 10.1016/j.jns.2014.12.004 25524526 88. Kalra S Vitale A Cashman NR Genge A Arnold DL Cerebral degeneration predicts survival in amyotrophic lateral sclerosis J Neurol Neurosurg Psychiatry 2006 77 11 1253 1255 10.1136/jnnp.2006.090696 16835288 89. Kalra S Hanstock CC Martin WW Allen PS Johnston WS Detection of cerebral degeneration in amyotrophic lateral sclerosis using high-field magnetic resonance spectroscopy Arch Neurol 2006 63 8 1144 1148 10.1001/archneur.63.8.1144 16908742 90. Cervo A Cocozza S Saccà F Giorgio SMDA Morra VB Tedeschi E The combined use of conventional MRI and MR spectroscopic imaging increases the diagnostic accuracy in amyotrophic lateral sclerosis Eur J Radiol 2015 84 1 151 157 10.1016/j.ejrad.2014.10.019 25466774 91. Pohl C Block W Träber F Schmidt S Pels H Grothe C Proton magnetic resonance spectroscopy and transcranial magnetic stimulation for the detection of upper motor neuron degeneration in ALS patients J Neurol Sci 2001 190 1–2 21 27 10.1016/S0022-510X(01)00568-8 11574102 92. Govind V Sharma KR Maudsley AA Arheart KL Saigal G Sheriff S Comprehensive evaluation of corticospinal tract metabolites in amyotrophic lateral sclerosis using whole-brain 1H MR spectroscopy PLoS One 2012 7 4 e35607 10.1371/journal.pone.0035607 22539984 93. Verma G Woo JH Chawla S Wang S Sheriff S Elman LB Whole-brain analysis of amyotrophic lateral sclerosis by using echo-planar spectroscopic imaging Radiology 2013 267 3 851 857 10.1148/radiol.13121148 23360740 94. Bowen BC Pattany PM Bradley WG Murdoch JB Rotta F Younis AA MR imaging and localized proton spectroscopy of the precentral gyrus in amyotrophic lateral sclerosis Am J Neuroradiol 2000 21 4 647 658 10782773 95. Foerster BR Callaghan BC Petrou M Edden RAE Chenevert TL Feldman EL Decreased motor cortex γ-aminobutyric acid in amyotrophic lateral sclerosis Neurology 2012 78 20 1596 1600 10.1212/WNL.0b013e3182563b57 22517106 96. Unrath A Ludolph AC Kassubek J Brain metabolites in definite amyotrophic lateral sclerosis: A longitudinal proton magnetic resonance spectroscopy study J Neurol 2007 254 8 1099 1106 10.1007/s00415-006-0495-2 17431700 97. Wang Q Zhang X Chen S Zhang X Zhang S Youdium M Prevention of Motor Neuron Degeneration by Novel Iron Chelators in SOD1(G93A) Transgenic Mice of Amyotrophic Lateral Sclerosis Neurodegener Dis 2011 8 5 310 321 10.1159/000323469 21346313 98. Deistung A Schäfer A Schweser F Biedermann U Turner R Reichenbach JR Toward in vivo histology: A comparison of quantitative susceptibility mapping (QSM) with magnitude-, phase-, and R2*-imaging at ultra-high magnetic field strength Neuroimage 2013 65 299 314 10.1016/j.neuroimage.2012.09.055 23036448 99. Haacke EM Cheng NYC House MJ Liu Q Neelavalli J Ogg RJ Imaging iron stores in the brain using magnetic resonance imaging Magn Reson Imaging 2005 23 1 1 25 10.1016/j.mri.2004.10.001 15733784 100. Adachi Y Sato N Saito Y Kimura Y Nakata Y Ito K Usefulness of SWI for the Detection of Iron in the Motor Cortex in Amyotrophic Lateral Sclerosis: Usefulness of SWI for the Diagnoses of ALS J Neuroimaging 2015 25 3 443 451 10.1111/jon.12127 24888543 101. Chavhan GB Babyn PS Thomas B Shroff MM Haacke EM Principles, Techniques, and Applications of T2*-based MR Imaging and Its Special Applications1 RadioGraphics 2009 29 5 1433 1449 10.1148/rg.295095034 19755604 102. Haacke EM Miao Y Liu M Habib CA Katkuri Y Liu T Correlation of putative iron content as represented by changes in R2* and phase with age in deep gray matter of healthy adults J Magn Reson Imaging JMRI 2010 32 3 561 576 10.1002/jmri.22293 20815053 103. Langkammer C Krebs N Goessler W Scheurer E Ebner F Yen K Quantitative MR imaging of brain iron: a postmortem validation study Radiology 2010 257 2 455 462 10.1148/radiol.10100495 20843991 104. Haacke EM Liu S Buch S Zheng W Wu D Ye Y Quantitative susceptibility mapping: current status and future directions Magn Reson Imaging 2015 33 1 1 25 10.1016/j.mri.2014.09.004 25267705 105. Langkammer C Schweser F Krebs N Deistung A Goessler W Scheurer E Quantitative susceptibility mapping (QSM) as a means to measure brain iron? A post mortem validation study Neuroimage 2012 62 3 1593 1599 10.1016/j.neuroimage.2012.05.049 22634862 106. Oba H Araki T Ohtomo K Monzawa S Uchiyama G Koizumi K Amyotrophic lateral sclerosis: T2 shortening in motor cortex at MR imaging Radiology 1993 189 3 843 846 10.1148/radiology.189.3.8234713 8234713 107. Hecht MJ Fellner C Schmid A Neundörfer B Fellner FA Cortical T2 signal shortening in amyotrophic lateral sclerosis is not due to iron deposits Neuroradiology 2005 47 11 805 808 10.1007/s00234-005-1421-5 16175348 108. Ignjatović A Stević Z Lavrnić S Daković M Bačić G Brain iron MRI: A biomarker for amyotrophic lateral sclerosis: Brain Iron MRI: A Biomarker for ALS J Magn Reson Imaging 2013 38 6 1472 1479 10.1002/jmri.24121 23564606 109. Yu J Qi F Wang N Gao P Dai S Lu Y Increased iron level in motor cortex of amyotrophic lateral sclerosis patients: An in vivo MR study Amyotroph Lateral Scler Front Degener 2014 15 5–6 357 361 10.3109/21678421.2014.906618 110. Langkammer C Enzinger C Quasthoff S Grafenauer P Soellinger M Fazekas F Mapping of iron deposition in conjunction with assessment of nerve fiber tract integrity in amyotrophic lateral sclerosis J Magn Reson Imaging JMRI 2010 31 6 1339 1345 10.1002/jmri.22185 20512885 111. Prell T, Hartung V, Tietz F, Penzlin S, Ilse B, Schweser F, et al. Susceptibility-Weighted Imaging Provides Insight into White Matter Damage in Amyotrophic Lateral Sclerosis. Kassubek J, editor. PLOS ONE. 2015 Jun 25;10(6):e0131114. 112. Schweitzer AD Liu T Gupta A Zheng K Seedial S Shtilbans A Quantitative Susceptibility Mapping of the Motor Cortex in Amyotrophic Lateral Sclerosis and Primary Lateral Sclerosis Am J Roentgenol 2015 204 5 1086 1092 10.2214/AJR.14.13459 25905946 113. Branco LMT De Albuquerque M De Andrade HMT Bergo FPG Nucci A França MC Spinal cord atrophy correlates with disease duration and severity in amyotrophic lateral sclerosis Amyotroph Lateral Scler Front Degener 2014 15 1–2 93 97 10.3109/21678421.2013.852589 114. Cohen-Adad J Mendili M-ME Morizot-Koutlidis R Lehéricy S Meininger V Blancho S Involvement of spinal sensory pathway in ALS and specificity of cord atrophy to lower motor neuron degeneration Amyotroph Lateral Scler Front Degener 2013 14 1 30 38 10.3109/17482968.2012.701308 115. El Mendili M-M Cohen-Adad J Pelegrini-Issac M Rossignol S Morizot-Koutlidis R Marchand-Pauvert V Multi-parametric spinal cord MRI as potential progression marker in amyotrophic lateral sclerosis PLoS One 2014 9 4 e95516 10.1371/journal.pone.0095516 24755826 116. Iglesias C Sangari S El Mendili M-M Benali H Marchand-Pauvert V Pradat P-F Electrophysiological and spinal imaging evidences for sensory dysfunction in amyotrophic lateral sclerosis BMJ Open 2015 5 2 e007659 10.1136/bmjopen-2015-007659 25712823 117. Ikeda K Murata K Kawase Y Kawabe K Kano O Yoshii Y Relationship between cervical cord 1H-magnetic resonance spectroscopy and clinoco-electromyographic profile in amyotrophic lateral sclerosis Muscle Nerve 2013 47 1 61 67 10.1002/mus.23467 23042532 118. Barkhof F Haller S Rombouts SARB Resting-state functional MR imaging: a new window to the brain Radiology 2014 272 1 29 49 10.1148/radiol.14132388 24956047 119. Habas C Physiological basis of functional MRI J Radiol 2002 83 11 1737 1741 12469010 120. Delmaire C Krainik A Lethuc V Reyns N Duffau H Capelle L Functional magnetic resonance imaging: physiopathology, techniques and applications J Radiol 2007 88 3 Pt 2 497 509 10.1016/S0221-0363(07)89849-0 17457260 121. Gore JC Principles and practice of functional MRI of the human brain J Clin Invest 2003 112 1 4 9 10.1172/JCI200319010 12840051 122. Raichle ME The Restless Brain Brain Connect 2011 1 1 3 12 10.1089/brain.2011.0019 22432951 123. Agosta F Canu E Valsasina P Riva N Prelle A Comi G Divergent brain network connectivity in amyotrophic lateral sclerosis Neurobiol Aging 2013 34 2 419 427 10.1016/j.neurobiolaging.2012.04.015 22608240 124. Jelsone-Swain LM, Fling BW, Seidler RD, Hovatter R, Gruis K, Welsh RC. Reduced Interhemispheric Functional Connectivity in the Motor Cortex during Rest in Limb-Onset Amyotrophic Lateral Sclerosis. Front Syst Neurosci [Internet]. 2010 [cited 2015 Jul 15];4. Available from: http://journal.frontiersin.org/article/10.3389/fnsys.2010.00158/abstract 125. Agosta F Valsasina P Absinta M Riva N Sala S Prelle A Sensorimotor Functional Connectivity Changes in Amyotrophic Lateral Sclerosis Cereb Cortex 2011 21 10 2291 2298 10.1093/cercor/bhr002 21368084 126. Mohammadi B Kollewe K Samii A Krampfl K Dengler R Münte TF Changes of resting state brain networks in amyotrophic lateral sclerosis Exp Neurol 2009 217 1 147 153 10.1016/j.expneurol.2009.01.025 19416664 127. Douaud G Filippini N Knight S Talbot K Turner MR Integration of structural and functional magnetic resonance imaging in amyotrophic lateral sclerosis Brain 2011 134 12 3470 3479 10.1093/brain/awr279 22075069 128. Luo C, Chen Q, Huang R, Chen X, Chen K, Huang X, et al. Patterns of Spontaneous Brain Activity in Amyotrophic Lateral Sclerosis: A Resting-State fMRI Study. Duce JA, editor. PLoS ONE. 2012 Sep 20;7(9):e45470. 129. Mohammadi B Kollewe K Samii A Dengler R Münte TF Functional neuroimaging at different disease stages reveals distinct phases of neuroplastic changes in amyotrophic lateral sclerosis Hum Brain Mapp 2011 32 5 750 758 10.1002/hbm.21064 20836159 130. Mohammadi B Kollewe K Cole DM Fellbrich A Heldmann M Samii A Amyotrophic lateral sclerosis affects cortical and subcortical activity underlying motor inhibition and action monitoring: ALS and Motor Inhibition Hum Brain Mapp 2015 36 8 2878 2889 10.1002/hbm.22814 25913637 131. Kollewe K Münte TF Samii A Dengler R Petri S Mohammadi B Patterns of cortical activity differ in ALS patients with limb and/or bulbar involvement depending on motor tasks J Neurol 2011 258 5 804 810 10.1007/s00415-010-5842-7 21128080 132. Konrad C Henningsen H Bremer J Mock B Deppe M Buchinger C Pattern of cortical reorganization in amyotrophic lateral sclerosis: a functional magnetic resonance imaging study Exp Brain Res 2002 143 1 51 56 10.1007/s00221-001-0981-9 11907690 133. Schoenfeld MA Tempelmann C Gaul C Kühnel GR Düzel E Hopf J-M Functional motor compensation in amyotrophic lateral sclerosis J Neurol 2005 252 8 944 952 10.1007/s00415-005-0787-y 15750701 134. Lule D Diekmann V Muller H-P Kassubek J Ludolph AC Birbaumer N Neuroimaging of multimodal sensory stimulation in amyotrophic lateral sclerosis J Neurol Neurosurg Psychiatry 2010 81 8 899 906 10.1136/jnnp.2009.192260 20543183 135. Stoppel CM Vielhaber S Eckart C Machts J Kaufmann J Heinze H-J Structural and functional hallmarks of amyotrophic lateral sclerosis progression in motor- and memory-related brain regions NeuroImage Clin 2014 5 277 290 10.1016/j.nicl.2014.07.007 25161894 136. Jelsone-Swain L, Persad C, Burkard D, Welsh RC. Action Processing and Mirror Neuron Function in Patients with Amyotrophic Lateral Sclerosis: An fMRI Study. Iacoboni M, editor. PLOS ONE. 2015 Apr 17;10(4):e0119862. 137. Bede P Hardiman O Lessons of ALS imaging: Pitfalls and future directions - A critical review NeuroImage Clin 2014 4 436 443 10.1016/j.nicl.2014.02.011 24624329 138. Brooks BR Miller RG Swash M Munsat TL World Federation of Neurology Research Group on Motor Neuron Diseases El Escorial revisited: revised criteria for the diagnosis of amyotrophic lateral sclerosis Amyotroph Lateral Scler Mot Neuron Disord Off Publ World Fed Neurol Res Group Mot Neuron Dis 2000 1 5 293 299 139. Andersen PM Borasio GD Dengler R Hardiman O Kollewe K Leigh PN EFNS task force on management of amyotrophic lateral sclerosis: guidelines for diagnosing and clinical care of patients and relatives Eur J Neurol 2005 12 12 921 938 10.1111/j.1468-1331.2005.01351.x 16324086 140. Turner MR Kiernan MC Leigh PN Talbot K Biomarkers in amyotrophic lateral sclerosis Lancet Neurol 2009 8 1 94 109 10.1016/S1474-4422(08)70293-X 19081518 141. Pradat P-F, Bruneteau G. [Clinical characteristics of amyotrophic lateral sclerosis subsets]. Rev Neurol (Paris). 2006 Jun;162 Spec No 2:4S29-24S33. 142. Zoccolella S Beghi E Palagano G Fraddosio A Samarelli V Lamberti P Predictors of delay in the diagnosis and clinical trial entry of amyotrophic lateral sclerosis patients: a population-based study J Neurol Sci 2006 250 1–2 45 49 10.1016/j.jns.2006.06.027 16920152 143. Pradat P-F El Mendili M-M Neuroimaging to investigate multisystem involvement and provide biomarkers in amyotrophic lateral sclerosis BioMed Res Int 2014 2014 467560 10.1155/2014/467560 24949452 144. Turner MR Grosskreutz J Kassubek J Abrahams S Agosta F Benatar M Towards a neuroimaging biomarker for amyotrophic lateral sclerosis Lancet Neurol 2011 10 5 400 403 10.1016/S1474-4422(11)70049-7 21511189 145. Filippi M Agosta F Grosskreutz J Benatar M Kassubek J Verstraete E Progress towards a neuroimaging biomarker for amyotrophic lateral sclerosis Lancet Neurol 2015 14 8 786 788 10.1016/S1474-4422(15)00134-9 26194923 146. Müller H-P Turner MR Grosskreutz J Abrahams S Bede P Govind V A large-scale multicentre cerebral diffusion tensor imaging study in amyotrophic lateral sclerosis J Neurol Neurosurg Psychiatry 2016 87 6 570 579 10.1136/jnnp-2015-311952 26746186

PMC005xxxxxx/PMC5002332.txt

==== Front Clin EpigeneticsClin EpigeneticsClinical Epigenetics1868-70751868-7083BioMed Central London 25310.1186/s13148-016-0253-yResearchNext-generation sequencing reveals broad down-regulation of microRNAs in secondary progressive multiple sclerosis CD4+ T cells Sanders Katherine A. 123Benton Miles C. 4Lea Rod A. 42Maltby Vicki E. 23Agland Susan 5Griffin Nathan 23Scott Rodney J. 236Tajouri Lotti 1Lechner-Scott Jeannette Jeannette.lechner-scott@hnehealth.nsw.gov.au 2571 Faculty of Health Sciences and Medicine, Bond University, Robina, Queensland 4226 Australia 2 Centre for Information-Based Medicine, Hunter Medical Research Institute, Newcastle, New South Wales 2305 Australia 3 School of Biomedical Sciences and Pharmacy, University of Newcastle, Newcastle, New South Wales 2308 Australia 4 Institute of Health and Biomedical Innovation, Genomics Research Centre, Brisbane, Queensland 4059 Australia 5 Department of Neurology, Division of Medicine, John Hunter Hospital, Locked Bag 1, Hunter Region Mail Centre, Newcastle, NSW 2310 Australia 6 Division of Molecular Genetics, Pathology North, Newcastle, New South Wales 2305 Australia 7 School of Medicine and Public Health, University of Newcastle, Newcastle, New South Wales 2308 Australia 27 8 2016 27 8 2016 2016 8 1 8729 6 2016 9 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Immunoactivation is less evident in secondary progressive MS (SPMS) compared to relapsing-remitting disease. MicroRNA (miRNA) expression is integral to the regulation of gene expression; determining their impact on immune-related cell functions, especially CD4+ T cells, during disease progression will advance our understanding of MS pathophysiology. This study aimed to compare miRNA profiles of CD4+ T cells from SPMS patients to healthy controls (HC) using whole miRNA transcriptome next-generation sequencing (NGS). Total RNA was extracted from CD4+ T cells and miRNA expression patterns analyzed using Illumina-based small-RNA NGS in 12 SPMS and 12 HC samples. Results were validated in a further cohort of 12 SPMS and 10 HC by reverse transcription quantitative polymerase chain reaction (RT-qPCR). Results The ten most dysregulated miRNAs identified by NGS were selected for qPCR confirmation; five (miR-21-5p, miR-26b-5p, miR-29b-3p, miR-142-3p, and miR-155-5p) were confirmed to be down-regulated in SPMS (p < 0.05). SOCS6 is targeted by eight of these ten miRNAs. Consistent with this, SOCS6 expression is up-regulated in SPMS CD4+ T cells (p < 0.05). This is of particular interest as SOCS6 has previously been shown to act as a negative regulator of T cell activation. Conclusions Ninety-seven percent of miRNA candidates identified by NGS were down-regulated in SPMS. The down-regulation of miRNAs and increased expression of SOCS6 in SPMS CD4+ T cells may contribute to reduced immune system activity in progressive MS. Electronic supplementary material The online version of this article (doi:10.1186/s13148-016-0253-y) contains supplementary material, which is available to authorized users. Keywords Multiple sclerosisSecondary progressiveMicroRNAsImmunologyCD4+ T cellsNext-generation sequencinghttp://dx.doi.org/http://dx.doi.org/10.13039/501100000924Multiple Sclerosis Research AustraliaTrish Multiple Sclerosis Research FoundationJohn Hunter Hospital Charitable TrustBloomfield Group Foundationhttp://dx.doi.org/http://dx.doi.org/10.13039/501100000024Canadian Institutes of Health Researchissue-copyright-statement© The Author(s) 2016 ==== Body Background Multiple sclerosis (MS) is an autoimmune disease characterized by multifocal inflammatory attacks in the CNS [1]. In the relapsing-remitting (RRMS) stage of the disease, CD4+ T cells are among the primary infiltrators moving from the periphery, through the blood-brain barrier, and into the CNS [2]. These cells then initiate an immune response that results in localized demyelination and corresponding symptoms. The later stage of MS, secondary progressive (SPMS), is characterized by compounding neurodegeneration and increasing disability; however, the relevance of inflammation is unclear [3]. As key regulators of gene expression, microRNAs (miRNAs) may be affecting the immune-related functions of CD4+ T cells in SPMS and may help to elucidate the actions of these cells in SPMS. MiRNAs are short, non-coding RNA molecules (~22 bp) that regulate gene expression at the posttranscriptional stage by targeting the 3′ untranslated region of target genes. Their small size and stable structure make them ideal biomarkers. In recent years, miRNA expression patterns in MS have been the focus of numerous studies, many of which have concentrated on using miRNAs as biomarkers for diagnosis and prognosis [4]. These studies predominantly use easily acquired (and often highly heterogeneous) samples such as whole blood, peripheral blood mononuclear cells (PBMCs), serum, and plasma. Numerous dysregulated miRNAs have been identified, however which cell types are actually responsible for differing miRNA profiles, and the consequences of altered miRNA expression is not clear in many studies. Furthermore, it is likely that these heterogeneous samples are masking the signal of differentially expressed miRNA in specific cell subtypes. To overcome this, we have focused on CD4+ T cells in this study on SPMS. Next-generation sequencing (NGS) allows for stringent examination of cell-specific miRNA expression profiles as well as discovery of previously uncharacterized miRNAs. Here, we have used small-RNA NGS analysis of CD4+ T cells from SPMS patients and healthy controls (HC). The total coverage approach of NGS generates expression information on all small RNA species including all known and novel miRNAs, as well as other small RNA species (isomiRs and snoRNAs)—a clear advantage over microarray and candidate approach assays. Three previous studies in MS have used NGS to effectively identify miRNA expression profiles in the whole blood [5, 6], serum [6], and PBMCs [7] from RRMS patients. However, NGS techniques have not been used for specific cell types or in SPMS samples. The miRNA expression profile of CD4+ T cells, either as instigating molecules or by-products of erroneous molecular mechanisms, will provide insight into the function of these cells in SPMS. Here, we used NGS to provide a comprehensive analysis of the miRNA expression profiles of CD4+ T cells from SPMS patients and healthy controls (HC) and confirmed these results using targeted expression assays. Methods Sample collection Whole blood was collected at a single study center from an initial cohort of 12 SPMS patients and 12 HC and a replication cohort of 12 SPMS and 10 HC. All patients were diagnosed with SPMS according to the McDonald criteria [8] and demonstrated EDSS progression without evidence of relapse in the 24 months prior to collection. Controls were age (±5 years) and gender matched (Table 1). The SPMS patient group was free of MS-specific treatments for a minimum period of 6 months prior to collection. Samples were collected at the John Hunter Hospital, and laboratory work was conducted at the Hunter Medical Research Institute, Newcastle.Table 1 Details of SPMS and healthy control individuals Next generation sequencing Replication cohort SPMS HC SPMS HC Number 12 12 12 10 Female 9 9 8 5 Age in years (mean ± SD) 60.2 ± 8.3 61.3 ± 9.5 61.4.0 ± 6.5 60.1 ± 5.9 EDSS (mean ± SD) 6.9 ± 0.9 NA 5.9 ± 1.0 NA Active SPMS 3 NA 4 NA Disease duration in years (mean ± SD) 25.6 ± 11.1 NA 18.3 ± 6.5 NA Progression duration (mean ± SD) 10.8 ± 8.1 NA 8.9 ± 6.2 NA EDSS expanded disability status scale, SD standard deviation, NA not applicable Blood sample processing PBMCs were isolated from 45 mL of heparinized whole blood by density gradient centrifugation on lymphoprep (Axis-Shield PoC AS, Norway). CD4+ T cells were enriched from the PBMCs using EasySep magnetic negative selection according to the manufacturer’s protocol (StemCell Technologies, Canada). The purity of the CD4+ selection was assessed by flow cytometry using a FITC-conjugated anti-CD4 antibody (anti-human CD4 antibody, clone OTK4, FITC, catalog# 60016FI, StemCell Technologies, Canada) on a BD FACSCanto II flow cytometer and then analyzed using FACSDiva software (BD Biosciences, USA) at the Analytical Biomolecular Research Facility of the University of Newcastle. All samples met a minimum purity threshold of >90 %. RNA isolation Total RNA was isolated from the CD4+ T cells using the miRNeasy Mini kit (Qiagen, USA) following the manufacturer’s instructions. The quality of the RNA was assessed using the RNA 6000 Nano kit on a 2100 Bioanalyzer (Agilent Technologies, USA); a RNA integrity number greater than 8 was deemed suitable for sequencing and reverse transcription quantitative polymerase chain reaction (RT-qPCR). Purity was measured on an Epoch spectrophotometer (BioTek, USA), and concentration was measured using the high-sensitivity RNA kit on Qubit 2.0 Fluorometer (Life Technologies, Thermo Fisher Scientific, USA). miRNA sequencing and analysis A cohort of 12 SPMS and 12 HC samples was run through NGS at the Diamantina Institute, University of Queensland, Brisbane, Australia. Samples were individually barcoded and then sequenced in two multiplexed pools each containing 12 samples. The sequencing libraries were prepared from 1-μg total RNA, using the TruSeq small RNA preparation kit (Illumina, USA) and sequenced using the 50-bp fragment protocol on the HiSeq 2500 platform. The sequencing generated four to nine million reads per sample, more than sufficient for expression and discovery applications. The sample sequencing reads were demultiplexed using the CASAVA 1.8 software package (Illumina, USA). The Illumina adapter sequences were trimmed from the fastq files using Trimmomatic [9]. All reads were aligned and counted against miRBase 21 [10]. RT-qPCR Mature miRNA TaqMan assays (Applied Biosystems, Thermo Fisher Scientific, USA) were used to determine expression of the ten most differentially expressed miRNAs in the initial NGS cohort as well as a replication cohort of 12 SPMS and 10 HC (assay IDs in miRNA numerical order: 000397, 000399, 000407, 000408, 000409, 000413, 002223, 000464, 002623, 000524). The small RNA RNU44 (ref: 001094) was used as an endogenous control. RNU44 has previously been demonstrated to be a stable control in CD4+ T cells [11], and its stability has been shown in our 47 samples (mean ± standard deviation Ct value of 23.58 ± 0.63). RNU44 was used for normalization using the ΔCt method. The relative expression (2−ΔCt) of all samples (24 SPMS and 22 HC) was calculated. Statistical analysis The two-sample Kolmogorov-Smirnov test (K-S test) was used to test whether differences in expression levels were statistically significant between the case and control groups as implemented in R. The K-S test was chosen (over the F test comparison of means) because of the non-normality of the expression level distributions among miRNAs. Our statistical significance threshold allowing for multiple testing correction was determined using the False Discovery Rate (FDR) procedure of Benjamini-Hochberg [12]. Based on the number of miRNA elements, this threshold was set at 1.2 × 10−4. We also considered a relaxed (or nominal) significance threshold of 0.05. In addition to using statistical significance thresholds for miRNA selection, we also included a count threshold of >800 to exclude miRNAs that had very low expression levels and were unlikely to be replicated with the less-sensitive RT-qPCR. The K-S test was also used to determine significant differential miRNA and SOCS6 expression from the RT-qPCR relative expression data. Correlation to patient characteristics The Pearson correlation coefficient was calculated using RT-qPCR data for MS samples (n = 24) and patient characteristics: EDSS, age, disease duration, and progression duration. A correlation coefficient (r value) >±0.5 was considered strong, ±0.3–0.49 moderate, and <±0.29 weak. Gene target prediction miRSystem integrates seven different target gene prediction algorithms and contains experimentally validated data on miRNA:mRNA interactions [13]. This integration system was used to identify genes that may be targeted by more than one of our identified dysregulated miRNAs. Analysis of SOCS6 expression Five hundred nanograms of total RNA was reverse transcribed using high-capacity cDNA reverse transcription kits (Applied Biosystems, Thermo Fisher Scientific, USA) in 21 SPMS and 21 HC samples. qPCR was performed using an exon-spanning TaqMan probe for SOCS6 (ref: Hs00377781_m1). Expression of SOCS6 was determined as relative expression to the housekeeping genes GAPDH (ref: 4326317E) and β-actin (ref: 4326215E) using a ViiA 7 (Applied Biosystems, Thermo Fisher Scientific, USA). Results We used NGS to establish miRNA expression profiles in CD4+ T cells from a cohort of 12 SPMS and 12 HC samples. RT-qPCR was then employed to validate differences in miRNA expression in the NGS cohort as well as a replication cohort of 12 SPMS and 10 HC samples (total 24 SPMS and 22 HC). NGS We observed three statistically significant miRNAs (miR-451a, miR-1246, and miR-144-5p) at the FDR-corrected threshold (Additional file 1: Figure S1), which probably reflects the modest sample size. These miRNAs were very lowly expressed (<100 reads per sample), and we were unable to confirm this dysregulation with RT-qPCR. We also observed 42 miRNAs at the nominal significance threshold (97 % of these were down-regulated). Of these 42 miRNAs, only 10 met our secondary criteria of having a read count >800: miR-21-5p (p = 0.031), miR-23a-3p (p = 0.007), miR-26b-5p (p = 0.031), miR-27a-3p (p = 0.031), miR-27b-3p (p = 0.031), miR-29b-3p (p = 0.007), miR-30e-5p (p = 0.031), miR-142-3p (p = 0.031), miR-155-5p (p = 0.031), and miR-221-3p (p = 0.031). Each of these miRNAs was found to be down-regulated in SPMS as summarized in Fig. 1 and was forwarded for replication testing in an independent cohort.Fig. 1 Tukey boxplot demonstrating the ten most significantly dysregulated microRNAs identified using NGS. Data is presented as log10 of the read count and clearly exhibits the down-regulation of miRNAs in SPMS (purple) compared to HC (gray). Whiskers represent data within 1.5 interquartile range (IQR) of the upper and lower quartile. Data points outside of the 1.5 IQR are represented by black dots. *p < 0.05, **p < 0.01 RT-qPCR To confirm our NGS findings, the top ten most dysregulated miRNAs were selected for further analysis in 24 SPMS and 22 HC samples using RT-qPCR (including the 12 SPMS and 12 HC samples that underwent NGS analysis). Of these ten miRNAs, RT-qPCR confirmed significant down-regulation of miR-21-5p (p = 0.0048), miR-26b-5p (p = 0.007) miR-29b-3p (p = 0.00001), miR-142-3p (p = 0.05), and miR-155-5p (p = 0.001) in SPMS CD4+ T cells (Fig. 2). These five miRNAs were confirmed in the original NGS cohort, the replication cohort, and the combined cohort. This provides statistically significant evidence of replication, indicating that these five miRNAs are very unlikely to be false positives. A trend of down-regulation of miRNA in SPMS samples was still observed across all ten miRNAs.Fig. 2 Tukey boxplot of top ten miRNAs expression (relative to RNU44) using RT-qPCR. Significant down-regulation of miR-21-5p, miR-26b-3p, miR-29b-3p, miR-142-3p, and miR-155-5p in SPMS was confirmed. Whiskers represent data within 1.5 interquartile range (IQR) of the upper and lower quartile. Data points outside of the 1.5 IQR are represented by black dots. p < 0.05, **p < 0.01, ***p < 0.001 Comparison of methods Concordance of differential expression can vary between quantitation methods [14]. To determine the magnitude of fold-change in SPMS vs. HC, we compared RT-qPCR and NGS results and found no change in the degree of decreased expression between NGS and RT-qPCR methods in the miRNAs confirmed by RT-qPCR (Fig. 3).Fig. 3 Comparison of miRNA fold-change between NGS and RT-qPCR. Magnitude of change is consistent between NGS and RT-qPCR methods Correlation to patient characteristics No strong correlations between miRNA expression and patient characteristics were identified (Table 2). However, moderate positive correlation between EDSS and miR-21-5p, miR-26b-5p, and miR-29b-3p was seen. Further positive correlation was also found between disease duration and miR-21-5p and miR-155-5p. All miRNAs demonstrated weak correlation to patient age and progression duration.Table 2 Correlation coefficients calculated from RT-qPCR data against patient characteristics miR-21-5p miR-26b-5p miR-29b-3p miR-142-3p miR-155-5p EDSS 0.34 0.42 0.41 0.28 0.26 Age (HC) 0.22 0.17 0.31 0.21 −0.08 Age (SPMS) −0.07 −0.17 −0.17 −0.30 −0.01 Disease duration 0.49 0.15 0.23 −0.08 0.49 Progression duration 0.12 0.12 0.11 −0.07 0.17 Correlation of miRNA expression and age of HC has also been calculated as a reference point for age of patients. Moderate correlations are in bold text. Target prediction miRNA fold-change was <2 for all miRNAs. It is therefore unlikely that any one particular miRNA is causing a significant effect on gene expression alone. It is more likely to be a combination of multiple miRNAs targeting a few specific genes. Furthermore, as RT-qPCR is a less-sensitive methodology than NGS, and the trend of down-regulation is still observed (though not significant) in the other miRNAs, all ten miRNAs were cross-analyzed for potential gene targets. miRSystem was used to identify genes that have multiple target genes in common, both in the five confirmed miRNAs and all ten miRNAs identified by NGS. One gene, bromodomain and WD repeat domain containing 1 (BRWD1), is targeted by all five confirmed miRNAs. No genes are targeted by all ten miRNAs; however, eight genes are targeted by eight of the miRNAs (Table 3).Table 3 Genes identified by miRSystem targeted by eight of the ten microRNAs miR-21-5p miR-23a-3p miR-26b-5p miR-27a-3p miR-27b-3p miR-29b-3p miR-30e-5p miR-142-3p miR-155-5p miR-221-3p ACVR2B V V V V V V V V ZBTB41 V V V V V V V V BRWD1 V V V V V V V V CAMTA1 V V V V V V V V CFL2 V V V V V V V V SOCS6 V V V V V V V V MIER3 V V V V V V V V KLF12 V V V V V V V V Verified targeting miRNAs are identified with a “V” These genes are involved in transmembrane ligand binding, regulation of actin filaments, or are transcription factors. However, only one gene is specifically linked to immune cell function, SOCS6 (suppressor of cytokine signaling 6). This gene has previously been reported to negatively regulate T cell activation by promoting ubiquitin-dependent proteolysis [15] and was consequently selected for further investigated. SOCS6 expression Gene expression analysis using RT-qPCR was conducted to determine whether SOCS6 is up-regulated in SPMS CD4+ T cells in direct negative correlation to the miRNA expression (Fig. 4). Both the preliminary and validation cohorts were analyzed, and SOCS6 expression is increased in SPMS compared to HC. Normalization against GAPDH and β-actin generated the same results (data for β-actin not shown).Fig. 4 Expression of SOCS6 relative to GAPDH. Up-regulation of SOCS6 in SPMS is significant though widely distributed (*p = 0.042) Discussion This is the first study in MS to utilize NGS for miRNA expression profiling in the CD4+ T cells of SPMS patients. We found 42 miRNAs that are dysregulated in the CD4+ T cells of SPMS patients as compared to controls: 97 % of which were down-regulated. TaqMan assays confirmed five of these miRNAs (miR-21-5p, miR-26b-5p, miR-29b-3p, miR-142-3p, and miR-155-5p) to be down-regulated in SPMS. Each of these miRNAs (excluding miR-26b) has been reported on previously in MS though not necessarily in SPMS or CD4+ T cells. Lindberg et al. [11] identified seven miRNAs dysregulated in CD4+ T cells from RRMS patients but did not identify dysregulation in any of the five miRNA in this study. Thus, down-regulation of these miRNAs may be exclusive to SPMS. Here, we report a decrease in miR-155-5p expression in MS. miR-155-5p has a pro-inflammatory role in MS and is up-regulated in a number of tissues. Studies of postmortem brain tissue find a gradient of miR-155-5p expression that peaks in active lesions [16] and associated neurovascular units [17] and decreases through chronic lesions and normal appearing white matter to a low baseline in healthy control (non-MS) white matter [17, 18]. This increased expression of miR-155 has been associated with the suppression of CD47 in active lesions that creates a permissive environment for myelin phagocytosis [16]; focal adhesion and cell-cell complex molecules in the blood-brain barrier, thus increasing permeability [17] and; AKR1C1 and AKR1C2, essential for biosynthesis of allopregnanolone (a neuroprotective steroid) [18]. Interestingly, a study of miR-155 in the EAE mouse model found that miR-155 expression in CD4+ T cells increases during EAE and that miR-155−/− mice had an attenuation of EAE [19]. Specifically, Th17 cells lacking miR-155-5p are unable to cause EAE [20]. miR-155-5p is required for normal immune function [21], and together, these studies confirm that the significant role miR-155-5p over-expression plays in the inflammatory process of MS. In contrast, our finding of miR-155-5p down-regulation may be exclusive to SPMS patients and/or CD4+ T cells and is consistent with SPMS as a non-inflammatory mediated disease. miR-155-5p and miR-142-3p have been identified as dysregulated in RRMS PBMCs [22], and a recent study on autologous hematopoietic stem cell transplant (AHSCT) also found co-dysregulation of miR-155-5p and miR-142-3p [23]. Contrary to our results, Arruda et al. found these miRNAs to be up-regulated in MS patient CD4+ T cells before treatment (cohort was 75 % SPMS). However, AHSCT is most effective in active MS disease, and six of the 19 SPMS patients enrolled in the Arruda et al. study presented with gadolinium-enhancing lesions in the year approaching the treatment indicating inflammatory activity. Further, the average disease duration in the Arruda et al. study was 8.1 years, as opposed to 25.6 (primary cohort) or 18.6 (replication cohort) years in our study. Our data is corroborated further by NGS expression analysis, which is a more sensitive measure of expression changes. In a study of potential biomarkers in Alzheimer’s disease (AD), miR-26b-5p was shown to be down-regulated in the serum and CSF of AD patients when compared to patients with inflammatory neurological diseases [24], supporting the predominantly neurodegenerative pathology of SPMS. Over-expression of miR-29b insystemic lupus erythematosus (SLE) has been linked to hypomethylation of DNA in CD4+ T cells [25]. While there are currently no studies on DNA methylation in SPMS, it would be interesting to see if the down-regulation of miR-29b that we have identified here in CD4+ T cells is associated with genome-wide hypermethylation in SPMS. Increased miR-21-5p promotes differentiation of Th17 cells in the EAE mouse model, and miR-21-5p knock-out mice are resistant to EAE [26]. Fenoglio et al. found increased miR-21-5p expression in RRMS (active relapse phase) PMBCs compared to controls, though no difference in SPMS. Again, this may be attributed to the relatively small sample size (n = 6) [27]. Also of interest, we previously reported miR-20a-5p down-regulation in the whole blood of all MS subtypes [28]. This miRNA was one of the 42 dysregulated miRNAs identified by NGS and is significantly down-regulated in SPMS compared to HC. However, it narrowly missed the 800 read cut-off for qPCR confirmation. miR-20a-5p is also predicted to target SOCS6. Eight of the top ten dysregulated miRNAs were predicted to target SOCS6 using MirSystem. Consistent with this, increased expression of SOCS6 in the SPMS cohort is in direct negative correlation with the miRNA expression profiles, strongly indicating a mRNA:miRNA relationship. To our knowledge, this is the first study to identify SOCS6 as a gene of interest in MS. It is a highly conserved gene with very low expression levels in healthy thymus and brain tissues and is down-regulated in gastric, colorectal, and pancreatic cancers [29–32]. In colorectal cancer, methylation changes have been ruled out as the mechanism of down-regulation [31]; therefore, down-regulation may be due to altered miRNA expression. MiR-424-5p is responsible for the down-regulation of SOCS6 in pancreatic cancer [32]; however, we found no differences in miR-424-5p expression between SPMS and HC in this study. The function of SOCS6 as a negative regulator of T cell activation [15] and its observed over-expression in SPMS CD4+ T cells supports the notion of reduced immune activity in SPMS. Very little is known about SOCS6, and more studies are required to determine if it may be a novel therapeutic target. This is the first study to use NGS miRNA profiling to assess miRNA expression in the CD4+ T cells of SPMS patients. Future studies should focus on using the same technique in treatment naïve RRMS patients to determine if this is a SPMS exclusive trend and remove the confounding factor of treatment effects. Furthermore, miRNA expression profiles of other cell subtypes should be investigated, as whole blood analysis is likely masking significant changes in individual cell subsets. Ideally, all of our patients would have had inactive SPMS; however, as SPMS is a difficult disease stage to define and collect, we have included some active SPMS patients in this study. In this study, we chose to focus on CD4+ T cells as they are thought to be the main cell infiltrates. Our previous studies also show that CD4+ T cells exhibit significant changes in methylation profiles in RRMS [33, 34]. Conclusions Here, we have shown a general down-regulation of miRNAs in CD4+ T cells compared to HC, with five miRNAs confirmed as significant in two independent assays. This indicates that miRNA expression may be over-normalizing in SPMS CD4+ T cells. SOCS6 is a predicted target of the majority of these miRNAs and, consistent with this, we found SOCS6 to be up-regulated in this cohort. These are novel findings that point towards a diminished role for CD4+ T cells in SPMS and add further evidence for SPMS being a neurodegenerative disease stage, not an inflammation-driven one. Additional file Additional file 1: Figure S1. Volcano plot of differentially expressed miRNAs identified with NGS. The FDR-corrected significance threshold is demarked with a green line at p < 1.2 × 10−4. Three miRNAs were identified at the threshold. Mean read counts were low in all three miRNAs: miR-451a (SPMS mean = 76.3, HC mean = 18.9), miR-1246 (SPMS mean = 94.9, HC mean = 51.9), and miR-144-5p (SPMS mean = 15.1, HC mean = 5.5). Differential expression could not be replicated with RT-qPCR. (PNG 57 kb) Abbreviations ADAlzheimer’s disease AHSCTAutologous hematopoietic stem cell transplant CNSCentral nervous system DNADeoxyribonucleic acid EAEExperimental autoimmune encephalitis EDSSExpanded disability status scale FDRFalse discovery rate GAGlatiramer acetate HCHealthy controls K-S testKolmogorov-Smirnov test miRNAMicroRNA MSMultiple sclerosis NGSNext-generation sequencing PBMCPeripheral blood mononuclear cells RNARibonucleic acid RRMSRelapsing remitting MS RT-qPCRReverse transcription quantitative polymerase chain reaction SDStandard deviation SOCS6Suppressor of cytokine signaling 6 SPMSSecondary progressive MS Acknowledgements We would like to thank the MS patients and clinical team at the John Hunter Hospital MS clinic who participated in this study. We also acknowledge the Analytical Biomolecular Research Facility at the University of Newcastle for the flow cytometry support and the Diamantina Institute at the University of Queensland for the NGS services. Funding This study was supported by the John Hunter Hospital Charitable Trust and the Bloomfield Group Foundation. KAS, VEM, and RAL are supported by fellowships from Multiple Sclerosis Research Australia. KAS is also funded by the Trish MS Research Foundation postgraduate scholarship. VEM is supported by a postdoctoral fellowship from the Canadian Institutes of Health Research. Availability of data and material Data available on request and subject to ethics committee approval. Authors’ contributions KAS, MCB, and RAL wrote the main manuscript text. KAS, VEM, SA, and NG performed the experiments. MCB and RAL analyzed the NGS data. RJS, LT, and JLS supervised the study. All authors reviewed the manuscript. All authors read and approved the final manuscript. Competing interests JL-S receives non-direct funding as well as honoraria for presentations and membership on advisory boards from Genzyme, Biogen, Bayer Health Care, Merck, Teva, and Novartis, Australia. The authors, KAS, MCB, RAL, VEM, SA, NG, RJS, and LT, declare that they have no competing interests. Consent for publication Not applicable. Ethics approval and consent to participate The Hunter New England Health Research Ethics Committee and University of Newcastle Ethics committee approved this study (05/04/13.09 and H-505-0607, respectively), and methods were carried out in accordance with institutional guidelines on human subject experiments. Written and informed consent was obtained from all patient and control subjects. ==== Refs References 1. Compston A Coles A Multiple sclerosis Lancet 2008 372 1502 1517 10.1016/S0140-6736(08)61620-7 18970977 2. Broux B Stinissen P Hellings N Which immune cells matter? The immunopathogenesis of multiple sclerosis Crit Rev Immunol 2013 33 283 306 10.1615/CritRevImmunol.2013007453 23971528 3. Segal BM Stage-specific immune dysregulation in multiple sclerosis J Interferon Cytokine Res 2014 34 633 640 10.1089/jir.2014.0025 25084180 4. Raphael I Webb J Stuve O Haskins W Forsthuber T Body fluid biomarkers in multiple sclerosis: how far we have come and how they could affect the clinic now and in the future Expert Rev Clin Immunol 2015 11 69 91 10.1586/1744666X.2015.991315 25523168 5. Keller A Leidinger P Steinmeyer F Stahler C Franke A Hemmrich-Stanisak G Kappel A Wright I Dorr J Paul F Comprehensive analysis of microRNA profiles in multiple sclerosis including next-generation sequencing Mult Scler 2013 20 295 303 10.1177/1352458513496343 23836875 6. Keller A Leidinger P Meese E Haas J Backes C Rasche L Behrens JR Pfuhl C Wakonig K Giess RM Next-generation sequencing identifies altered whole blood microRNAs in neuromyelitis optica spectrum disorder which may permit discrimination from multiple sclerosis J Neuroinflammation 2015 12 196 10.1186/s12974-015-0418-1 26521232 7. De Felice B Mondola P Sasso A Orefice G Bresciamorra V Vacca G Biffali E Borra M Pannone R Small non-coding RNA signature in multiple sclerosis patients after treatment with interferon-beta BMC Med Genomics 2014 7 26 10.1186/1755-8794-7-26 24885345 8. Polman CH Reingold SC Banwell B Clanet M Cohen JA Filippi M Fujihara K Havrdova E Hutchinson M Kappos L Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria Ann Neurol 2011 69 292 302 10.1002/ana.22366 21387374 9. Bolger AM Lohse M Usadel B Trimmomatic: a flexible trimmer for Illumina sequence data Bioinformatics 2014 30 2114 2120 10.1093/bioinformatics/btu170 24695404 10. Kozomara A Griffiths-Jones S miRBase: annotating high confidence microRNAs using deep sequencing data Nucleic Acids Res 2014 42 D68 73 10.1093/nar/gkt1181 24275495 11. Lindberg RL Hoffmann F Mehling M Kuhle J Kappos L Altered expression of miR-17-5p in CD4+ lymphocytes of relapsing-remitting multiple sclerosis patients Eur J Immunol 2010 40 888 898 10.1002/eji.200940032 20148420 12. Benjamini Y Hochberg Y Controlling the false discovery rate: a practical and powerful approach to multiple testing J R Stat Soc Ser B Methodol 1995 57 289 300 13. Lu T-P Lee C-Y Tsai M-H Chiu Y-C Hsiao CK Lai L-C Chuang EY miRSystem: an integrated system for characterizing enriched functions and pathways of microRNA targets PLoS ONE 2012 7 e42390 10.1371/journal.pone.0042390 22870325 14. Mestdagh P Hartmann N Baeriswyl L Andreasen D Bernard N Chen C Cheo D D'Andrade P DeMayo M Dennis L Evaluation of quantitative miRNA expression platforms in the microRNA quality control (miRQC) study Nat Methods 2014 11 809 815 10.1038/nmeth.3014 24973947 15. Choi YB Son M Park M Shin J Yun Y SOCS-6 negatively regulates T cell activation through targeting p56lck to proteasomal degradation J Biol Chem 2010 285 7271 7280 10.1074/jbc.M109.073726 20007709 16. Junker A Krumbholz M Eisele S Mohan H Augstein F Bittner R Lassmann H Wekerle H Hohlfeld R Meinl E MicroRNA profiling of multiple sclerosis lesions identifies modulators of the regulatory protein CD47 Brain 2009 132 3342 3352 10.1093/brain/awp300 19952055 17. Lopez-Ramirez MA Wu D Pryce G Simpson JE Reijerkerk A King-Robson J Kay O de Vries HE Hirst MC Sharrack B MicroRNA-155 negatively affects blood-brain barrier function during neuroinflammation FASEB J 2014 28 6 2551 65 10.1096/fj.13-248880 24604078 18. Noorbakhsh F Ellestad KK Maingat F Warren KG Han MH Steinman L Baker GB Power C Impaired neurosteroid synthesis in multiple sclerosis Brain 2011 134 2703 2721 10.1093/brain/awr200 21908875 19. Murugaiyan G Beynon V Mittal A Joller N Weiner HL Silencing microRNA-155 ameliorates experimental autoimmune encephalomyelitis J Immunol 2011 187 2213 2221 10.4049/jimmunol.1003952 21788439 20. Hu R Huffaker TB Kagele DA Runtsch MC Bake E Chaudhuri AA Round JL O'Connell RM MicroRNA-155 confers encephalogenic potential to Th17 cells by promoting effector gene expression J Immunol 2013 190 5972 5980 10.4049/jimmunol.1300351 23686497 21. Rodriguez A Vigorito E Clare S Warren MV Couttet P Soond DR van Dongen S Grocock RJ Das PP Miska EA Requirement of bic/microRNA-155 for normal immune function Science 2007 316 608 611 10.1126/science.1139253 17463290 22. Waschbisch A Atiya M Linker RA Potapov S Schwab S Derfuss T Glatiramer acetate treatment normalizes deregulated microRNA expression in relapsing remitting multiple sclerosis PLoS One 2011 6 e24604 10.1371/journal.pone.0024604 21949733 23. Arruda LC Lorenzi JC Sousa AP Zanette DL Palma PV Panepucci RA Brum DS Barreira AA Covas DT Simoes BP Autologous hematopoietic SCT normalizes miR-16, -155 and -142-3p expression in multiple sclerosis patients Bone Marrow Transplant 2015 50 380 389 10.1038/bmt.2014.277 25486582 24. Galimberti D Villa C Fenoglio C Serpente M Ghezzi L Cioffi SM Arighi A Fumagalli G Scarpini E Circulating miRNAs as potential biomarkers in Alzheimer’s disease J Alzheimers Dis 2014 42 1261 1267 25024331 25. Qin H Zhu X Liang J Wu J Yang Y Wang S Shi W Xu J MicroRNA-29b contributes to DNA hypomethylation of CD4+ T cells in systemic lupus erythematosus by indirectly targeting DNA methyltransferase 1 J Dermatol Sci 2013 69 61 67 10.1016/j.jdermsci.2012.10.011 23142053 26. Murugaiyan G da Cunha AP Ajay AK Joller N Garo LP Kumaradevan S Yosef N Vaidya VS Weiner HL MicroRNA-21 promotes Th17 differentiation and mediates experimental autoimmune encephalomyelitis J Clin Invest 2015 125 1069 1080 10.1172/JCI74347 25642768 27. Fenoglio C Cantoni C De Riz M Ridolfi E Cortini F Serpente M Villa C Comi C Monaco F Mellesi L Expression and genetic analysis of miRNAs involved in CD4+ cell activation in patients with multiple sclerosis Neurosci Lett 2011 504 9 12 10.1016/j.neulet.2011.08.021 21875645 28. Cox MB Cairns MJ Gandhi KS Carroll AP Moscovis S Stewart GJ Broadley S Scott RJ Booth DR Lechner-Scott J Consortium ANMSG: MicroRNAs miR-17 and miR-20a inhibit T cell activation genes and are under-expressed in MS whole blood PLoS One 2010 5 e12132 10.1371/journal.pone.0012132 20711463 29. Lai RH Wang MJ Yang SH Chen JY Genomic organization and functional characterization of the promoter for the human suppressor of cytokine signaling 6 gene Gene 2009 448 64 73 10.1016/j.gene.2009.07.022 19716864 30. Lai RH Hsiao YW Wang MJ Lin HY Wu CW Chi CW Li AF Jou YS Chen JY SOCS6, down-regulated in gastric cancer, inhibits cell proliferation and colony formation Cancer Lett 2010 288 75 85 10.1016/j.canlet.2009.06.025 19646809 31. Letellier E Schmitz M Baig K Beaume N Schwartz C Frasquilho S Antunes L Marcon N Nazarov PV Vallar L Identification of SOCS2 and SOCS6 as biomarkers in human colorectal cancer Br J Cancer 2014 111 726 735 10.1038/bjc.2014.377 25025962 32. Wu K Hu G He X Zhou P Li J He B Sun W MicroRNA-424-5p suppresses the expression of SOCS6 in pancreatic cancer Pathol Oncol Res 2013 19 739 748 10.1007/s12253-013-9637-x 23653113 33. Graves M Benton M Lea R Boyle M Tajouri L Macartney-Coxson D Scott R Lechner-Scott J Methylation differences at the HLA-DRB1 locus in CD4+ T-Cells are associated with multiple sclerosis Mult Scler 2014 20 1033 1041 10.1177/1352458513516529 24336351 34. Maltby VE Graves MC Lea RA Benton MC Sanders KA Tajouri L Scott RJ Lechner-Scott J Genome-wide DNA methylation profiling of CD8+ T cells shows a distinct epigenetic signature to CD4+ T cells in multiple sclerosis patients Clin Epigenetics 2015 7 118 10.1186/s13148-015-0152-7 26550040

PMC005xxxxxx/PMC5002333.txt

==== Front J Orthop Surg ResJ Orthop Surg ResJournal of Orthopaedic Surgery and Research1749-799XBioMed Central London 42710.1186/s13018-016-0427-5Research ArticleThe morphological features of different Schatzker types of tibial plateau fractures: a three-dimensional computed tomography study Chen Pengbo 403807657@qq.com Shen Hao 13917481191@163.com Wang Wei drwang7597@aliyun.com Ni Binbin orthoxinhua@163.com Fan Zhiyuan fzhiyuan1993@126.com Lu Hua +862125078999hualushe@sina.com Department of Orthopaedics, Xinhua Hospital Affiliated to Shanghai Jiaotong University School of Medicine, No. 1665 Kongjiang RD, Shanghai, 200092 China 27 8 2016 27 8 2016 2016 11 1 942 6 2016 11 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Tibial plateau fractures are of great challenge to treat with open reduction and internal fixation, because fractures vary from simple to complex, with little or extensive articular involvement. Hence, recognition and comprehension of the fracture features will help orthopedic surgeons understand the injury mechanism better and manage these fractures by planning optimal surgical procedures. This study aimed to evaluate the morphological characteristics of tibial plateau fractures based on the Schatzker classification. Methods A total of 186 patients with 188 tibial plateau fractures from 2010 to 2014 in our hospital were reviewed using a computed tomography scan and three-dimensional (3D) reconstruction. The main fracture line angles (FLA) of Schatzker types I, II, and IV were measured. For each fracture, depression depth was measured, and the depression zone was also located. Depression zones were overlapped to obtain a frequency diagram. Results Schatzker type I and II fractures had three subtypes: single anterolateral fracture, single posterolateral fracture, and complex fracture (the anterolateral and posterolateral parts). Schatzker type IV fractures were also divided into three subtypes: single posteromedial fracture, single anteromedial fracture, and the whole medial fracture. For various Schatzker types and subtypes of fracture, fracture depression clustered and occurred at different locations of the tibial plateau. A significant difference was observed in the depression depth among the different Schatzker types (P < 0.01, Kruskal-Wallis test), especially between Schatzker type III and other types (Nemenyi test). There was no difference in the depression depth among the subtypes of Schatzker type II, whereas the difference was significant between the two subtypes of Schatzker type IV. Conclusions Schatzker type I, II, and IV fractures could be divided into three corresponding subtypes by FLA. Various Schatzker types of fractures differed in location and depth of depression. A proper operative approach should be made based on the morphological characteristics of individual types of tibial plateau fractures. Keywords Tibial plateau fractureSchatzker typeMorphological featureComputed tomographyissue-copyright-statement© The Author(s) 2016 ==== Body Background Tibial plateau fractures, one intra-articular fracture caused by high-velocity trauma, are of great challenge to treat [1, 2]. Treatment with open reduction and internal fixation is difficult [2–4] because fractures vary from simple to complex, with little or extensive articular involvement [5, 6]. Hence, recognition and comprehension of the fracture features will help orthopedic surgeons understand the injury mechanism better and manage these fractures by planning optimal surgical procedures. Now, three classification systems have been commonly used for classifying tibial plateau fractures in clinical practice, bicondylar including Schatzker, AO/OTA, and Three-Column classifications [7–9]. Based on X-ray plain radiographs, Schatzker and AO classifications describe the location and general pattern of fracture without consideration of the fracture line orientation, which usually determines the right position of the bone plate. They were also lacking in adequate details of depression morphological characteristics. However, identification of depression morphological characterization would facilitate surgical plan and therapeutic effect postoperatively [10]. Failure to accomplish the reduction of depression is associated with residual pain, post-traumatic arthritis, and deformity [10–12]. Computed tomography (CT) is nowadays indispensable in the understanding of fracture patterns precisely, especially in consideration of fracture line orientation, location, and magnitude of depression components [13, 14]. It provides surgeons with an opportunity to promote the ability of reduction and internal fixation. Although the Three-Column Classification system locates the fracture based on CT [8], it overlooks the fracture line orientation and morphological characteristics of depression. Hence, this study was to clarify the morphologic characteristics of different Schatzker types [7] of tibial plateau fractures based on CT scanning and three-dimensional (3D) reconstruction, by measuring the fracture line angles (FLA), mapping the most common zones of depression, and measuring the depression depth. Our study would give the surgeons instructions during diagnosis, preoperative planning, and execution of surgical strategies. Methods Patients The retrospective study reviewed a total 186 patients with tibial plateau fractures who were treated in our hospital between January 2010 and December 2014. For each patient, computed tomography (CT) scans of the knee have been performed. Exclusion criteria were age of less than 16 years old, incomplete CT images, previous knee surgery, and/or existing knee deformity. Additionally, any patient with an open proximal tibial fracture was excluded. Demographic and clinical data such as age, gender, causes of fracture, affected side, Schatzker type, and accompanied injury were collected from the medical records department. 3D reconstruction CT images were saved in an interactive medical image processing software (Mimics 16.0, Materialise, Leuven, Belgium) [15], which allowed identifying the issues with different CT values visible on the scans. Mimics featured extended visualization and segmentation functions based on image density thresholding. We selected the CT data of bone according to the “Thresholding” (CT value > 225) command on the entire stack of scans, and built the mask of tibia by the “Region Growing” command, followed by noise filtering. Then, the 3D analogous image of each fracture was generated. FLA In the top view, the line connecting the middle point of the posterior cruciate ligament’s insertion (B) on the tibial plateau and the medial 1/3 point of the tibial tuberosity (F) acted as a neutral axis. The FLA was determined by counterclockwise measuring the included angle between BF and major fracture line MN of the separated fragment when the fractures occurred in the left leg (Fig. 1a); otherwise, the clockwise FLA was determined. Considering that Schatzker type V and VI fractures were too complex to be measured accurately, only the FLA of Schatzker type I, II, and IV fractures were measured.Fig. 1 The method to calculate the fracture line angle (FLA) and depict the depression. a The FLA is determined by measuring the included angle between BF and major fracture line MN. B is the middle point of the posterior cruciate ligament’s insertion on the tibial plateau; F is the medial 1/3 point of the tibial tuberosity. BF is used as the reference line to determine the major FLA. b The depression zone is depicted with a certain gray value and transparency level. The red points at the border are the position with ≥2 mm depression, but not all the borders are of less than 2 mm of depression Mapping of fracture depression zone As all tibial plateau fractures may combine with more or less depression, a 2-mm depression was considered as a threshold for depression mapping and depth measurement in this study. The method of mapping the fracture line or depression for a given bone, as Cole and colleagues previously described [16, 17], was used for this study. After 3D reconstruction, the border of the depression zone was marked on the top view image. Then, the fracture image was imported into Macromedia Fireworks (v7.0, Macromedia Inc., San Francisco, California) and superimposed to a normal tibial plateau image according to the posterior margin and the line BF, with the size adjusted properly to match the normal. According to the marked border on the fracture image, the depression area was depicted on the normal image with a certain gray value and transparency level (Fig. 1b). Every depression zone was depicted on this normal image as well. With the fracture images superimposed, the gray value of the overlapped area increased. The final gray value (frequency diagram) indicated the incidence of depression at the certain location. Depression depth Using Mimics software (v16.0, Materialise, Leuven, Belgium), the depression depth was measured. The most depressed point X of the articular surface (Fig. 2a), as well as the normal articular surface PL (Fig. 2b), was marked for each fracture after the accurate evaluation by CT scans and the 3D models. Then, the depression depth was obtained by automatically calculating the distance from X to PL (Fig. 2c, d) with Mimics software.Fig. 2 The method to calculate the depression depth. a X is the most depressed point of the articular surface. b PL is the normal articular surface plane. c Anteroposterior. d Lateral. The depression depth is obtained by calculating the distance from X to PL Statistical analyses Qualitative data are shown as n (percentage) and quantitative data are expressed as mean ± SD. The depression depths among different Schatzker types and subtypes were compared using Kruskal-Wallis and Nemenyi tests. All statistical analyses were performed using SAS V8 software (SAS Institute Inc., North Carolina, USA). Difference was considered significant if the P value was less than 0.05. Results Patient demographic and clinical data are summarized in Table 1. In our study, there were 106 male and 80 female patients, with an average age of 54.9 years (range 17–87 years) at the time of injury. Two of them had bilateral tibial plateau fractures. A total of 188 CT scans were reviewed: 115 showed Schatzker type I (12), II (74), or III (29) tibial plateau fractures; 15 showed a Schatzker type IV tibial plateau fracture; and 58 showed Schatzker type V (30) or type VI (28) tibial plateau fractures. There were 159 (84.6 %) tibial plateau fractures with depression.Table 1 Demographic and clinical data of patients with tibial plateau fractures Parameter Patients Mean age (range), years 54.9 (17–87) Gender (male/female), n 106/80 Causes of fracture, n (%)  Road traffic accidents 79 (42.5 %)  Falling from a height 45 (24.2 %)  Riding electromobile injuries 38 (20.4 %)  Others 24 (12.9 %) Affected side, n (%)  Left 112 (60.2 %)  Right 72 (38.7 %)  Bilateral 2 (1.1 %) Simple fracture, n (%) 130 (69.9 %) Complicated fracture, n (%) 56 (30.1 %) Schatzker classification (N = 188)  Type I 12  Type II 74  Type III 29  Type IV 15  Type V 30  Type VI 28 Associated injury, n  Other fracture within the knee joint 45  Non-knee joint fracture 38  Organic lesion 32  Craniocerebral injury 23 The FLA of Schatzker type I and II fractures ranged from 90° to 270°, whereas the Schatzker type IV fracture distributed between 270° and 450°. Mean FLA of Schatzker type I, II, and IV tibial plateau fractures is summarized in Table 2. On the basis of the FLA, both Schatzker type I and II fractures were further categorized into three subtypes: single anterolateral fracture (SALF), single posterolateral fracture (SPLF), and complex fracture (CF), including anterolateral and posterolateral parts. The average FLA of these fractures were 145.92° ± 16.06°, 208.87° ± 6.85°, 149.59° ± 9.33° and 248.80° ± 11.53° (Schatzker type I) and 149.66° ± 13.17°, 235.79° ± 19.85°, 127.41° ± 16.51° and 231.62° ± 14.71° (Schatzker type II), respectively. The Schatzker type IV fracture was also divided into three subtypes: single posteromedial fracture (SPMF), single anteromedial fracture (SAMF), and whole medial fracture (WMF), with the average FLA of 293.27° ± 26.76°, 409.66° ± 17.24°, and 341.10° ± 16.34°, respectively.Table 2 Fracture line angle (FLA) of different Schatzker types of tibial plateau fractures Schatzker type FLA (°) Type I  Single anterolateral fracture (SALF) 145.92 ± 16.06  Single posterolateral fracture (SPLF) 208.87 ± 6.85  Anterolateral part of complex fracture (CF) 149.59 ± 9.33  Posterolateral part of CF 248.80 ± 11.53 Type II  SALF 149.66 ± 13.17  SPLF 235.79 ± 19.85  Anterolateral part of CF 127.41 ± 16.51  Posterolateral part of CF 231.62 ± 14.71 Type IV  Single posteromedial fracture (SPMF) 293.27 ± 26.76  Single anteromedial fracture (SAMF) 409.66 ± 17.24  Whole medial fracture (WMF) 341.10 ± 16.34 Mapping of the 159 tibial plateau fractures resulted in a diverse diagram of fracture zones of depression (Fig. 3). The complete fracture map was divided according to fracture types for visual comparison and analysis. From Schatzker types II to VI, the “heat maps” of depression were depicted. The Schatzker type II fracture had three subtypes according to the FLA, leading to three predictable pattern “heat maps” (Fig. 3a–c), as the tibial plateau depressed surrounding the fracture lines. The depression zone of Schatzker type III is located at the anterolateral tibial plateau (Fig. 3d). Despite the three subtypes of the Schatzker type IV fracture, only two “heat maps” were depicted because SPMF patients had no depression in our cohort (Fig. 3e, f). Beyond our expectations, WMF depressed more frequently at the center and posterolateral side of the tibial plateau. The depression zones of Schatzker type V and VI fractures distributed at the whole lateral plateau (Fig. 3g, h). However, the Schatzker type V fracture was most likely to cause depression at the posterolateral side whereas the Schatzker type VI fracture at the center of the tibial plateau.Fig. 3 The “heat maps” of tibial plateau fractures with depression zone. a–c single anterolateral fracture (SALF), single posterolateral fracture (SPLF), and complex fracture (CF) of Schatzker type II. d Schatzker Type III. e–f Single anteromedial fracture (SAMF) and whole medial fracture (WMF) of Schatzker type IV. g Schatzker type V. h Schatzker type VI. The final gray value (frequency diagram) indicated the incidence of depression at the certain location Comparisons of the depression depths between different types are shown in Table 3. There was a significant difference in depression depth among different Schatzker type fractures by Kruskal-Wallis test (P < 0.01). The severity order of the average depression depth and mean Wilcoxon score were Schatzker types III, II, V, VI, and IV (Table 3). Furthermore, the analysis by Nemenyi test also confirmed a less depression depth of the Schatzker type III fracture than the other types of fractures (Table 3). Further analyses revealed a similar depression depth among the three subtypes of the Schatzker type II fracture (P = 0.2374, Table 3); however, as for the Schatzker type IV fracture, the difference was significant between the SAMF and WMF subtypes (P = 0.0404).Table 3 Depression depth of different Schatzker types and subtypes of tibial plateau fractures Schatzker type Depression rate (%) Depression depth, mm Mean score Type II 74 (100 %) 7.85 ± 4.24 a 73.43  Single anterolateral fracture (SALF) 48 (64.9 %) 8.43 ± 4.26 –  Single posterolateral fracture (SPLF) 4 (5.4 %) 5.59 ± 2.74 –  Complex fracture (CF) 22 (29.7 %) 7.13 ± 4.35 – Type III 29 (100 %) 5.16 ± 1.49 b 44.71 Type IV 9 (60.00 %) 16.7 ± 8.84 a 120.44  Single posteromedial fracture (SPMF) 0 – –  Single anteromedial fracture (SAMF) 3 (33.3 %) 8.15 ± 3.40 –  Whole medial fracture (WMF) 6 (66.7 %) 17.93 ± 10.05 – Type V 26 (86.67 %) 11.99 ± 6.93 a 101.81 Type VI 21 (75.00 %) 13.25 ± 8.73 a 107.57 Fracture types with different letters means that depression depth were significant different between different Schatzker types Discussion Tibial plateau fracture requires great and precise understanding of fracture features for optimal preoperative planning. On the base of radiographic images, Schatzker and AO/OTA classifications describe the location and general pattern of fracture line but without considering the orientation of fracture line [7, 9]. CT and 3D reconstruction software have been recognized as important tools in preoperative protocol to understand radiographic images [8, 18]. Despite the use of CT scanning to assess the location of fracture lines [8], the Three-Column Classification also lacks the discrimination of fracture line orientation. Meanwhile, none of the classifications aforementioned has taken into account the morphological characteristics of depression. Fracture line location is of great importance for the planning of an appropriate surgical approach [19–21]. Depression is the other factor that will influence the choice of surgical approach, anatomical reduction or functional restoration, and even the development of periarticular plates [22]. After the analyses of CT images and 3D reconstruction from 188 different Schatzker types of tibial plateau fractures, this study revealed the distribution of fracture line orientation by the FLA and exhibited the associated morphological characteristics of depression by “heat maps” and depth. Schatzker type I, II and III fracture occur at the lateral tibial plateau. Both Schatzker type I and II fractures had three subtypes (SALF, SPLF, and CF). The SALF can be exposed well by the anterolateral approach according to the statistics of the FLA. However, it was still controversial to expose and treat the posterolateral fracture parts of both SPLF and CF; posterolateral and anterolateral approaches with or without fibula osteotomy were suggested [23–25]. Frosch et al. proposed a posterolateral skin incision with two approaches, which allowed the exposure and examination of the posterolateral joint surfaces of the tibia [24]. Solomon et al. described a posterolateral trans-fibular neck approach to the proximal tibia, which needed an osteotomy and fixation [25]. Meantime, Yu et al. suggested partial fibular head osteotomy [23]. According the average FLA of SPLF, the fibular head will be a great obstacle for exposure and reduction. Screws are most efficient when placed perpendicular to the fracture plane [26]. Consequently, the posterolateral approach with fibula osteotomy could be an efficient and direct way to achieve the goal of anatomical reduction and stable fixation. As for the CF, anterolateral and posterolateral fractures might have a better outcome when treated by bilateral-approach and dual-plate corresponding according to the FLA. Sun and his colleagues also approved of this surgical treatment after clinical observations [27]. Besides, a proper operative approach would make for the stabilizing pattern of fracture reduction. Furthermore, a systemic understanding of fracture morphology such as the FLA would make the inserter internal fixities (hollow screw, blade plate, or intramedullary pin) designed more specially for individual patient, which will also be conducive to the postoperative fracture stability. In this study, SALF and CF of Schatzker type II were both more likely to depress at the anterolateral side of the tibial plateau, so with Schatzker type III. The larger depth of Schatzker type II compared with Schatzker type III might suggest that the Schatzker type II fracture was caused by a more powerful violence. The SPLF of Schatzker type II had an associated depression area with the fracture line. As the depression depths of the three subtypes of the Schatzker type II fracture were not significantly different, it was presumed that the violence might be also similar. The WMF of the Schatzker type IV fracture was shown with a depression at the center and posterolateral side of the tibial plateau. The violence of WMF might be greater than that of the SAMF according to the depression depth (P < 0.01). Sufficient care should be taken for the depression of it [28]. The mean WMF line angle was 341.10° ± 16.34°, suggesting that the fracture lines went more frequently from anteromedial to posterolateral. Considering this, a posteromedial approach or a medial approach was needed for this subtype of the Schatzker type IV fracture. SAMF and SPMF with or without depression surrounding the fracture line accounted for a low proportion of fractures. The SAMF with an average FLA of 409.66° ± 17.24° would be well-exposed and fixed through an anteromedial approach, whereas SPMF with a FLA of 293.27° ± 26.76° needed to be treated via a posteromedial approach [22, 29]. The “heat maps” of Schatzker type V and VI fractures revealed that the anatomical reduction or functional restoration for depression of them were difficult as the depression was located more frequently in the center and the posterolateral side of the tibial plateau. However, not all tibial plateau fractures require open reduction, and fractures of less than 2 mm of depression are stabilized with casting, pinning, or subcutaneously plating without fragment reduction. Rasmussen also advised that an operation should be performed when the depression depth arrives at 5 mm [30]. Therefore, distinguishing the depression with 2 mm for fractures is critical. Apart from the morphological features of different Schatzker types of tibial plateau fractures, we further summarized two key methods of measurement in this study, one for measuring FLA and the other for depression zone and depression depth. Although the FLA was measured only for Schatzker type I, II, and IV fractures because of the difficulty in identifying the anatomical biomarker in Schatzker type V and VI fractures, this can be regularly applied to all preoperative cases as long as the anatomical biomarkers were recognized. When a patient was encountered with a tibial plateau fracture in clinic, Schatzker type was first determined for preliminary judgment, and then the FLA was detected using the method as described in this study, followed by the determination of fracture subtype. Thereby, special management for this subtype of tibial plateau fracture could be performed. For example, if a patient was classified as a Schatzker type IV as well as WMF subtype after the measurement of FLA, the actual depression area and depth of the patient could be determined, as this subtype has been proved with depression at the center and posterolateral areas of the tibial plateau. Accordingly, based on comprehensive, complete, and clear understanding, a posteromedial approach was preferred with a specially designed blade plate that was conducive to the insertion of the screw perpendicular to the major fracture line, so was the case with other Schatzker types of tibial plateau fractures. Hence, this study, to a certain extent, would provide guidance or some experiences to the surgeons during diagnosis, preoperative planning, and execution of surgical strategies. It was a limitation of this study that the FLA of Schatzker type V and VI fractures was not measured due to their complexity. The other limitation was the small sample size, especially the number of Schatzker type IV fractures, which may lead to statistics deviation when compared with the total population of tibial plateau fractures. Conclusions The data and “heat maps” here elucidated the patterns of the tibia plateau fractures: Schatzker type I, II, and IV fractures could be divided into three corresponding subtypes by FLA. The location and depression depth of different Schatzker types of fractures were highly variable. Proper preoperative planning, operative approach, and fixation method should be selected based on the morphological characteristics of individual types of plateau fractures. Abbreviations 3DThree-dimensional ANOVAAnalysis of variance CFComplex fracture CTComputed tomography FLAFracture line angles SALFSingle anterolateral fracture SAMFSingle anteromedial fracture SPLFSingle posterolateral fracture SPMFSingle posteromedial fracture WMFWhole medial fracture Acknowledgements None. Funding Science and Technology Commission of Shanghai Municipality (funding number:13441902500). Availability of data and materials Not applicable. This study was only the primary research, and further study has been in progress. Authors’ contributions PC and HS participated in the design of this study. WW and BN performed the statistical analysis. ZF collected the important background information. HL, PC, and HS drafted the manuscript. All authors read and approved the final manuscript. Competing interests The authors declare that they have no competing interests. Consent for publication Not applicable. Ethics approval and consent to participate This study was approved by the Ethics Committee of Xin Hua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine. ==== Refs References 1. Lowe JA Tejwani N Yoo B Wolinsky P Surgical techniques for complex proximal tibial fractures J Bone Joint Surg Am 2011 93 16 1548 1559 22204013 2. Hall JA, Beuerlein MJ, McKee MD. Open reduction and internal fixation compared with circular fixator application for bicondylar tibial plateau fractures. J Bone Joint Surg Am. 2009;91 Supp 2 (Part 1):74-88. 3. Johnson EE Timon S Osuji C Surgical technique: Tscherne-Johnson extensile approach for tibial plateau fractures Clin Orthop Relat Res 2013 471 9 2760 2767 10.1007/s11999-013-2962-2 23670670 4. Berkson EM Virkus WW High-energy tibial plateau fractures J Am Acad Orthop Surg 2006 14 1 20 31 10.5435/00124635-200601000-00005 16394164 5. Barei DP O’Mara TJ Taitsman LA Dunbar RP Nork SE Frequency and fracture morphology of the posteromedial fragment in bicondylar tibial plateau fracture patterns J Orthop Trauma 2008 22 3 176 182 10.1097/BOT.0b013e318169ef08 18317051 6. Streubel PN Glasgow D Wong A Barei DP Ricci WM Gardner MJ Sagittal plane deformity in bicondylar tibial plateau fractures J Orthop Trauma 2011 25 9 560 565 10.1097/BOT.0b013e318200971d 21654524 7. Schatzker J Mcbroom R Bruce D The tibial plateau fracture: the Toronto experience 1968-1975 Clin Orthop Relat Res 1979 138 94 104 445923 8. Zhu Y Yang G Luo CF Smith WR Hu CF Gao H Computed tomography-based Three-Column Classification in tibial plateau fractures: introduction of its utility and assessment of its reproducibility J Trauma Acute Care Surg 2012 73 3 731 737 10.1097/TA.0b013e31825c17e7 22929503 9. Müller ME, Nazarian S, Koch P, Schatzker J. The comprehensive classification of fractures of long bones. Berlin, Heidelberg: Springer Science & Business Media; 2012. 10. Tscherne H Lobenhoffer P Tibial plateau fractures: management and expected results Clin Orthop Relat Res 1993 292 87 100 8519141 11. Onderko LL Rehman S Treatment of articular fractures with continuous passive motion Orthop Clin North Am 2013 44 3 345 356 10.1016/j.ocl.2013.04.002 23827837 12. Urruela AM Davidovitch R Karia R Khurana S Egol KA Results following operative treatment of tibial plateau fractures J Knee Surg 2013 26 03 161 166 10.1055/s-0032-1324807 23288754 13. Atesok K Finkelstein J Khoury A Peyser A Weil Y Liebergall M The use of intraoperative three-dimensional imaging (ISO-C-3D) in fixation of intraarticular fractures Injury 2007 38 10 1163 1169 10.1016/j.injury.2007.06.014 17884047 14. Gösling T Klingler K Geerling J Shin H Fehr M Krettek C Improved intra-operative reduction control using a three-dimensional mobile image intensifier—a proximal tibia cadaver study Knee 2009 16 1 58 63 10.1016/j.knee.2008.07.012 18945620 15. Magne P Stanley K Schlichting LH Modeling of ultrathin occlusal veneers Dent Mater 2012 28 7 777 782 10.1016/j.dental.2012.04.002 22575740 16. Cole PA Mehrle RK Bhandari M Zlowodzki M The pilon map: fracture lines and comminution zones in OTA/AO type 43C3 pilon fractures J Orthop Trauma 2013 27 7 e152 e156 10.1097/BOT.0b013e318288a7e9 23360909 17. Armitage BM Wijdicks CA Tarkin IS Schroder LK Marek DJ Zlowodzki M Mapping of scapular fractures with three-dimensional computed tomography J Bone Joint Surg Am 2009 91 9 2222 2228 10.2106/JBJS.H.00881 19724000 18. Suero EM Hüfner T Stübig T Krettek C Citak M Use of a virtual 3D software for planning of tibial plateau fracture reconstruction Injury 2010 41 6 589 591 10.1016/j.injury.2009.10.053 19939376 19. Berber R Lewis CP Copas D Forward DP Moran CG Postero-medial approach for complex tibial plateau injuries with a postero-medial or postero-lateral shear fragment Injury 2014 45 4 757 765 10.1016/j.injury.2013.11.028 24377482 20. Tornetta P III Weiner L Bergman M Watnik N Steuer J Kelley M Pilon fractures: treatment with combined internal and external fixation J Orthop Trauma 1993 7 6 489 496 10.1097/00005131-199312000-00001 8308599 21. Yang G Zhu Y Luo CF Putnis S Morphological characteristics of Schatzker type IV tibial plateau fractures: a computer tomography based study Int Orthop 2012 36 11 2355 2360 10.1007/s00264-012-1646-y 22915237 22. Hahnhaussen J Hak DJ Weckbach S Heiney JP Stahel PF Percutaneous inflation osteoplasty for indirect reduction of depressed tibial plateau fractures Orthopedics 2012 35 9 768 772 10.3928/01477447-20120822-04 22955384 23. Yu BQ Han KW Zhan C Zhang CC Ma H Su JC Fibular head osteotomy: a new approach for the treatment of lateral or posterolateral tibial plateau fractures Knee 2010 17 5 313 318 10.1016/j.knee.2010.01.002 20163966 24. Frosch KH Balcarek P Walde T Stürmer KM A new posterolateral approach without fibula osteotomy for the treatment of tibial plateau fractures J Orthop Trauma 2010 24 8 515 520 10.1097/BOT.0b013e3181e5e17d 20657262 25. Solomon LB Stevenson AW Baird RP Pohl AP Posterolateral transfibular approach to tibial plateau fractures: technique, results, and rationale J Orthop Trauma 2010 24 8 505 514 10.1097/BOT.0b013e3181ccba4b 20657261 26. Thomas P, Richard E, Christopher G. AO Principles of fracture management, second expanded edition. Switzerland: AO Publishing; 2007. 27. Sun H Zhai QL Xu YF Wang YK Luo CF Zhang CQ Combined approaches for fixation of Schatzker type II tibial plateau fractures involving the posterolateral column: a prospective observational cohort study Arch Orthop Trauma Surg 2015 135 2 209 221 10.1007/s00402-014-2131-4 25519181 28. Sciadini MF Sims SH Proximal tibial intra-articular osteotomy for treatment of complex Schatzker type IV tibial plateau fractures with lateral joint line impaction: description of surgical technique and report of nine cases J Orthop Trauma 2013 27 1 e18 e23 10.1097/BOT.0b013e31825316ea 22495525 29. Carlson DA Bicondylar fracture of the posterior aspect of the tibial plateau. A case report and a modified operative approach J Bone Joint Surg Am 1998 80 7 1049 1052 9698010 30. Rasmussen PS Tibial condylar fractures J Bone Joint Surg Am 1973 55 7 1331 1350 4586086

PMC005xxxxxx/PMC5002334.txt

==== Front BMC Res NotesBMC Res NotesBMC Research Notes1756-0500BioMed Central London 222910.1186/s13104-016-2229-6Research ArticleAre Danish doctors comfortable teaching in English? Nilas L. Lisbeth.Nilas@regionh.dk 12Løkkegaard E. C. Ellen.Christine.Leth.Loekkegaard@regionh.dk 23Laursen J. B. Jacob.Brink.Laursen@regionh.dk 1Kling J. joyce@hum.ku.dk 4Cortes D. Dina.Cortes@regionh.dk 251 Department of Obstetrics and Gynecology 537, Hvidovre University Hospital, 2650 Hvidovre, Denmark 2 Faculty of Health Science, University of Copenhagen, Copenhagen, Denmark 3 Department of Obstetrics and Gynecology, Nord Zeeland University Hospital, Hilleroed, Denmark 4 Centre for Internationalization and Parallel Language Use, University of Copenhagen, Copenhagen, Denmark 5 Department of Pediatrics, Hvidovre University Hospital, Hvidovre, Denmark 27 8 2016 27 8 2016 2016 9 1 42020 4 2015 17 8 2016 © Nilas et al 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background From 2012–2015, the Departments of Obstetrics and Gynecology and of Pediatrics at the University of Copenhagen conducted a project, “Internationalization at Home ”, offering clinical teaching in English. The project allowed international students to work with Danish speaking students in a clinical setting. Using semi-quantitative questionnaires to 89 clinicians about use of English and need for training, this paper considers if Danish clinical doctors are prepared to teach in English. Results The majority self-assessed their English proficiency between seven and eight on a 10 unit visual analogue scale, with 10 equivalent to working in Danish, while 15 % rated five or less. However, one-fourth found teaching and writing in English to be twice as difficult than in Danish, and 12 % rated all teaching tasks in English at four or less compared to Danish. The self-assessed need for additional English skills was perceived low. Conclusion Teaching in English was rated as 30 % more difficult than in Danish, and a significant subgroup of doctors had difficulties in all forms of communication in English, resulting in challenges when introducing international students in non-native English speaking medical departments. Keywords International studentsClinical teachingTeaching in foreign languageDoctors’ English skillsSelf-assessmentissue-copyright-statement© The Author(s) 2016 ==== Body Background Current strategic policy at the University of Copenhagen (UCPH) aims to internationally promote education, create partnerships and collaborate in high level scientific study [1]. One specific area of internationalization is the inclusion of non-Danish students in traditionally Danish degree programs. For example, in 2013 the Faculty of Health at UCPH enrolled 116 international students, both full time students and Erasmus exchange students [2, 3]. To meet the needs of this more diverse student body, courses traditionally taught in Danish were taught in English. At present, the majority of English medium instruction (EMI) clinical courses at the School of Medical Sciences (MedSchool) are taught through traditional lectures and elective clinical training sessions. Since 2006, UCPH has also offered a limited number of clinical courses in hospital departments taught in English. In contrast to classroom based EMI medical courses, these clinical training courses offer local and international students the opportunity to study and work together in a hospital setting with local patients and offer the Danish students to interact with the international students. While enrolment of international students in courses at UCPH’s MedSchool has resulted in challenges related to both linguistic and cultural diversity for the students, it has also affected the teaching staff, and lecturers are adjusting to this change. Whereas some of the clinical doctors have been positive to teaching in English, there has also been some resistance to internationalization of courses and the use of English as the language of instruction and teaching. To investigate possible causes and determine whether any of these are related to linguistic proficiency, we asked Danish doctors in this clinical setting to consider how comfortable they are working and teaching in English. Methods Setting In 2012, a 3-year educational project entitled Internationalization at Home, was introduced at the departments of Obstetrics and Gynecology (ObGyn) and Pediatrics (Ped) at Hvidovre Hospital and Nordsjaellands Hospital in Hilleroed. Each 5-week course concluded with an oral examination. Of the 6–10 students in each course, half were local UCPH students, who voluntarily registered for this English medium programme, while the other half came from various European countries. The language of instruction of these courses was English, not only during formal lectures and case-based learning session, but was also intended at departmental meetings. The morning meetings, where acute referrals, plans for patient care, and teaching are discussed, were scheduled to be conducted in English when the international teams consisting of both the Danish and the international students were present. One of the departments chose to keep the daily morning meetings in Danish, but instead offer more intensive special English-spoken morning meetings for the international student team. Survey Using a paper-based questionnaire, distributed at morning meetings at three of the departments, we asked the doctors to reflect on their English proficiency, including frequency of language use and level of difficulty communicating in English in the workplace, as compared to Danish usage. Information about the frequency with which the doctors listen to, read, write, speak and lecture in English was collected using a semi-quantitative questionnaire with five response options: never, rarely, sometimes, often, and very often. To further investigate whether attitudes about English proficiency were related to age and teaching experience, the questionnaire included questions about age, experience and gender, but not charge. On a continuous 10 cm visual analogue scale (VAS) with questions related to specific tasks, we asked the doctors to state the ease by which they read, understand, write, guide, teach, and possibly conduct examinations in English compared to in Danish, where 0 represents very difficult and 10 unproblematic and as easy as in Danish. For comparison, information was also collected on how the doctor’s perceived these tasks in Danish. Furthermore, doctors were asked to self-assess their overall proficiency in English from very poor to expert level. Finally, the doctors were requested to assess their need for additional English language skills for the future on a scale ranging from currently satisfied to in great need for improved proficiency. Statistical analyses Of 91 questionnaires distributed during a morning meeting, two were only partly completed leaving 89 for analyses. Responses on the 10 cm VAS were read by 0.5 cm accuracy. A response of having performed an examination once was interpreted as having examination experience. For the questions comparing the ease of performing in English compared to Danish, the values on the VAS were interpreted in percentage differences i.e., a VAS of seven for teaching is interpreted as 30 percent more difficult than in Danish. Numeric outcomes are given as geographical means (means calculated on log transformed values), ranges, and 25–75 percentiles. Statistical differences were tested by Student’s-t test using a level of significance of 5 %. Ethics Informed consent was not formally asked for, but the respondents were anonymous and they could decline participation. No formal ethics approval was required for our study according to Danish law. Results Use of English by the respondents More than 60 % of the respondents indicate they often or very often read English medical texts; the corresponding figures for reading fiction and newspapers were 20 and 5 % respectively (Fig. 1). Written English was mostly used in an academic context, and less than one-third of the respondents often speak English. About 40 % rarely or never participate in discussions in English. The majority of the respondents had been involved in clinical bedside teaching of medical students, mostly in Danish (82 %, 74/89), whereas half (49 %, 44/89) have experienced bed-side teaching of English speaking students (Fig. 2). Only a minority of the respondents had experience in performing lectures or examinations in English.Fig. 1 How often do you read, write, speak, and participate in discussions in English? Fig. 2 How often do you guide clinically, teach in classrooms, and conduct examinations in Danish and English? Self-assessment of English skills The majority of respondents reported a self-assessment of between seven and eight on a VAS scale for general English performance, but 17 % rated their competencies at five or less (Fig. 3). The general skills in English showed a declining trend with age with average values of 7.3, 7.2, 6.7 and 6.0 for the age groups 29–39, 40–49, 50–59 and 60+ years, respectively. Those who examined occasionally or frequently (n = 17) were older than the 72 without experience in examinations (50.2 versus 42.8 years, p < 0.05), but there was no difference between the two groups self-assessment of their general, professional or daily English language skills. There was no difference between the overall English skills in obstetricians/gynecologists and pediatricians, nor between men and women.Fig. 3 Self-assessed English skills: How do assess your overall proficiency in English? VAS ranging from very poor (0) to expert-level (10) Regarding the more specific tasks, the respondents generally rated their skills for reading English high, but about 30 % more difficult than in Danish (Table 1). The respondents reported that they were comfortable with writing professionally, lecturing, classroom teaching and guiding clinically in Danish, with mean values of 8.6, 8.6, 8.9 and 9.2, respectively, and 25–75 percentiles between eight and ten. The mean values for performing these tasks in English compared to in Danish were 6.4, 6.4, 6.6 and 6.5, respectively, with 25 percentiles around five (Table 1), indicating that one in four respondents found teaching and writing in English to be twice as difficult as in Danish. Eleven (12 %) of the respondents rated all teaching tasks in English at four or less compared to Danish, indicating great difficulties in all forms of communication in English.Table 1 Self-assessed language proficiency English compared to Danish Danish Mean Median (range), 25–75 percentiles Mean Median (range), 25–75 percentiles Read medical literature 7.7 7.5 (0.2–10), 6.8–8.5 Read novels 7.2 8 (3–10), 7–10 Read English newspapers 6.7 7 (1.5–10), 6–9 Understand spoken daily language 7.9 8 (3.0–10), 7–10 Understand news programs 7.7 8 (2.0–10), 7–9 Write professionally 7.5 8 (3–10), 8–9 8.6 9 (2–9), 8–10 Lecture students/colleagues 6.4 7 (0.5–10), 5.5–8.5 8.6 9 (2–10), 8–10 Guide clinically 6.4 7 (0.5–10), 5–9 9.2 9 (7–10), 9–10 Teach in classrooms 6.6 7 (1.0–10), 6–8 8.9 9 (6–10), 8–10 Conduct examinations (n = 17) 6.5 7 (0–10), 5–8 How easy is it for you to read, write and teach in English compared to in Danish? VAS ranging from 0, very difficult, to 10, which is just as easy as in Danish n = 89. Data are given as geometric means, range and 25–75 percentiles The self-assessed need for further skills in English varied from a no need (0) to a great need (10), but the majority had little interest in improving their linguistic skills (Fig. 4). The need of additional language skills was not different between those with teaching and examination duties and those without.Fig. 4 Self-assessed need for further linguistic English competencies. VAS ranging from no need (0) to large need (10) Discussion The study shows that teaching and communication in English compared to Danish is assessed to be about 30 % more difficult, and that approximately 10 % of doctors find it very difficult to teach and communicate orally in English. English is the primary foreign language in Denmark and is taught for at least 7 years during primary school. Some of the medical text books at the MedSchool are in English, but English has not previously been used as the language of instruction. Specialist doctors are to some degree trained to acquire medical knowledge in English, but an academic career demands ability also to communicate in English, as 90 % of the medical scientific literature is disseminated in English. The participating departments are common clinical departments where approximately half of doctors are specialists and half are younger doctors at different stages of specialized training. Some doctors have been on the wards for many years and others have just joined the team. Two of the departments have had a few English speaking students for several years, while teaching in English had only just been introduced at the third department at the time of the study. With 30–52 new students per semester at each department, it is necessary to draw on the skills of all doctors at the departments, as confirmed by the fact that over 80 % of the respondents, regardless of level of medical responsibility, stated that they work with students in the clinical setting. The daily classroom lessons are shared between lecturers/professors, but may also be led by senior trainees/residents, as training in teaching is an essential part of the specialist medical training program, a common trend at university hospitals [4–6]. The introduction of EMI courses was a pragmatic decision. The doctors received no formal supervision or offer of English language training. Those who had difficulties communicating in English could to some extent avoid formal teaching, but they were still faced with meetings held in English and interaction with students on the ward. The questionnaire used in this study was not validated and relies on the physicians’ self-assessment, and generally self-assessment has been found to correlate poorly with performance in some contexts [5]. Doctors generally rate themselves highly as teachers [4], and often assume they have teaching qualifications without courses in medical pedagogy [6, 7]. It is also recognized that teachers often overestimate their linguistic abilities, their teaching and their communicative skills [8]. However, as we asked about the clinician’s assessment of the ease of teaching and instructing in English compared to in Danish, we assume that any overestimation may be equivalent in the two situations. The survey was performed during morning meetings and the respondents were expected to be representative for the staff. We have no comparable data from other departments, but find no reason to believe that the doctors in our departments differ substantially from other clinicians in a hospital setting. This study reports the doctors’ self-assessed perceptions of strong receptive English language skills, i.e., listening and reading comprehension. Several of the doctors stated that they have fair competences in writing medical English, but while many found it difficult, they often also found writing medical Danish challenging. However, overall, the respondents in this study found oral production in teaching in English to be the most challenging, indicating that receptive language skills not were sufficient to ensure productive competences. This supports previous research, which suggests that the teachers find themselves less confident when teaching in English compared to their own Scandinavian language [9, 10]. In addition, previous studies have reported that doctors’ EMI teaching was more formal and contained less small-talk compared to teaching in Danish [9–11]. The self-assessed need for further English skills varied. Those who found their linguistic abilities sufficient ranged from one mother-tongue English speaker, to others who had lived in English speaking counties either as children or had studied abroad. In addition, there were some who may have felt their English knowledge was sufficient, because they were on their way toward careers as general practitioners. Introducing education in a foreign language is associated with many challenges [10, 11], and in a clinical setting the challenges are not only linguistic, but also involve attitudes and opinions, and demand structural changes in the organization, and are time consuming. Danish Universities demand internationalization and acknowledge that linguistic competences for both medical students and teachers need to improve. UCPH has established a center for support, the Centre for Internationalisation and Parallel Language Use (CIP), which develops courses for improving the English skills for university staff [12–14]. A study from CIP [14] reported, that the attitudes of the academic staff towards increased use of English as the educational language at the University of Copenhagen, depended on the teaching load in English. While the MedSchool is increasingly offering courses held in English, it is doubtful if such initiatives can ameliorate the linguistic shortcomings in a clinical department, as a clinical course also involves communication with patients and other groups of employees, who may find English communication even more difficult. We find that teaching English-speaking students is a challenge to a clinical department, and our results question whether internationalization through teaching in English in clinical departments is the best possible model. Abbreviations EMIenglish-medium instruction IaHInternationalisation at Home MedSchoolSchool of Medical Sciences (MedSchool) UCPHUniversity of Copenhagen VASvisual analogue scale Authors’ contributions LN wrote the first draft of the manuscript, LN and JB made the calculations and all authors contributed to and accepted the final manuscript. All authors read and approved the final manuscript. Competing interests The authors declare that they have no competing interest. Funding “Internationalization at Home” was funded by Undervisningskvalitetspuljen, The Capital Region of Denmark and University of Copenhagen. ==== Refs References 1. University of Copenhagen, Strategy 2016. http://research.ku.dk/strategy/. Retrieved 25 April 2016. 2. University of Copenhagen, Studieophold i udlandet. http://udrejse.ku.dk/Udveksling_Feltarbejde_og_Praktik/Europa/english/. Retrieved 22 Feb 2015. 3. University of Copenhagen, Statistikberedskab. http://tal.ku.dk/boks1/Statistikberedskab_2013__Studienoegletal.pdf. Retrieved 25 April 2016. 4. Busari J Scherpbier AJJA van der Vleuten CPM Essed GGM The perceptions of attending doctors of the role of residents as teachers of undergraduate students Med Educ. 2003 237 241 247 10.1046/j.1365-2923.2003.01436.x 12603763 5. Eva KW Regehr G Exploring the divergence between self-assessment and self-monitoring Adv Health Sci. Educ. 2011 16 311 329 10.1007/s10459-010-9263-2 6. Erie JA Starkman SJ Pawlina W Lachman N Developing medical students as teachers: an anatomy-based student-as-teacher program with emphasis on core teaching competencies Anat Sci Educ 2013 6 385 392 10.1002/ase.1364 23907836 7. Butani L Paterniti DA Tancredi DJ Li S-T Attributes of residents as teachers and role models—a mixed methods study of stakeholders Med Teach 2013 35 e1052 e1059 10.3109/0142159X.2012.733457 23137246 8. Hellekjaer GO, Westergaard MR. An exploratory survey of content learning through English at Scandinavian universities. In: Simensen AM editor. Teaching and learning foreign languages. Issues and ideas. Oslo: Unipub; 2002. p. 47–61. 9. Tange H Caught in the tower of Babel: University lecturer’s experiences with internationalisation Lang Intercult Commun. 2010 10 137 149 10.1080/14708470903342138 10. Hellekjær GO Wilkin R Zegars V The implementation of undergraduate level English medium programs in Norway: An explorative case study Researching content and language integration in higher education 2007 Maastricht Universitaire Pers Maastricht 68 81 11. Borgan M, Kase BF, Bjerkeset S, Handal G. Engelskspråklig undervisning i medisinutdanningen. En evaluering af prøveordning med engelsk som undervisningsspråk i 9. semester. Oppdragsgruppen ved Pedagogisk Forskningsinstututt. Universitetet i Oslo, det Utdannensvitenskapelige Fakultet. 2005. 12. Internationalisation and language competencies at the University of Copenhagen. Centre for Internationalisation and Parallel Language Use. http://cip.ku.dk/satsningsomraader/pulje2016internationalisering/. Retrieved 25 April 2016. 13. Gregersen F (Ed). 2014. Hvor parallelt. Om parallellspåkighet på Nordens universitet. http://rafhladan.is/bitstream/handle/10802/6165/TN2014535%20web.pdf?sequence=1. Retrieved 25 April 2016. 14. Attitudes towards English as the language of instruction among Danish university lecturers at the University of Copenhagen. Centre for Internationalisation and Parallel Language Use. University of Copenhagen. 2009. http://cip.ku.dk/pdf/publ/Underviseres_holdninger_til_engelsk_som_undervisningssprog.pdf. Retrieved 23 April 2016.

PMC005xxxxxx/PMC5002335.txt

==== Front Crit CareCritical Care1364-85351466-609XBioMed Central London 141410.1186/s13054-016-1414-2ReviewSevere community-acquired pneumonia: timely management measures in the first 24 hours Phua Jason jason_phua@nuhs.edu.sg 12Dean Nathan C. nathan.dean@imail.org 34Guo Qi qiguo007@sina.com 56Kuan Win Sen win_sen_kuan@nuhs.edu.sg 78Lim Hui Fang hui_fang_lim@nuhs.edu.sg 12Lim Tow Keang tow_keang_lim@nuhs.edu.sg 121 Division of Respiratory and Critical Care Medicine, University Medicine Cluster, National University Hospital, National University Health System, Tower Block, Level 10, 1E Kent Ridge Road, Singapore, 119228 Singapore 2 Department of Medicine, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore 3 Department of Medicine, University of Utah School of Medicine, Salt Lake City, UT USA 4 Division of Pulmonary and Critical Care Medicine, Department of Medicine, Intermountain Medical Center, Salt Lake City, UT USA 5 Department of Respiratory Medicine, Affiliated Futian Hospital, Guangdong Medical College, Shenzhen, Guangdong China 6 Guangzhou Institute of Respiratory Diseases (State Key Laboratory of Respiratory Diseases), First Affiliated Hospital, Guangzhou Medical University, Guangzhou, Guangdong China 7 Department of Emergency Medicine, National University Hospital, National University Health System, Singapore, Singapore 8 Department of Surgery, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore 28 8 2016 28 8 2016 2016 20 1 237© The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Mortality rates for severe community-acquired pneumonia (CAP) range from 17 to 48 % in published studies. In this review, we searched PubMed for relevant papers published between 1981 and June 2016 and relevant files. We explored how early and aggressive management measures, implemented within 24 hours of recognition of severe CAP and carried out both in the emergency department and in the ICU, decrease mortality in severe CAP. These measures begin with the use of severity assessment tools and the application of care bundles via clinical decision support tools. The bundles include early guideline-concordant antibiotics including macrolides, early haemodynamic support (lactate measurement, intravenous fluids, and vasopressors), and early respiratory support (high-flow nasal cannulae, lung-protective ventilation, prone positioning, and neuromuscular blockade for acute respiratory distress syndrome). While the proposed interventions appear straightforward, multiple barriers to their implementation exist. To successfully decrease mortality for severe CAP, early and close collaboration between emergency medicine and respiratory and critical care medicine teams is required. We propose a workflow incorporating these interventions. Keywords SepsisCare bundlesResuscitationEmergency departmentIntensive care unitissue-copyright-statement© The Author(s) 2016 ==== Body Background Community-acquired pneumonia (CAP) has plagued humankind for millennia. Hippocrates described pneumonia as a disease which the “ancients” named, and stated that “when pneumonia is at its height, the case is beyond remedy if he is not purged” [1]. Prognosis remained bleak through the centuries. Osler, widely recognised as the father of modern medicine, called pneumonia the “captain of the men of death” in 1901 [2]. Although outcomes improved with the advent of antibiotics, CAP continues to be one of the world’s leading causes of hospitalisation, morbidity, and mortality [3, 4]. Studies have shown that between 2 and 24 % of patients presenting to hospitals with CAP require admission to an ICU [5–7]. Hospital mortality rates for these patients range from 17 to 49 % in large multicentre cohort studies [8–10], and there are conflicting data on whether these rates are increasing or decreasing over time [11, 12]. Calls have been made to treat CAP as an emergency, with aggressive interventions to lower mortality [13]. Unfortunately, while national and international clinical practice guidelines for CAP typically review severity assessment, diagnostic tools, and choice of antibiotics, they do not emphasise the importance of timely resuscitation and care of respiratory failure [14–17]. In this review, we will explore how early and aggressive management measures may result in decreased mortality in severe CAP. We focus on the impact of management measures implemented within the first 24 hours and carried out both in the emergency department (ED) and in the ICU. These measures comprise those which specifically target severe CAP, such as identification and antibiotics, as well as those which target its complications, including septic shock and respiratory failure. Methods We electronically searched PubMed (1981–June 2016) using a sensitive search strategy without language restrictions and with the following Medical Subject Headings (MeSH®) terms: “pneumonia”, “mortality”, “severe”, “severity assessment tools”, “severity scores”, “emergency service, hospital”, and “intensive care units”, “antibiotics”, “sepsis”, “shock septic”, “resuscitation”, “early goal-directed therapy”, “hypoxaemia”, “acute respiratory distress syndrome”, “patient care bundles”, and “quality improvement”. We supplemented the search by reviewing references of included studies and our files. The findings of key papers on the impact of various management measures on mortality are summarised in Table 1. For key studies on sepsis and respiratory failure which also enrolled patients without severe CAP, we recorded the proportion of patients with pneumonia. We used the principles of the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system to assess the quality of evidence as high (well-done randomised trials), moderate (downgraded randomised trials or upgraded observational studies), low (well-done observational studies), or very low (downgraded observational studies) [18].Table 1 Early management measures and impact on mortality in severe community-acquired pneumonia and its complications Intervention Patient population Number of patients Number with pneumoniaa Mortality reduction Mortality definition Risk (95 % CI) Evidence quality Selected reference CAP severity assessment tools to guide management Severe CAP in hospital 348 348 with CAP Yes Hospital Adjusted OR 0.24 (0.09–0.67) Low; before-and-after study [27] Guideline-concordant antibiotics CAP in hospital 1288 1288 with CAP Yes Hospital mortality Adjusted OR 0.55 (0.30–0.90) Low; observational study [39] Macrolide combination treatment Severe CAP in ICUs 8872 8872 with CAP Yes Hospital, ICU, 28-day or 30-day RR 0.84 (0.71–1.00) Moderate; systematic review of observational studies [48] Early antibiotics within 1 hour Sepsis and septic shock 11,017 Data not available Uncertain Hospital or 28-day OR 0.68 (0.42–1.12) Moderate; systematic review of observational studies [56] Early antibiotics after septic shock Septic shock 2154 838 Yes Hospital Adjusted OR 0.89 (0.88–0.91) per hour earlier Low; observational study [55] Early goal-directed therapy Septic shock 4201 1278 Uncertain 90-day OR 0.99 (0.86–1.15) High; systematic review of RCTs [66] High-flow nasal cannula PaO2/FIO2 ≤ 300 mmHg 310 254 (197 with CAP) Yes 90-day HR 0.50 (0.25–0.99) High; RCT [71] Low tidal volumes and plateau pressures ARDS with PaO2/FIO2 ≤ 300 mmHg 861 69 Yes Hospital RR 0.78 (0.65–0.93) High; RCT [76] High positive-end expiratory pressure ARDS with PaO2/FIO2 ≤ 200 mmHg 2299 1145 Yes Hospital Adjusted RR 0.90 (0.81–1.00) Moderate; systematic review of RCTs; conflicting results with other systematic reviews [79] Low driving pressure ARDS with PaO2/FIO2 ≤ 300 mmHg 3562 1314b Yes Hospital at 60 days RR 0.71 (0.66–0.76) for each 1-SD decrease in driving pressure Moderate; systematic review of RCTs of different objectives and methods [77] Neuromuscular blockade ARDS with PaO2/FIO2 < 150 mmHg 339 262 (130 with CAP) Yes 90-day Adjusted HR 0.68 (0.48–0.98) High; RCT [81] Prone positioning ARDS with PaO2/FIO2 < 150 mmHg 466 281 Yes 28-day HR 0.39 (0.25–0.63) High; RCT [83] Corticosteroids Severe CAP according to severity assessment tools 388 388 with CAP Uncertain Varies between RCTs RR 0.39 (0.20–0.77) Moderate; systematic review of RCTs; conflicting results with other systematic reviews [87] Care bundles CAP in hospital 2118 2118 with CAP Yes 30-day Adjusted OR 0.59 (0.37–0.95) Low; before-and-after study [97] aNumbers include all forms of pneumonia, including CAP and hospital-acquired pneumonia. Number of CAP patients is stated where available bA total of 1314 out of 3449 patients had pneumonia (specific data on pneumonia were not available from one study of 113 patients [106]) CI confidence interval, OR odds ratio, RCT randomised controlled trial, PaO 2 partial pressure of arterial oxygen, FIO 2 fraction of inspired oxygen, HR hazard ratio, ARDS acute respiratory distress syndrome, RR risk ratio, SD standard deviation, CAP community-acquired pneumonia Early recognition of severe CAP Severity assessment tools may help clinicians recognise severe CAP and select patients for early intervention. In 1997, Fine et al. [19] showed that the Pneumonia Severity Index (PSI) predicts mortality. This was followed by validation of the CURB65 (Confusion, Urea, Respiratory Rate, Blood pressure, Age > 65 years) score [20]. However, while both the PSI and the CURB65 score guide decisions to hospitalise patients, they were not designed to guide ICU admission. After two earlier iterations [21, 22], the American Thoracic Society (ATS) and the Infectious Diseases Society of America (IDSA) recommended a set of major and minor criteria for ICU admission in 2007 [14]. The major criteria are invasive mechanical ventilation and/or the need for vasopressors. Fulfilment of the minor criteria requires three or more of the following: tachypnoea, hypoxaemia, multilobar infiltrates, confusion, uraemia, leukopaenia, thrombocytopaenia, hypothermia, and hypotension. Several groups have validated these criteria using existing pneumonia databases [23–26]. Two systematic reviews and meta-analyses found 40 studies on different severity assessment tools to guide ICU admission [5, 6]. Marti et al. [5] found that the 2007 IDSA/ATS rule had a pooled sensitivity and specificity of 84 % and 78 % respectively for the prediction of ICU admission, while Chalmers et al. [6] found a pooled sensitivity and specificity of 61 % and 89 % respectively. When the major criteria were removed from the rule, the minor criteria had a pooled sensitivity and specificity of 57 % and 90 % respectively according to Marti et al. [5], and of 56 % and 92 % respectively according to Chalmers et al. [6]. Other similar tools include the SMART-COP (Systolic blood pressure, Multilobar infiltrates, Albumin, Respiratory rate, Tachycardia, Confusion, low Oxygen, low PH) which had a pooled sensitivity and specificity of 79 % and 64 % respectively, and the SCAP (Severe Community-Acquired Pneumonia) score which had a pooled sensitivity and specificity of 94 % and 46 % respectively [5]. Severity assessment tools have several limitations [13]. Firstly, they do not differentiate preventable mortality (which may be reduced by timely management measures) and non-preventable mortality (such as in older patients with multiple co-morbidities and do-not-resuscitate orders), and evidence that adoption of these tools may improve outcomes remains weak [27]. This concern is partly mitigated by tools such as the IDSA/ATS minor criteria, which do not contain age or co-morbid illness factors and therefore select patients based on acute physiology, as opposed to tools such as the PSI [14, 19]. Secondly, studies of severity assessment tools often did not exclude patients with orders to withhold life-sustaining treatments, thus compromising the validity of their findings [5, 6]. Thirdly, because their sensitivity and specificity are not 100 %, some patients who would have done well even without close monitoring may be admitted to an already resource-limited ICU, while others at risk for death would not be detected [5, 6]. Fourthly, decisions for ICU admission are dependent on local culture and resources [28]. Some groups have used the need for mechanical ventilation and vasoactive agents rather than ICU admission per se as end points [29–31]. Fifthly, it remains to be seen how these tools which are specifically for CAP compare with more generic criteria for sepsis such as the recently introduced quick Sequential Organ Failure Assessment (qSOFA) score, which predicts poor outcomes in the presence of at least two of the following: systolic blood pressure ≤ 100 mmHg, respiratory rate ≥ 22/min, or altered mentation [32]. Early action with or without ICU admission Much work has been devoted to finding the ideal severity assessment tool to guide ICU admission because several studies have linked delayed ICU admission for severe CAP with increased mortality [33–36]. Such a finding, however, does not necessarily mean that earlier ICU admissions will result in improved survival. Some patients are admitted late to the ICU because of subsequent deterioration and not because they are unwell to begin with [37]. This notwithstanding, the adverse effect on mortality of admitting patients from an ED to a general ward rather than straight to an ICU persisted even when accounting for late deteriorations by adding radiographic progression to a logistic regression model [38]. Other patients are admitted to the ICU late because clinicians may sometimes inadvertently underestimate the severity of illness in the ED [38]. In these cases, earlier ICU admission alone is insufficient, and must be paired with aggressive management measures in the ED. Indeed, patients in a Singaporean study deemed critically ill and directly transferred from the ED to the ICU received more fluids and antibiotics in the ED than those triaged to the general wards and transferred to the ICU later [35]. Given the limitations of severity assessment tools, we argue that it is time to move on from the search for the perfect tool to effective translation of currently available tools into action. Lim et al. [27] linked CAP severity assessment tools to management prior to initiation of mechanical ventilation or vasopressors. Their multifaceted intervention used the 2007 IDSA/ATS minor criteria to identify severe CAP early in the ED and to trigger empiric antibiotics within 3 hours of triage, prompt intubation for respiratory failure, fluid resuscitation, and vasopressors for patients with hypotension or shock. This before-and-after study reported a decrease in the hospital mortality rate from 24 % to 6 % (Table 1). While inappropriately delayed ICU admissions decreased from 32 % to 15 %, the intervention also decreased the overall ICU admission rates from 53 % to 39 % because early and appropriate resuscitative measures in the ED helped to stabilise patients who could then be managed on a general ward. Early antimicrobial therapy Common causative organisms for severe CAP include Streptococcus pneumoniae, Staphylococcus aureus, Legionella species, Gram-negative bacilli, Haemophilus influenzae, and influenza A and B viruses [14, 15, 17]. Adherence to guidelines for empiric antibiotics is associated with improved survival [39, 40] (Table 1), although microbial aetiology varies across time and place, and different guidelines have proposed slightly different antimicrobial regimes. In general, however, American, British, and European guidelines all recommend empirically starting a beta-lactam (such as amoxicillin-clavulanate, ampicillin-sulbactam, cefotaxime, or ceftriaxone) plus a macrolide (such as azithromycin or clarithromycin) [14–17]. American and European guidelines suggest that a fluoroquinolone (such as levofloxacin or moxifloxacin) may be substituted for the macrolide [14, 17]. Several systematic reviews which explored the role of macrolides in CAP have arrived at different conclusions, depending on the type of studies included. In general, lower-quality observational studies tend to suggest survival benefits [41, 42], as opposed to studies on non-critically ill patients and randomised trials [43, 44]. Two high-quality randomised trials were recently published: a Dutch trial concluded that beta-lactam monotherapy was non-inferior to beta-lactam plus macrolide combination for 90-day mortality [45], but a Swiss trial could not demonstrate non-inferiority [46]. While these trials included non-ICU patients, macrolides have immune-modulatory effects in addition to antimicrobial properties which may benefit sicker and more septic patients [47]. Indeed, in a systematic review of 25 observational studies of 8872 patients with severe CAP, combination treatment with macrolides was associated with lower mortality compared with treatment without macrolides, including when guideline-concordant regimens of beta-lactam/macrolide and beta-lactam/fluoroquinolone were specifically compared [48] (Table 1). Multidrug resistance is increasingly of concern, and American and European guidelines also suggest that when a Pseudomonas infection is suspected, the beta-lactam used should have antipseudomonal effects (such as piperacillin-tazobactam, cefepime, imipenem, or meropenem) [14, 17]. In addition, the IDSA/ATS recommended in 2005 [49] that extended-spectrum antibiotics should be used for patients admitted from the community if they have risk factors for harbouring resistant organisms, such as a recent hospitalisation, residence in a nursing home or extended care facility, home infusion therapy, chronic dialysis, home wound care, or a family member with a multidrug-resistant pathogen. This form of pneumonia was termed healthcare-associated pneumonia (HCAP). Recent work, however, has shown that the HCAP criteria poorly predict the presence of resistant organisms, and that administration of broad-spectrum antibiotics based on these criteria does not improve outcomes [50–52]. A recent systematic review concluded from four large observational studies that administration of antibiotics for CAP within 4–8 hours of hospital arrival was associated with a 5–43 % relative reduction in mortality, even in non-ICU patients [41]. Multiple investigators have associated early administration of appropriate antibiotics with improved survival in sepsis [53, 54]. Kumar et al. [55] reported that each hour of delay in antibiotic administration lowered survival by 8 % in septic shock (838 out of 2154 patients had pneumonia) (Table 1). However, a recent systematic review of 11 studies which reported the time from recognition of sepsis or septic shock to antibiotic administration did not find such an association [56] (Table 1). The latter analysis, however, must be interpreted with caution due to selection bias from excluded studies, the lack of microbiological data, and the lack of confidence that all patients studied had bacterial sepsis [57]. The Surviving Sepsis Campaign recommends administering antibiotics within the first hour of recognition of sepsis and septic shock, and the same is prudent for severe CAP [54, 58]. Early haemodynamic support Pneumonia is the most common cause of sepsis and often presents with septic shock [59, 60]. In a recent multicentre Spanish study, one-third of hospitalisations for CAP were complicated by sepsis [61]. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) [32] state that patients with septic shock can be identified clinically by a vasopressor requirement to maintain a mean arterial pressure ≥ 65 mmHg, and a serum lactate level ≥ 2 mmol/L after adequate fluid resuscitation, although some controversy exists because most studies from which these criteria were derived measured lactate upon presentation and before fluids. In 2001, Rivers et al. [62] showed that a protocol for haemodynamic optimisation known as early goal-directed therapy (EGDT) decreased hospital mortality in sepsis and septic shock from 46.5 % to 30.5 %. The EGDT bundle included lactate measurement, fluid resuscitation according to the central venous pressure, vasoactive agents to keep the mean arterial pressure at 65–90 mmHg, and red blood cell transfusion and/or inotropes according to the central venous oxygen saturation. Pneumonia accounted for 39 % of the enrolled patients and was the commonest cause of sepsis in the study. In 2014 and 2015, however, three large multicentre randomised trials—the ProCESS, ARISE, and ProMISe studies [63–66]—reported that EGDT was not superior to usual care for ED patients with septic shock (Table 1). While these studies examined sepsis in general, the findings are probably applicable to pneumonia which was the leading source of infection, affecting 1278 out of the enrolled 4201 patients. Subgroup analysis by site of infection in the ProCESS study did not reveal significant differences when compared with the full analysis [63]. It is likely that the data from Rivers et al. [62] pushed clinicians towards more aggressive and early resuscitation in recent years, such that the majority of patients in the usual care arms of the ProCESS, ARISE, and ProMISe studies received prompt fluids and vasopressors, even though central venous pressure and central venous oxygenation were not targeted [63–68]. Early respiratory support Acute respiratory failure frequently complicates severe CAP; the combined impact of acute respiratory failure and sepsis on mortality is exponential [69]. Delay in oxygenation assessment using pulse oximetry or arterial blood gas measurements beyond 3 hours from the time of triage at hospital admission is independently associated with increased mortality [70]. In the multicentre FLORALI study, 310 patients with acute hypoxaemic respiratory failure and a ratio of the partial pressure of arterial oxygen to the fraction of inspired oxygen (PaO2/FIO2) ≤ 300 mmHg, but without haemodynamic instability, were randomised within 3 hours of meeting inclusion criteria to high-flow oxygen therapy through a nasal cannula, standard oxygen therapy delivered through a face mask, or non-invasive ventilation [71]. There were 197 patients with CAP, 37 patients with hospital-acquired pneumonia, and 20 patients with pneumonia and an immunocompromising condition. The assigned treatments were commenced a median time of 60 minutes after randomisation. Although high-flow nasal cannulae did not reduce intubation rates, ventilator-free days were increased at day 28 and survival was improved at day 90 (Table 1). However, use of high-flow nasal cannulae and non-invasive ventilation must not delay endotracheal intubation when needed. Delay in intubation beyond 3 days from the onset of CAP symptoms has been associated with increased mortality [72]. Among patients in a French study who were intubated after failure of non-invasive ventilation for severe CAP, those with a longer delay in intubation were more likely to die [73]. A Korean study showed that delay in intubation after a failed trial of high-flow nasal cannulae increased mortality [74]. Criteria to prompt clinicians to intubate may help prevent such delays. For example, the FLORALI study advised intubation in the presence of haemodynamic instability, neurological deterioration, or persisting or worsening respiratory failure, as defined by two or more of the following: respiratory rate > 40 breaths per minute, signs of high respiratory-muscle workload, copious tracheal secretions, arterial pH < 7.35, and pulse oximetry reading < 90 % for >5 minutes [71]. Pneumonia is the most common cause of acute respiratory distress syndrome (ARDS) [75], and as detailed in Table 1 accounts for a large proportion of participants in multiple ARDS trials. Lung-protective ventilation with low tidal volumes of 6 ml/kg of predicted body weight and limitation of the driving pressure (tidal volume divided by respiratory system compliance) after intubation have been associated with reduced mortality [76, 77] (Table 1). The impact of limiting tidal volumes on mortality is greatest at the start of mechanical ventilation [78]. A patient-level meta-analysis of three large multicentre randomised trials suggested that higher positive-end expiratory pressure may improve survival [79], although subsequent systematic reviews have not found a similar association [80] (Table 1). Early initiation of a 48-hour infusion of cisatracurium for neuromuscular blockade for patients with severe hypoxaemia (PaO2/FIO2 < 150 mmHg) lowers mortality [81, 82], as does early prone positioning [83, 84] (Table 1). Early corticosteroids While systemic corticosteroids attenuate the inflammatory response in CAP, recent data on their role are conflicting. In a multicentre randomised controlled trial of 785 patients, 386 of whom had severe CAP as defined by PSI classes IV and V, Blum et al. [85] found that corticosteroids shortened the time to clinical stability. In another trial of 120 patients with severe CAP and C-reactive protein levels > 150 mg/L, Torres et al. [86] found that corticosteroids decreased treatment failure, although predominantly by halting radiographic progression rather than improving patient-centric outcomes. Three recent meta-analyses came to different conclusions, with two analyses suggesting that corticosteroids decrease mortality in severe CAP [87, 88] (Table 1) and one analysis finding no impact on mortality [89]. The latter review, however, suggested that corticosteroids are safe, and may reduce the risk of ARDS, lengths of hospital and ICU stay, and time to clinical stability. These meta-analyses are limited by the heterogeneity of the included studies, and in particular are skewed by an outlying randomised trial by Confalonieri et al. [90] which found that hydrocortisone reduced mortality from 38 % to 0 %. More data are thus needed before corticosteroids become routine treatments for severe CAP [91]. Early use of care bundles Care bundles combine several management practices to improve outcomes [53]. In 2002, the Surviving Sepsis Campaign recommended that a resuscitation bundle should be performed within 6 hours for septic patients [92]. The bundle included principles from EGDT, in addition to blood cultures and broad-spectrum antibiotics. Recent international analyses by the Surviving Sepsis Campaign showed that compliance to the guidelines was associated with decreased mortality [58, 93]. A few before-and-after studies have focused on sepsis care bundles derived from the Surviving Sepsis Campaign guidelines [58] for severe CAP. In a single-centre study in China, Guo et al. [94] defined severe CAP according to the IDSA/ATS major criteria and applied the Surviving Sepsis Campaign’s 6-hour resuscitation bundle (and a 24-hour management bundle) for these patients. While the intervention was associated with a decrease in overall hospital mortality from 44 % to 29 %, full compliance to the bundles was associated with a greater than twofold decrease. Georges et al. [95] reported that a similar bundle including antibiotic therapy, fluids, and vasoactive agents was associated with a decrease in mortality from 43 % to 31 % in patients with severe CAP as defined by ICU admission in a single French centre. Hortmann et al. [96] implemented a care bundle for all patients with CAP in a German ED. The bundle comprised checklists on history, clinical examination, investigations including lactate measurement and blood cultures, risk stratification according to CRB65, and treatment including antibiotic guidelines and fluid resuscitation. Hospital mortality decreased from 14 % to 11 %. Sixteen United Kingdom hospital trusts participated in a quality improvement programme incorporating a British Thoracic Society care bundle which included the use of the CURB65 score, standardised oxygen assessment and prescription, and chest X-ray scan and targeted antibiotics within 4 hours of hospital admission [97]. Bundle implementation was associated with improved 30-day inpatient mortality (Table 1). Dean et al. [98] showed that four hospital EDs experienced lower CAP mortality after introduction of an electronic clinical decision support tool compared with three usual care hospitals. The tool had multiple features, including electronic calculation of the IDSA/ATS minor criteria coupled with logic for ICU admission. Treatment protocols in the EDs and ICUs guided management [99]. Early collaboration between teams Without explicit guidelines, clinical management differs between individual physicians [100]. Workflows that incorporate the interventions reviewed here are needed, but multiple barriers slow their adoption. Using the framework drawn by Cabana et al. [101], these barriers include: a lack of awareness of, familiarity with, or agreement with the workflows, lack of self-efficacy (e.g. confidence in inserting central venous catheters for vasopressors), disbelief that the workflows can reduce mortality, inertia, and external barriers such as cumbersome workflows, insufficient staff, and insufficient time. Overcoming these barriers requires close liaison between emergency medicine, respiratory medicine, and critical care medicine clinicians, as well as clinical decision support tools. These are unfortunately not emphasised in current guidelines for CAP and sepsis [14–17]. To illustrate, Lim et al.’s intervention [27] was designed and sustained by the same multidisciplinary team over 7 years, comprising representatives from the ED and the respiratory and critical care medicine department. Local champions trained nurses and physicians on the definitions and management of severe CAP. Orientation tutorials were provided for new staff, and posters and forms were displayed prominently. Data on compliance were reviewed every 2–4 weeks during business meetings and email discussions, and regular feedback was obtained to improve the workflow. These processes continue to this day. Guo et al. [94] used a personal pocket information card as a daily reminder of the steps required for severe CAP, and provided weekly feedback to staff on compliance to their workflow through group discussions and posters. Dean et al. [98] not only deployed their electronic pneumonia clinical decision support tool, but also closely followed the principles of change management. The first 24 hours We propose in Fig. 1 a bundled approach to early and aggressive treatment to improve survival for patients with severe CAP. The algorithm is guided by clinical decision support tools available in the ED in paper or electronic form [27, 94, 98]. The approach relies on ED nurses to rapidly diagnose CAP at triage, and then suspect sepsis using the qSOFA score by identifying low blood pressure, tachypnoea, and altered mental state [32]. It empowers the nurses to order lactate measurements, blood cultures, and chest X-ray scans, to insert large-bore intravenous cannulae, and to alert the emergency physicians [27]. Broad-spectrum antibiotics must be given early, and include a beta-lactamase-stable beta-lactam and a macrolide [14–17, 39, 48, 56]. Patients with hypotension and/or elevated lactate are given boluses of crystalloids [32, 58]. Severe CAP is defined by the need for intubation and/or vasopressors, or by more subtle features such as the IDSA/ATS minor criteria [14]. High-flow nasal cannulae may be considered for respiratory failure [71], but the application of stringent semi-elective intubation criteria is preferred to emergent intubation near the point of cardiorespiratory arrest [71–73]. The respiratory and critical care medicine department(s) should be engaged early. Patients who do not improve with initial management measures should be managed in the ICU [27, 98, 102], where norepinephrine may be infused for septic shock [103] and lung-protective ventilation with low tidal volumes and driving pressures is provided for ARDS. Prone positioning and/or neuromuscular blockade may be needed for moderate to severe cases [76, 77, 79, 81, 83].Fig. 1 Suggested approach to early and aggressive management measures for severe community-acquired pneumonia (CAP). ED emergency department, PaO 2 partial pressure of arterial oxygen, FIO 2 fraction of inspired oxygen The proposed approach has its limitations and must be adapted to each individual setting. Selected antimicrobials should be tailored to the local antibiogram. While we have focused on antibiotics for bacterial pathogens, empiric treatment with neuraminidase inhibitors for influenza in the presence of typical symptoms should be considered [14, 104]. Resource-limited settings may not be able to measure lactate levels, and more work is needed to evaluate the clinical utility of the qSOFA score [32]. The IDSA/ATS minor criteria are featured because they are well recognised, validated, and easy to use, but no severity assessment tool has perfect sensitivity or specificity [5, 6, 105] so the physicians’ clinical judgment remains key. Importantly, the workflow is highly dependent on a seamless working relationship between nurses and doctors, and between ED and respiratory and critical care medicine teams. Conclusions Severe CAP has claimed too many lives for too long. The emergency medicine and respiratory and critical care medicine communities should work together to decrease mortality by implementing early and aggressive management measures upon recognition of severe CAP. Abbreviations ARDS, acute respiratory distress syndrome; ATS, American Thoracic Society; CAP, community-acquired pneumonia; CI, confidence interval; CURB65, Confusion, Urea, Respiratory Rate, Blood pressure, Age > 65 years; ED, emergency department; EGDT, early goal-directed therapy; FIO2, fraction of inspired oxygen; GRADE, Grading of Recommendations Assessment, Development and Evaluation; HCAP, healthcare-associated pneumonia; IDSA, Infectious Diseases Society of America; MeSH®, Medical Subject Headings; OR, odds ratio; PaO2, partial pressure of arterial oxygen; PSI, Pneumonia Severity Index; qSOFA, quick Sequential Organ Failure Assessment; SCAP, Severe Community-Acquired Pneumonia; SMART-COP, Systolic blood pressure, Multilobar infiltrates, Albumin, Respiratory rate, Tachycardia, Confusion, low Oxygen, low PH Acknowledgements Not applicable. Funding This work is unfunded. Availability of data and materials Not applicable. Authors’ contributions TKL and JP conceived and designed the manuscript. All authors contributed to the acquisition, analysis, and interpretation of papers included the manuscript. All authors contributed to the drafting and revising of the manuscript. All authors approved the final version of the manuscript. Competing interests The authors declare that they have no competing interests. Consent for publication Not applicable. Ethics approval and consent to participate Not applicable. ==== Refs References 1. Adams F The genuine works of Hippocrates 1849 London Sydenham Society 2. Osler W The principles and practice of medicine, designed for the use of practitioners and students of medicine 1901 4 New York D. Appleton and Company 3. Wunderink RG Waterer GW Clinical practice. Community-acquired pneumonia N Engl J Med. 2014 370 543 551 10.1056/NEJMcp1214869 24499212 4. Quan TP Fawcett NJ Wrightson JM Finney J Wyllie D Jeffery K Increasing burden of community-acquired pneumonia leading to hospitalisation, 1998–2014 Thorax. 2016 71 535 542 10.1136/thoraxjnl-2015-207688 26888780 5. Marti C Garin N Grosgurin O Poncet A Combescure C Carballo S Prediction of severe community-acquired pneumonia: a systematic review and meta-analysis Crit Care. 2012 16 R141 10.1186/cc11447 22839689 6. Chalmers JD Mandal P Singanayagam A Akram AR Choudhury G Short PM Severity assessment tools to guide ICU admission in community-acquired pneumonia: systematic review and meta-analysis Intensive Care Med. 2011 37 1409 1420 10.1007/s00134-011-2261-x 21660535 7. Jain S Self WH Wunderink RG Fakhran S Balk R Bramley AM Community-acquired pneumonia requiring hospitalization among U.S. adults N Engl J Med 2015 375 415 27 10.1056/NEJMoa1500245 26172429 8. Kellum JA Kong L Fink MP Weissfeld LA Yealy DM Pinsky MR Understanding the inflammatory cytokine response in pneumonia and sepsis: results of the Genetic and Inflammatory Markers of Sepsis (GenIMS) Study Arch Intern Med. 2007 167 1655 1663 10.1001/archinte.167.15.1655 17698689 9. Angus DC Marrie TJ Obrosky DS Clermont G Dremsizov TT Coley C Severe community-acquired pneumonia: use of intensive care services and evaluation of American and British Thoracic Society diagnostic criteria Am J Respir Crit Care Med. 2002 166 717 723 10.1164/rccm.2102084 12204871 10. Woodhead M, Welch CA, Harrison DA, Bellingan G, Ayres JG. Community-acquired pneumonia on the intensive care unit: secondary analysis of 17,869 cases in the ICNARC Case Mix Programme Database. Crit Care. 2006;10 Suppl 2:S1. 11. Cavallazzi R Wiemken T Arnold FW Luna CM Bordon J Kelley R Outcomes in patients with community-acquired pneumonia admitted to the intensive care unit Respir Med. 2015 109 743 750 10.1016/j.rmed.2015.04.007 25956021 12. Gattarello S Borgatta B Sole-Violan J Valles J Vidaur L Zaragoza R Decrease in mortality in severe community-acquired pneumococcal pneumonia: impact of improving antibiotic strategies (2000–2013) Chest. 2014 146 22 31 10.1378/chest.13-1531 24371840 13. Ewig S Woodhead M Torres A Towards a sensible comprehension of severe community-acquired pneumonia Intensive Care Med. 2011 37 214 223 10.1007/s00134-010-2077-0 21080155 14. Mandell LA Wunderink RG Anzueto A Bartlett JG Campbell GD Dean NC Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults Clin Infect Dis. 2007 44 Suppl 2 S27 S72 10.1086/511159 17278083 15. Lim WS Baudouin SV George RC Hill AT Jamieson C Le Jeune I BTS guidelines for the management of community acquired pneumonia in adults: update 2009 Thorax 2009 64 Suppl 3 iii1 iii55 10.1136/thx.2009.121434 19783532 16. Eccles S Pincus C Higgins B Woodhead M Guideline Development Group Diagnosis and management of community and hospital acquired pneumonia in adults: summary of NICE guidance BMJ. 2014 349 g6722 10.1136/bmj.g6722 25471702 17. Woodhead M Blasi F Ewig S Garau J Huchon G Ieven M Guidelines for the management of adult lower respiratory tract infections—full version Clin Microbiol Infect. 2011 17 Suppl 6 E1 E59 10.1111/j.1469-0691.2011.03672.x 21951385 18. Guyatt G Oxman AD Akl EA Kunz R Vist G Brozek J GRADE guidelines: 1. Introduction—GRADE evidence profiles and summary of findings tables J Clin Epidemiol 2011 64 383 394 10.1016/j.jclinepi.2010.04.026 21195583 19. Fine MJ Auble TE Yealy DM Hanusa BH Weissfeld LA Singer DE A prediction rule to identify low-risk patients with community-acquired pneumonia N Engl J Med. 1997 336 243 250 10.1056/NEJM199701233360402 8995086 20. Lim WS van der Eerden MM Laing R Boersma WG Karalus N Town GI Defining community acquired pneumonia severity on presentation to hospital: an international derivation and validation study Thorax. 2003 58 377 382 10.1136/thorax.58.5.377 12728155 21. Niederman MS Bass JB Jr Campbell GD Fein AM Grossman RF Mandell LA Guidelines for the initial management of adults with community-acquired pneumonia: diagnosis, assessment of severity, and initial antimicrobial therapy. American Thoracic Society. Medical Section of the American Lung Association Am Rev Respir Dis 1993 148 1418 1426 10.1164/ajrccm/148.5.1418 8239186 22. Niederman MS Mandell LA Anzueto A Bass JB Broughton WA Campbell GD Guidelines for the management of adults with community-acquired pneumonia. Diagnosis, assessment of severity, antimicrobial therapy, and prevention Am J Respir Crit Care Med 2001 163 1730 1754 10.1164/ajrccm.163.7.at1010 11401897 23. Brown SM Jones BE Jephson AR Dean NC Infectious Disease Society of America/American Thoracic Society. Validation of the Infectious Disease Society of America/American Thoracic Society 2007 guidelines for severe community-acquired pneumonia Crit Care Med 2009 37 3010 3016 10.1097/CCM.0b013e3181b030d9 19789456 24. Phua J See KC Chan YH Widjaja LS Aung NW Ngerng WJ Validation and clinical implications of the IDSA/ATS minor criteria for severe community-acquired pneumonia Thorax. 2009 64 598 603 10.1136/thx.2009.113795 19386583 25. Liapikou A Ferrer M Polverino E Balasso V Esperatti M Piner R Severe community-acquired pneumonia: validation of the Infectious Diseases Society of America/American Thoracic Society guidelines to predict an intensive care unit admission Clin Infect Dis. 2009 48 377 385 10.1086/596307 19140759 26. Chalmers JD Taylor JK Mandal P Choudhury G Singanayagam A Akram AR Validation of the Infectious Diseases Society of America/American Thoratic Society minor criteria for intensive care unit admission in community-acquired pneumonia patients without major criteria or contraindications to intensive care unit care Clin Infect Dis. 2011 53 503 511 10.1093/cid/cir463 21865188 27. Lim HF Phua J Mukhopadhyay A Ngerng WJ Chew MY Sim TB IDSA/ATS minor criteria aid pre-intensive care unit resuscitation in severe community-acquired pneumonia Eur Respir J. 2014 43 852 862 10.1183/09031936.00081713 24176994 28. Wunsch H Angus DC Harrison DA Linde-Zwirble WT Rowan KM Comparison of medical admissions to intensive care units in the United States and United Kingdom Am J Respir Crit Care Med. 2011 183 1666 1673 10.1164/rccm.201012-1961OC 21471089 29. Charles PG Wolfe R Whitby M Fine MJ Fuller AJ Stirling R SMART-COP: a tool for predicting the need for intensive respiratory or vasopressor support in community-acquired pneumonia Clin Infect Dis. 2008 47 375 384 10.1086/589754 18558884 30. Espana PP Capelastegui A Gorordo I Esteban C Oribe M Ortega M Development and validation of a clinical prediction rule for severe community-acquired pneumonia Am J Respir Crit Care Med. 2006 174 1249 1256 10.1164/rccm.200602-177OC 16973986 31. Kolditz M Ewig S Klapdor B Schutte H Winning J Rupp J Community-acquired pneumonia as medical emergency: predictors of early deterioration Thorax. 2015 70 551 558 10.1136/thoraxjnl-2014-206744 25782758 32. Singer M Deutschman CS Seymour CW Shankar-Hari M Annane D Bauer M The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) JAMA. 2016 315 801 810 10.1001/jama.2016.0287 26903338 33. Renaud B Santin A Coma E Camus N Van Pelt D Hayon J Association between timing of intensive care unit admission and outcomes for emergency department patients with community-acquired pneumonia Crit Care Med. 2009 37 2867 2874 10.1097/CCM.0b013e3181b02dbb 19770748 34. Restrepo MI Mortensen EM Rello J Brody J Anzueto A Late admission to the ICU in patients with community-acquired pneumonia is associated with higher mortality Chest. 2010 137 552 557 10.1378/chest.09-1547 19880910 35. Phua J Ngerng WJ Lim TK The impact of a delay in intensive care unit admission for community-acquired pneumonia Eur Respir J. 2010 36 826 833 10.1183/09031936.00154209 20185424 36. Renaud B Brun-Buisson C Santin A Coma E Noyez C Fine MJ Outcomes of early, late, and no admission to the intensive care unit for patients hospitalized with community-acquired pneumonia Acad Emerg Med. 2012 19 294 303 10.1111/j.1553-2712.2012.01301.x 22435862 37. Garcia-Vidal C Carratala J Early and late treatment failure in community-acquired pneumonia Semin Respir Crit Care Med. 2009 30 154 160 10.1055/s-0029-1202934 19296415 38. Brown SM Jones JP Aronsky D Jones BE Lanspa MJ Dean NC Relationships among initial hospital triage, disease progression and mortality in community-acquired pneumonia Respirology. 2012 17 1207 1213 10.1111/j.1440-1843.2012.02225.x 22805170 39. Menendez R Torres A Zalacain R Aspa J Martin-Villasclaras JJ Borderias L Guidelines for the treatment of community-acquired pneumonia: predictors of adherence and outcome Am J Respir Crit Care Med. 2005 172 757 762 10.1164/rccm.200411-1444OC 15937289 40. Dean NC Bateman KA Donnelly SM Silver MP Snow GL Hale D Improved clinical outcomes with utilization of a community-acquired pneumonia guideline Chest. 2006 130 794 799 10.1378/chest.130.3.794 16963677 41. Lee JS Giesler DL Gellad WF Fine MJ Antibiotic therapy for adults hospitalized with community-acquired pneumonia: a systematic review JAMA. 2016 315 593 602 10.1001/jama.2016.0115 26864413 42. Nie W Li B Xiu Q Beta-lactam/macrolide dual therapy versus beta-lactam monotherapy for the treatment of community-acquired pneumonia in adults: a systematic review and meta-analysis J Antimicrob Chemother. 2014 69 1441 1446 10.1093/jac/dku033 24535276 43. Asadi L Sligl WI Eurich DT Colmers IN Tjosvold L Marrie TJ Macrolide-based regimens and mortality in hospitalized patients with community-acquired pneumonia: a systematic review and meta-analysis Clin Infect Dis. 2012 55 371 380 10.1093/cid/cis414 22511553 44. Eliakim-Raz N, Robenshtok E, Shefet D, Gafter-Gvili A, Vidal L, Paul M, et al. Empiric antibiotic coverage of atypical pathogens for community-acquired pneumonia in hospitalized adults. Cochrane Database Syst Rev. 2012;(9):CD004418. doi:10.1002/14651858.CD004418.pub4. 45. Postma DF van Werkhoven CH van Elden LJ Thijsen SF Hoepelman AI Kluytmans JA Antibiotic treatment strategies for community-acquired pneumonia in adults N Engl J Med. 2015 372 1312 1323 10.1056/NEJMoa1406330 25830421 46. Garin N Genne D Carballo S Chuard C Eich G Hugli O Beta-lactam monotherapy vs beta-lactam-macrolide combination treatment in moderately severe community-acquired pneumonia: a randomized noninferiority trial JAMA Intern Med. 2014 174 1894 1901 10.1001/jamainternmed.2014.4887 25286173 47. Giamarellos-Bourboulis EJ Immunomodulatory therapies for sepsis: unexpected effects with macrolides Int J Antimicrob Agents. 2008 32 Suppl 1 S39 S43 10.1016/j.ijantimicag.2008.06.004 18707849 48. Sligl WI Asadi L Eurich DT Tjosvold L Marrie TJ Majumdar SR Macrolides and mortality in critically ill patients with community-acquired pneumonia: a systematic review and meta-analysis Crit Care Med. 2014 42 420 432 10.1097/CCM.0b013e3182a66b9b 24158175 49. American Thoracic S, Infectious Diseases Society of America Guidelines for the management of adults with hospital-acquired, ventilator-associated, and healthcare-associated pneumonia Am J Respir Crit Care Med 2005 171 388 416 10.1164/rccm.200405-644ST 15699079 50. Dean NC Webb BJ Health care-associated pneumonia is mostly dead. Long live the acronym PES? Ann Am Thorac Soc 2015 12 239 240 10.1513/AnnalsATS.201501-002ED 25706489 51. Webb BJ Dascomb K Stenehjem E Dean N Predicting risk of drug-resistant organisms in pneumonia: moving beyond the HCAP model Respir Med. 2015 109 1 10 10.1016/j.rmed.2014.10.017 25468412 52. Chalmers JD Rother C Salih W Ewig S Healthcare-associated pneumonia does not accurately identify potentially resistant pathogens: a systematic review and meta-analysis Clin Infect Dis. 2014 58 330 339 10.1093/cid/cit734 24270053 53. Barochia AV Cui X Vitberg D Suffredini AF O’Grady NP Banks SM Bundled care for septic shock: an analysis of clinical trials Crit Care Med. 2010 38 668 678 10.1097/CCM.0b013e3181cb0ddf 20029343 54. Gattarello S Lagunes L Vidaur L Sole-Violan J Zaragoza R Valles J Improvement of antibiotic therapy and ICU survival in severe non-pneumococcal community-acquired pneumonia: a matched case-control study Crit Care. 2015 19 335 10.1186/s13054-015-1051-1 26369551 55. Kumar A Roberts D Wood KE Light B Parrillo JE Sharma S Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock Crit Care Med. 2006 34 1589 1596 10.1097/01.CCM.0000217961.75225.E9 16625125 56. Sterling SA Miller WR Pryor J Puskarich MA Jones AE The impact of timing of antibiotics on outcomes in severe sepsis and septic shock: a systematic review and meta-analysis Crit Care Med. 2015 43 1907 1915 10.1097/CCM.0000000000001142 26121073 57. Sivapalan P Staehr Jensen JU Timing of therapy in sepsis: are early antibiotics important for prognosis? Crit Care Med. 2015 43 2030 2031 10.1097/CCM.0000000000001179 26274712 58. Dellinger RP Levy MM Rhodes A Annane D Gerlach H Opal SM Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012 Crit Care Med. 2013 41 580 637 10.1097/CCM.0b013e31827e83af 23353941 59. Angus DC Linde-Zwirble WT Lidicker J Clermont G Carcillo J Pinsky MR Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care Crit Care Med. 2001 29 1303 1310 10.1097/00003246-200107000-00002 11445675 60. Kaukonen KM Bailey M Pilcher D Cooper DJ Bellomo R Systemic inflammatory response syndrome criteria in defining severe sepsis N Engl J Med. 2015 372 1629 1638 10.1056/NEJMoa1415236 25776936 61. Montull B Menendez R Torres A Reyes S Mendez R Zalacain R Predictors of severe sepsis among patients hospitalized for community-acquired pneumonia PLoS One. 2016 11 e0145929 10.1371/journal.pone.0145929 26727202 62. Rivers E Nguyen B Havstad S Ressler J Muzzin A Knoblich B Early goal-directed therapy in the treatment of severe sepsis and septic shock N Engl J Med. 2001 345 1368 1377 10.1056/NEJMoa010307 11794169 63. ProCESS Investigators Yealy DM Kellum JA Huang DT Barnato AE Weissfeld LA A randomized trial of protocol-based care for early septic shock N Engl J Med. 2014 370 1683 1693 10.1056/NEJMoa1401602 24635773 64. ARISE Investigators ANZICS Clinical Trials Group Peake SL Delaney A Bailey M Bellomo R Goal-directed resuscitation for patients with early septic shock N Engl J Med. 2014 371 1496 1506 10.1056/NEJMoa1404380 25272316 65. Mouncey PR Osborn TM Power GS Harrison DA Sadique MZ Grieve RD Trial of early, goal-directed resuscitation for septic shock N Engl J Med. 2015 372 1301 1311 10.1056/NEJMoa1500896 25776532 66. Angus DC Barnato AE Bell D Bellomo R Chong CR Coats TJ A systematic review and meta-analysis of early goal-directed therapy for septic shock: the ARISE, ProCESS and ProMISe Investigators Intensive Care Med. 2015 41 1549 1560 10.1007/s00134-015-3822-1 25952825 67. Russell JA Moller MH Annane D Early goal-directed therapy: from discovery through enthusiasm to equipoise? Intensive Care Med. 2015 41 1676 1678 10.1007/s00134-015-3857-3 26077056 68. Stevenson EK Rubenstein AR Radin GT Wiener RS Walkey AJ Two decades of mortality trends among patients with severe sepsis: a comparative meta-analysis Crit Care Med. 2014 42 625 631 10.1097/CCM.0000000000000026 24201173 69. Aliberti S Brambilla AM Chalmers JD Cilloniz C Ramirez J Bignamini A Phenotyping community-acquired pneumonia according to the presence of acute respiratory failure and severe sepsis Respir Res. 2014 15 27 10.1186/1465-9921-15-27 24593040 70. Blot SI Rodriguez A Sole-Violan J Blanquer J Almirall J Rello J Effects of delayed oxygenation assessment on time to antibiotic delivery and mortality in patients with severe community-acquired pneumonia Crit Care Med. 2007 35 2509 2514 10.1097/01.CCM.0000287587.43801.9C 17901833 71. Frat JP Thille AW Mercat A Girault C Ragot S Perbet S High-flow oxygen through nasal cannula in acute hypoxemic respiratory failure N Engl J Med. 2015 372 2185 2196 10.1056/NEJMoa1503326 25981908 72. Hraiech S Alingrin J Dizier S Brunet J Forel JM La Scola B Time to intubation is associated with outcome in patients with community-acquired pneumonia PLoS One. 2013 8 e74937 10.1371/journal.pone.0074937 24069367 73. Carrillo A Gonzalez-Diaz G Ferrer M Martinez-Quintana ME Lopez-Martinez A Llamas N Non-invasive ventilation in community-acquired pneumonia and severe acute respiratory failure Intensive Care Med. 2012 38 458 466 10.1007/s00134-012-2475-6 22318634 74. Kang BJ Koh Y Lim CM Huh JW Baek S Han M Failure of high-flow nasal cannula therapy may delay intubation and increase mortality Intensive Care Med. 2015 41 623 632 10.1007/s00134-015-3693-5 25691263 75. Bellani G Laffey JG Pham T Fan E Brochard L Esteban A Epidemiology, patterns of care, and mortality for patients with acute respiratory distress syndrome in intensive care units in 50 countries JAMA. 2016 315 788 800 10.1001/jama.2016.0291 26903337 76. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med. 2000;342:1301–8. 77. Amato MB Meade MO Slutsky AS Brochard L Costa EL Schoenfeld DA Driving pressure and survival in the acute respiratory distress syndrome N Engl J Med. 2015 372 747 755 10.1056/NEJMsa1410639 25693014 78. Needham DM Yang T Dinglas VD Mendez-Tellez PA Shanholtz C Sevransky JE Timing of low tidal volume ventilation and intensive care unit mortality in acute respiratory distress syndrome. A prospective cohort study Am J Respir Crit Care Med 2015 191 177 185 10.1164/rccm.201409-1598OC 25478681 79. Briel M Meade M Mercat A Brower RG Talmor D Walter SD Higher vs lower positive end-expiratory pressure in patients with acute lung injury and acute respiratory distress syndrome: systematic review and meta-analysis JAMA. 2010 303 865 873 10.1001/jama.2010.218 20197533 80. Santa Cruz R Rojas JI Nervi R Heredia R Ciapponi A High versus low positive end-expiratory pressure (PEEP) levels for mechanically ventilated adult patients with acute lung injury and acute respiratory distress syndrome Cochrane Database Syst Rev. 2013 6 CD009098 23740697 81. Papazian L Forel JM Gacouin A Penot-Ragon C Perrin G Loundou A Neuromuscular blockers in early acute respiratory distress syndrome N Engl J Med. 2010 363 1107 1116 10.1056/NEJMoa1005372 20843245 82. Alhazzani W Alshahrani M Jaeschke R Forel JM Papazian L Sevransky J Neuromuscular blocking agents in acute respiratory distress syndrome: a systematic review and meta-analysis of randomized controlled trials Crit Care. 2013 17 R43 10.1186/cc12557 23497608 83. Guerin C Reignier J Richard JC Beuret P Gacouin A Boulain T Prone positioning in severe acute respiratory distress syndrome N Engl J Med. 2013 368 2159 2168 10.1056/NEJMoa1214103 23688302 84. Sud S Friedrich JO Adhikari NK Taccone P Mancebo J Polli F Effect of prone positioning during mechanical ventilation on mortality among patients with acute respiratory distress syndrome: a systematic review and meta-analysis CMAJ. 2014 186 E381 E390 10.1503/cmaj.140081 24863923 85. Blum CA Nigro N Briel M Schuetz P Ullmer E Suter-Widmer I Adjunct prednisone therapy for patients with community-acquired pneumonia: a multicentre, double-blind, randomised, placebo-controlled trial Lancet. 2015 385 1511 1518 10.1016/S0140-6736(14)62447-8 25608756 86. Torres A Sibila O Ferrer M Polverino E Menendez R Mensa J Effect of corticosteroids on treatment failure among hospitalized patients with severe community-acquired pneumonia and high inflammatory response: a randomized clinical trial JAMA. 2015 313 677 686 10.1001/jama.2015.88 25688779 87. Siemieniuk RA Meade MO Alonso-Coello P Briel M Evaniew N Prasad M Corticosteroid therapy for patients hospitalized with community-acquired pneumonia: a systematic review and meta-analysis Ann Intern Med. 2015 163 519 528 10.7326/M15-0715 26258555 88. Horita N Otsuka T Haranaga S Namkoong H Miki M Miyashita N Adjunctive systemic corticosteroids for hospitalized community-acquired pneumonia: systematic review and meta-analysis 2015 update Sci Rep. 2015 5 14061 10.1038/srep14061 26374694 89. Wan YD Sun TW Liu ZQ Zhang SG Wang LX Kan QC Efficacy and safety of corticosteroids for community-acquired pneumonia: a systematic review and meta-analysis Chest. 2016 149 209 219 10.1378/chest.15-1733 26501852 90. Confalonieri M Urbino R Potena A Piattella M Parigi P Puccio G Hydrocortisone infusion for severe community-acquired pneumonia: a preliminary randomized study Am J Respir Crit Care Med. 2005 171 242 248 10.1164/rccm.200406-808OC 15557131 91. Waterer G Systemic corticosteroids and community-acquired pneumonia-cautious optimism or wishful thinking? J Thorac Dis. 2015 7 E622 E624 26793375 92. Dellinger RP Carlet JM Masur H Gerlach H Calandra T Cohen J Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock Crit Care Med. 2004 32 858 873 10.1097/01.CCM.0000117317.18092.E4 15090974 93. Levy MM Rhodes A Phillips GS Townsend SR Schorr CA Beale R Surviving Sepsis Campaign: association between performance metrics and outcomes in a 7.5-year study Crit Care Med 2015 43 3 12 10.1097/CCM.0000000000000723 25275252 94. Guo Q Li HY Li YM Nong LB Xu YD He GQ Compliance with severe sepsis bundles and its effect on patient outcomes of severe community-acquired pneumonia in a limited resources country Arch Med Sci. 2014 10 970 978 10.5114/aoms.2014.46216 25395949 95. Georges H Journaux C Devos P Alfandari S Delannoy PY Meybeck A Improvement in process of care and outcome in patients requiring intensive care unit admission for community acquired pneumonia BMC Infect Dis. 2013 13 196 10.1186/1471-2334-13-196 23631630 96. Hortmann M Heppner HJ Popp S Lad T Christ M Reduction of mortality in community-acquired pneumonia after implementing standardized care bundles in the emergency department Eur J Emerg Med. 2014 21 429 435 10.1097/MEJ.0000000000000106 24384619 97. Lim WS Rodrigo C Turner AM Welham S Calvert JM British Thoracic S British Thoracic Society community-acquired pneumonia care bundle: results of a national implementation project Thorax. 2016 71 288 290 10.1136/thoraxjnl-2015-206834 26197815 98. Dean NC Jones BE Jones JP Ferraro JP Post HB Aronsky D Impact of an electronic clinical decision support tool for emergency department patients with pneumonia Ann Emerg Med. 2015 66 511 520 10.1016/j.annemergmed.2015.02.003 25725592 99. Miller RR 3rd Dong L Nelson NC Brown SM Kuttler KG Probst DR Multicenter implementation of a severe sepsis and septic shock treatment bundle Am J Respir Crit Care Med. 2013 188 77 82 10.1164/rccm.201212-2199OC 23631750 100. Dean NC Jones JP Aronsky D Brown S Vines CG Jones BE Hospital admission decision for patients with community-acquired pneumonia: variability among physicians in an emergency department Ann Emerg Med. 2012 59 35 41 10.1016/j.annemergmed.2011.07.032 21907451 101. Cabana MD Rand CS Powe NR Wu AW Wilson MH Abboud PA Why don’t physicians follow clinical practice guidelines? A framework for improvement JAMA. 1999 282 1458 1465 10.1001/jama.282.15.1458 10535437 102. Na S Kuan WS Mahadevan M Li CH Shrikhande P Ray S Implementation of early goal-directed therapy and the surviving sepsis campaign resuscitation bundle in Asia Int J Qual Health Care. 2012 24 452 462 10.1093/intqhc/mzs045 22899698 103. De Backer D Biston P Devriendt J Madl C Chochrad D Aldecoa C Comparison of dopamine and norepinephrine in the treatment of shock N Engl J Med. 2010 362 779 789 10.1056/NEJMoa0907118 20200382 104. Muthuri SG Venkatesan S Myles PR Leonardi-Bee J Al Khuwaitir TS Al Mamun A Effectiveness of neuraminidase inhibitors in reducing mortality in patients admitted to hospital with influenza A H1N1pdm09 virus infection: a meta-analysis of individual participant data Lancet Respir Med. 2014 2 395 404 10.1016/S2213-2600(14)70041-4 24815805 105. Riley PD Aronsky D Dean NC Validation of the 2001 American Thoracic Society criteria for severe community-acquired pneumonia Crit Care Med. 2004 32 2398 2402 10.1097/01.CCM.0000147443.38234.D2 15599142 106. Brochard L Roudot-Thoraval F Roupie E Delclaux C Chastre J Fernandez-Mondejar E Tidal volume reduction for prevention of ventilator-induced lung injury in acute respiratory distress syndrome. The Multicenter Trail Group on Tidal Volume reduction in ARDS Am J Respir Crit Care Med 1998 158 1831 1838 10.1164/ajrccm.158.6.9801044 9847275

PMC005xxxxxx/PMC5002336.txt

==== Front Biomed Eng OnlineBiomed Eng OnlineBioMedical Engineering OnLine1475-925XBioMed Central London 22110.1186/s12938-016-0221-yResearchPostreconstruction filtering of 3D PET images by using weighted higher-order singular value decomposition Liu Hongbo liuhongbo03@gmail.com 1Wang Kun kun.wang@ia.ac.cn 2Tian Jie tian@ieee.org 121 Engineering Research Center of Molecular and Neuro Imaging of the Ministry of Education and School of Life Science and Technology, Xidian University, 266 Xinglong Section of Xifeng Road, Xi’an, 710126 China 2 Key Laboratory of Molecular Imaging, Institute of Automation, Chinese Academy of Sciences, 95 Zhongguancun East Road, Beijing, 100190 China 27 8 2016 27 8 2016 2016 15 1 10221 4 2016 9 8 2016 © The Author(s) 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Background Positron emission tomography (PET) always suffers from high levels of noise due to the constraints of the injected dose and acquisition time, especially in the studies of dynamic PET imaging. To improve the quality of PET image, several approaches have been introduced to suppress noise. However, traditional filters often blur the image edges, or erase small detail, or rely on multiple parameters. In order to solve such problems, nonlocal denoising methods have been adapted to denoise PET images. Methods In this paper, we propose to use the weighted higher-order singular value decomposition for PET image denoising. We first modeled the noise in the PET image as Poisson distribution. Then, we transformed the noise to an additive Gaussian noise by use of the anscombe root transformation. Finally, we denoised the transformed image using the proposed higher-order singular value decomposition (HOSVD)-based algorithms. The denoised results were compared with results from some general filters by performing physical phantom and mice studies. Results Compared to other commonly used filters, HOSVD-based denoising algorithms can preserve boundaries and quantitative accuracy better. The spatial resolution and the low activity features in PET image also can be preserved by use of HOSVD-based methods. Comparing with the standard HOSVD-based algorithm, the proposed weighted HOSVD algorithm can suppress the stair-step artifact, and the time-consumption is about half of that needed by the Wiener-augmented HOSVD algorithm. Conclusions The proposed weighted HOSVD denoising algorithm can suppress noise while better preserving of boundary and quantity in PET images. Electronic supplementary material The online version of this article (doi:10.1186/s12938-016-0221-y) contains supplementary material, which is available to authorized users. Keywords Positron emission tomography (PET)Higher-order singular value decomposition (HOSVD)Anscombe variance-stabilizing transform (AVST)SimilarityWeightedDenoisingissue-copyright-statement© The Author(s) 2016 ==== Body Background Positron emission tomography (PET) is a non-invasive high-specificity imaging modality that is widely used for both clinical diagnosis and preclinical research because of its ability to reveal the metabolic evolution of molecular and biochemical processes in living organs. However, it suffers from relatively poor spatial resolution and high levels of noise due to the constraints of the injected dose and acquisition time, especially in the studies of dynamic PET imaging. High noise levels make the quantitative interpretation of PET images difficult; this is especially problematic when the image is to be used for early tumor detection when the tumor is small [1], myocardial perfusion [2], or hypometabolism of glucose in studies of early diagnosis of Alzheimer’s disease [3, 4]. In such applications, high noise level contained in PET images will cover small but important lesions and make the edges between organs blurred, further leading to mistakes in diagnosis and quantitation. Therefore, noise reduction is important for interpretation and computer-aided analysis of PET images. Several approaches have been introduced to suppress noise contained in PET images, during or after image reconstruction process. Statistical image reconstruction method such as maximum likelihood expectation maximization (MLEM) [5], its accelerated version ordered subset expectation maximization (OSEM) [6], or other modified algorithms [7–10] can model the noise in acquired data but suffer from increased noise as iterations increase. The regularization algorithm [11, 12], which reduces noise in the statistical reconstruction process by the way of adding a prior or penalty term to the reconstruction process, has promised a good denoising performance, but the determination of the appropriate regularization parameters is a challenging work, which is usually task-dependent. Alternatively, a smoothing filter [11, 13] after reconstruction, such as a Gaussian filter (GF), is widely used in current clinical applications. This kind of filter suppresses noise by averaging adjacent pixels by using a weight function of spatial distance between pixels [14], and has an inherent drawback of blurring edges, which may be clinically relevant. In [15], the authors used masked smoothing method for CT perfusion images denoising, but its performance for PET images has not been proved. Numerous edge-preserving image filters, such as the bilateral filter (BF) [16, 17], anisotropic diffusion filters (ADFs) [18, 19], or priors [11, 20], which are able to detect and maintain edges while smoothing over noise-induced fluctuations, have been introduced to denoise PET images. However, these approaches rely on multiple parameters specified by users according to the study to control the smoothness level and target intensity levels. Actually, it is impossible to preserve all of the interesting features simultaneously [21, 22]. Moreover, these filters usually erase small details, such as small tumors, which can lead to misdiagnosis. Algorithms that incorporate the anatomical information from CT or MRI into PET reconstruction [19, 23, 24] will improve the denoising performance, but the improvement of performance can only be achieved when the boundaries from anatomical image and PET image coincide with each other strongly; if not, these algorithms may introduce incorrect information leading to bias and artifacts [25]. Transform-based filters, such as wavelet-based filters [26–28] and singular value decomposition [29], have also been proposed to denoise PET images by exploiting the uncorrelated properties of noise in the transform domain. Recently, nonlocal denoising methods, such as the nonlocal mean (NLM) filters [30] and block-matching 3D (BM3D) filter [31], have been adapted to denoise medical images [32–35], which exploit the similarity and sparsity among small patches to preserve details while removing noise, and achieve outstanding performance. In [21], the authors extended the application of the NLM filter to the 3D PET image and modified it by incorporating anatomical information from the CT image to constrain the similarity measure within a subset of voxels. This algorithm can produce improved lesion contrast and signal-to-noise ratio (SNR) when compared with other traditional PET image denoising methods. The higher-order singular value decomposition (HOSVD) [36] is another kind of nonlocal denoising method, which is a development on high-order matrix or tensor of the singular value decomposition (SVD) of the two-dimension matrix. It offers a simple method for handling sparsity among similar patches by grouping them into a high-order matrix, then denoising the image in a transform-threshold-inverse transform fashion. The transform bases are learned from the original image and are more adaptable to the image content and may achieve a much sparser representation than fixed bases used in the BM3D method. The HOSVD-based method has been extended into the applications of MR image denoising, and demonstrated state-of-the-art performance [37]. In this study, we aim to investigate and extend the application of HOSVD to denoise 3D PET images. We examined its ability to preserve high-intensity features in denoised images, and the ability to detect potentially low remnant activity by the percentage recovered signal (PRS) in lesion volume of interest (VOI) and the contrast-to-noise ratio (CNR) in the denoised images. Results were compared to those obtained after GF, BF, and ADF filtering of the same data. Methods To extend HOSVD to PET image denoising and propose our HOSVD-based 3D PET images denoising method, we will first give a brief introduction about HOSVD in image denoising. Then, we will introduce the Anscombe variance-stabilizing transformation and its exact unbiased inverse. Finally, we will describe our method and its implementation. Non-local HOSVD denoising for 2D images Similar to most denoising algorithms, the HOSVD denoising method is originally designed for the additive zero mean Gaussian noise. The main assumption in the denoising process is that the observed noisy image Y from a noise free image X with additive Gaussian noise N can be modeled as Y=X+N, where N has a known variance σ2 and zero mean. To find a good estimate X~ of X from Y, the HOSVD denoising method first groups the similar patches for each reference image patch into a 3D stack. Then, the HOSVD transformation of the stack is performed to obtain its representation by the HOSVD bases and coefficients. Based on the uncorrelated properties of noise in the HOSVD transform domain, the coefficients are truncated by a hard thresholding operator, then one denoised estimate of this stack is obtained by the inverse HOSVD transform. Repeating this operation for each reference patch results in multiple estimates for each pixel. The final denoised image is obtained by averaging these estimates. The details of the standard HOSVD denoising process are described in the following. Given a p×p reference patch Pref in the original noisy 2D image, K similar patches are found including Pref and stacked into a 3D matrix Z∈Rp×p×K. Then, the HOSVD of Z is performed and can be formulated as [38] 1 Z=S×1U(1)×2U(2)×3U(3), where U(1),U(2)∈Rp×p, and U(3)∈RK×K are orthonormal unitary matrices. They are computed from the SVD of the unfolding matrices of Z along three directions respectively [38]. ×n stands for the n-mode tensor product. S is a 3D coefficient matrix of size p×p×K. Eq. (1) indicates that the HOSVD can exploit signal sparsity across each dimension. The coefficient matrix S can be computed as: 2 S=Z×3U(3)T×2U(2)T×1U(1)T, where T represents the transpose of a matrix. Then, stack Z can be filtered by a hard-threshold operation for S, and the threshold τ can be determined as follows [39]: 3 τ=σ2log(p2K). Thus, the truncated coefficient matrix S~ with entries s~i can be determined as follows: 4 s~i=siforsi≥τ0forsi<τ where the threshold τ is optimal from a statistical risk viewpoint for any orthonormal basis and for N∼N(0,σ2). This truncation of the HOSVD coefficients is based on the assumption that the underlying noise-free image can be sparsely represented by the HOSVD bases. Using (1), we can obtain an estimate Z~ of Z by following: 5 Z~=S~×1U(1)×2U(2)×3U(3). Repeating the above process for each reference patch in sliding window fashion, multiple estimates at each pixel will be obtained. The final filtered image is produced by averaging these estimates at each pixel. To improve the denoising performance even further, an additional Wiener filter step after the standard HOSVD denoising process can be performed. Let Z~ be a stack of similar patches from the standard HOSVD filtered image, which corresponds to Z from the original noisy image, then the coefficients of Z~ and Z in the HOSVD bases of Z~ are denoted as s~ and s respectively. Then, the filtered coefficients of Z, denoted as s¯, are computed as follows: 6 s¯=ss~2s~2+σ2. In this paper, the standard HOSVD and its Wiener filter-augmented version are referred to as HOSVD-S and HOSVD-W respectively. HOSVD-based denoising for 2D PET images The above-mentioned HOSVD denoising for 2D images cannot be straightly extended to denoise PET images, as (1) for nonlinear reconstruction algorithms, usually, the exact noise model in reconstructed PET images is unknown, even though raw PET data follow the Poisson distribution, and (2) the intensity value at each voxel in PET images is the representation of the activity concentration (Bq/ml) in the corresponding region of the space, which is low, so the assumption of an image-independent noise model does not hold. In fact, for images with low photon counts, such as a PET image or SPECT image, the pixel or voxel intensity can be modeled approximatively with the Poisson distribution, which follows the observations that the noise variation in PET images are multiplicative, asymmetric and varying in the image. In this paper, we approximate noise in reconstructed PET images with the Poisson model. Then, with the Anscombe root transformation [40] and its inverse, the HOSVD-based denoising algorithms can be used to denoise PET images. Here, we employ the closed-form approximation of the exact unbiased inverse of the Anscombe variance-stabilizing transformation [41] to alleviate the biased estimates caused by the nonlinearity of the Anscombe transformation. The Anscombe transformation and its inverse are denoted as AVST and IAVST, respectively, which are formulated as follows: 7 AVST(I)=2I+3/8 and 8 IAVST(Z)=14Z2+1432Z-1-118Z-2+5858Z-3-18, where I represents the noisy PET image, and Z is the denoised data in the HOSVD transform domain. The HOSVD-based denoising algorithms can then be extended to the denoising application of PET images by using the following formula: 9 I~=IAVST(DenHOSVD(AVST(I))), where DenHOSVD(·) represents the HOSVD denoising manipulation, I and I~ are the noisy and denoised PET images, respectively. Weighted HOSVD denoising scheme for 3D PET images Fig. 1 An example denoised image resulting from the straight usage of HOSVD-S in 3D PET image. We given the zoom of a local region of interest in b, as a contrast, we also shown the zoom of corresponding regions in original image a and the denoised image with method proposed in this study c The hard-threshold operation used in HOSVD-S may result quite different between reference and selected similar patches in structure; this is especially true when the exact structure is unavailable in the underlying clean images, as with PET image denoising. In such applications, straight use of HOSVD-S may lead to stair-step artifact in the denoised image, as shown in Fig. 1. Although the HOSVD-W can modify this disadvantage, its use also increases time consumption dramatically to about twice the time required for HOSVD-S. To improve the denoising performance of the HOSVD while keeping the time consumption from dramatically increasing, we assign different weights to different selected similar patches, depending on their level of similarity with the reference patch, in structure and statistics. This is different to the HOSVD-S method, which assign identical weight to each selected similar patch. The similarity level in statistics is measured by the p-value output of the Kolmogorov-Smirnov (K-S) test between reference patch and the selected similar patch. For the calculation of weights, we first consider the patch similarity measure used in the HOSVD-S. The distance between the three-dimension (3D) reference patch Pref and its ith similar patch Pi can be formulated as: 10 d(Pref,Pi)=‖Pref-Pi‖2. Note that the reference patch Pref and a patch P within the search window is similar only if the Euclidean distance d(Pref,P) between them meets the following rule [36]: 11 d2(Pref,P)<3σ2p3. Based on this, we calculate the similarity weight between Pi and Pref as follows: 12 s(Pref,Pi)=exp-12d2(Pref,Pi), which means that patches with high similarity to the reference patch will be assigned a larger weight. To alleviate the disadvantage caused by the original similarity measure, we used a hypothesis test (the one-sided Kolmogorov-Smirnov test) to check how well the values in Pref-Pi conform to N(0,2σ2), and the p-value output by the K-S tests are used to modify the weighting factor by a multiplication factor. We denote the p value output by the K-S test as pKS(Pref,Pi), then the weight of patches that are similar to the reference patch Pref can be formulated as follows: 13 c(Pref,Pi)=s(Pref,Pi)·pKS(Pref,Pi). Thus, the ith 3D patch that is similar to Pref can be formulated as follows: 14 P^i=c(Pref,Pi)Pi. Then, the clustering 4D stack can be formulated as: 15 C(Pref)=(Pref,P^1,P^2,...). Note that C(Pref) is a matrix of order four. The denoised 4D stack is formulated as: 16 C~(Pref)=DenHOSVD(C(Pref))=(P~ref,P~1,P~2,...). The estimation P˘i of Pi can then be obtained as follows: 17 P˘i=1c(Pref,Pi)P~i. Fig. 2 Flow diagram of the proposed HOSVD-KSW algorithm. Note that, the reference patches were obtained by certain voxels for saving execute time. A 4D stack was formed for every reference patch, by clustering itself and its similar patches, with a weighted manner. The stack was then processed orderly by HOSVD, truncated transformation coefficients, and inverse HOSVD, to obtain a estimation value of voxels included in the stack Repeating the above process for each reference patch and averaging the estimates at each voxel, we can obtain the denoised image in the Anscombe transform domain. The final filtered PET image can then be produced by performing the IAVST(·) operation. For clarity, the proposed weighted HOSVD denoising algorithm is referred to as HOSVD-KSW, and the flow diagram is shown in Fig. 2. HOSVD-KSW not only preserves all the advantages of the HOSVD-S method, but suppresses the stair-step artifact caused by the HOSVD-S with similar time consumption. Practical implementation of the HOSVD-based denoising algorithm In this work, three HOSVD-based methods are applied to denoise 3D PET images. The most time-consuming parts of HOSVD-based algorithms are 3D patch grouping and HOSVD transform. In the HOSVD-S and HOSVD-W implementation, the K-nearest-neighbors of the reference patch are found based on (11) and are then grouped into a stack. For HOSVD-KSW, the patches that correspond to the first K-maximum-weight are selected as the K-nearest-neighbors of the reference patch (including the reference patch itself) and similar patches obtained by using (11). To improve the computation efficiency while preserving the denoising efficacy of algorithms, we restricted the size of the searching space of similar patches and the sliding step of the reference patch. On one hand, bigger search windows do not always lead to better results, and a further increase of the search window will increase time consumption, especially in 3D PET image denoising, where similar structures of a special reference patch are usually concentrated in a restricted area close to it. On the other hand, the algorithm execution time is linearly corrected with the number of reference patches selected in the implementation. In our implementation of the HOSVD-based algorithms, the size of the search window is set to 20 voxels and the sliding step is set to p-1 at each image dimension, where p is the size of the reference patch in each dimension. Experiments and results Phantom study To assess various effects in image quality resulting from different denoising methods, two physical phantom studies were performed, one for the mini-Derenzo phantom and another for the small animal image quality phantom which is described by NEMA NU 4-2008 [42].Fig. 3 Physical phantom. The NEMA small animal image quality phantom (a) contains five rods (B) (diameter 1–5 mm) was used to measure the recovery coefficients, homogeneous region to measure the uniformity (A middle), and 2-chamber region (C) (water- and air-filled) to measure the spill-over ratio. The mini-Derenzo phantom (b) was used to measure the effects of different filters on the spatial resolution of PET images The NEMA small animal image quality phantom (homemade), created from polymethylmethacrylate, has three main parts, as shown in Fig. 3a. The first is a homogeneous cavity with a 30 mm diameter and 30 mm length, filled with radioactivity (hot region) to measure the uniformity. The second part consists of two chambers housed in the cavity and filled with water and air (cold regions, 15 mm in length, wall thickness and outer diameter are 1 and 10 mm respectively) to estimate the spill-over ratio. The third part of the phantom contains five rods which are drilled through a plastic region and have diameters of 1–5 mm respectively, and are used to measure the noise and recovery coefficients. The mini-Derenzo phantom with rod diameters ranging from 0.7 to 2.4 mm was used to compare the spatial resolution characteristics of the different methods. Fig. 3b is the schematic diagram of the mini-Derenzo phantom. The small animal image quality phantom was filled with 2.74 MBq of 18F-FDG, and the mini-Derenzo phantom had 1.11 MBq. Data were acquired for 15 min for both phantoms using a GENISYS4 PET scanner (Sofie Biosciences Inc., USA). The images were reconstructed on the GENISYS4 Acquisition Engine with 3D OSEM algorithm (without PSF) into 96 × 96 × 208 voxels, where the voxel dimensions were 0.457 × 0.457 × 0.457 mm. Nude mice 18F-FDG study To compare the different denoising approaches in terms of their effect on preclinical or clinical quantitative studies, we performed an 18F-FDG PET study of nude mice with 4T1 tumors. A 2.24 MBq injection of 18F-FDG mixed with isotonic saline to a total volume of 150 μL was administered into nude mice through a tail vein with the mouse immobilized in a restrainer. Image acquisition started 45 min after injection and was performed on the GENISYS4 PET scanner without attenuation correction. Quantitative evaluations Uniformity To quantitatively evaluate and compare the effect of different denoising algorithms on the uniformity of the PET image, a 22.5 mm diameter (75 % of the active diameter) by 10 mm long cylindrical volume of interest (VOI) was drawn over the center of the uniform region of the NEMA small animal image quality phantom. The average activity concentration in this VOI, and its percentage standard deviation (%STD) were calculated. The percentage standard deviation was calculated as follows: 18 %STD=100×STDuniformMeanuniform, where 19 STDuniform=1n-1∑i∈VOI(Acti-Meanuniform)2 is the standard derivation of activity concentration in a uniform region VOI, and Acti is the activity concentration in the ith voxel of the VOI, and n represents the number of voxels that are contained in uniform region VOI. STDuniform and Meanuniform are the standard deviation (STD) and mean in the uniform region VOI, respectively. Recovery coefficients A single image slice with lower noise characteristics was obtained by averaging image slices that cover the central 10 mm length of the rods. Circular region of interests (ROIs) were taken in this image, around each rod with diameters twice the physical diameter of the rods [42]. Transverse image pixel coordinates of the locations with the maximum voxel value within each of these ROIs were used to create line profiles along the rod in the axial direction. Then, the mean and standard deviation of the recovery coefficient (RC) for the rods were calculated as follows: 20 RC=MeanlineMeanuniform, where Meanline represents the mean of the activity concentration for each line profile. The percentage standard deviation of the recovery coefficient for each rod was calculated as follows: 21 %STDRC=100×STDlineMeanline2+STDuniformMeanuniform2. Spill-over ratio For the calculation of the spill-over ratio of water- and air-filled regions, two cylindrical VOIs with a diameter of 4 mm (half the physical diameter of the cylinders) and encompassing the central 7.5 mm in length were drawn in the water- and air-filled cylindrical inserts respectively. Then, the spill-over ratio (SOR) was calculated as follows: 22 SOR=MeanVOIMeanuniform where MeanVOI was the mean of the activity concentration in the water- or air-filled region VOI. The percentage standard deviation of SOR was calculated as follows: 23 %STDSOR=100×STDVOIMeanVOI2+STDuniformMeanuniform2. Spatial resolution The influence of different algorithms to the spatial resolution of the PET image was evaluated visually by the mini-Derenzo phantom study. Percentage recovery signal and contrast-to-noise ratio In the nude mice study, we compared the ability to preserve high-intensity features of different denoising approaches by two metrics. The first metric is the PRS in a VOI in the denoised image relative to noisy image, and the second is the CNR. These two metrics are important in clinical diagnosis and treatment monitoring. In comparison to PRS, which fails to capture the noise characteristics of the noisy images, CNR is more holistic, especially in studies where the goal is to detect potentially low remnant activity, the CNR can be quite critical. Two VOIs were drawn for the calculation of PSR and CNR, a hot VOI in the lesion region and a low-intensity VOI that has a relatively uniform uptake in the background region (as shown in Fig. 12a). The PRS and CNR were calculated as follows: 24 PRS=MlesdeMlesun×100% and 25 CNR=|Mles-Mback|STDback where Mlesde and Mlesun denote the mean intensity in the lesion VOI in the denoised images and noisy images respectively. Mles denotes the mean intensity in the lesion VOI in denoised or noisy images, and Mback and STDback represent the mean and standard deviation of intensity in the low intensity (background) VOI in the same images. Filtering parameters In this work, 3D GF, BF, and ADF were applied. The only parameter of GF is the standard deviation of the Gaussian kernel, which is denoted as σd and used to control the diffusion rate of the Gaussian kernel. For BF, there are two parameters to be determined to control the smoothing degree of the images, namely the geometric spread σd and photometric spread σr, which determine the geometric closeness and photometric similarity of the neighboring voxels to be used in the smoothing process, in domain and range respectively. There are many flow functions which can be used in ADF, however, considering that one purpose of PET image denoising is to preserve a prominent signal in a homogeneous area, such as a focal lesion within a single anatomical region, we chose the exponential flow function in this work. The smoothing performance of ADF was controlled by two parameters, the diffusion rate λ, and the edge preserving parameter κ. The performance of the HOSVD-based algorithms depends on the setting of two patch-related parameters, patch size p and the number K of similar patches in a group; however, in a rational range, K would not significantly affect the denoising performance. Therefore p is the only free parameter which influences the accuracy in HOSVD-based algorithms.Table 1 Filter parameters setting GF BF ADF (%) HOSVD-S σd σr κ p 0.5 25.5 60 5 0.8 51.0 65 7 1.0 76.5 70 8 1.5 112.0 75 10 2.0 127.5 80 13 In order to compare the performance of HOSVD-KSW with GF, BF, and ADF for PET image denoising, filtering parameters that make all these algorithms balanced should be determined; these selected parameters result similar standard deviation in the homogeneous regions in denoised images. In the phantom study, we chose this region as the uniform VOI, and in the mice study, the background VOI was used. The comparison between HOSVD-based algorithms was performed by using the same patch size p. For GF, the kernel size is fixed to five voxels. We fixed σd at 4.5 in BF. For ADF, keeping λ=0.15 (λ∈[0,0.25] should be sufficiently small to make the diffusion process stable [18]) unchanged. The residual parameters were swept over the range listed in Table 1. Note that the parameter κ in ADF was represented as the percentage of standard deviation inside the selected homogeneous region in this work.Fig. 4 The standard deviation in a uniform VOI as a function of parameters in each filter in NEMA phantom study. a for Gaussian filter (GF), b for bilateral filter (BF) and c for anisotropic diffusion filter (ADF), and d for standard HOSVD-based filter (HOSVD-S) Figure 4 shows the variation of the standard deviation in uniformity region VOI when different filtering parameters setting were used in the phantom study. The parameters corresponding to the standard deviation value circled in red dashed circles were chosen in the phantom study. For GF, σd=1 (1.07 mm FWHM) was used. For BF, σd=4.5 and σr=25.5. λ=0.15 and κ=75% for ADF. For HOSVD-S, HOSVD-W, and HOSVD-KSW, patch size was set to p=7. For the nude mice study, the filtering parameters were σd=1 for GF, σd=3.0 and σr=10 for BF, λ=0.15 and κ=80% for ADF, and p=9 for HOSVD-S, HOSVD-W, and HOSVD-KSW. Results Image quality Fig. 5 The mean and percentage standard deviation (PSR) of normalized counts in the uniform VOI in NEMA phantom study. a for the mean value, and b for the PSR Fig. 6 Counts profiles of the cross-section in uniform region in noisy and denoised images in NEMA phantom study Figure 5 shows the mean (Fig. 5a) and percentage standard deviation (Fig. 5b) of the noisy and denoised image in the uniform VOI. From this result, all of these filters can preserve the mean and improve the smoothness in the uniform region. However, HOSVD-KSW yielded a similar smoothing result with GF, BF, and ADF, which were smoother than HOSVD-S and HOSVD-W. As was expected, GF yielded an over-smoothed boundary in the uniform region while BF and ADF preserved boundary. A similar edge-preserving result also could be observed in HOSVD-based methods. This can be seen in Fig. 6.Fig. 7 The recovery coefficient (RC) and percentage standard deviation (PSTD) of normalized counts in VOI within rod regions in NEMA phantom study. a for the recovery coefficient and b for the percentage standard deviation Figure 7 shows how the recovery coefficients (RC) of rods varies with diameters in the noisy and denoised image under different algorithms. From Fig. 7a, the GF, BF, and ADF deteriorated the RC over all of the rods, while the HOSVD-based algorithms could preserve the RC, and the HOSVD-KSW improved RC slightly even with a smaller rod diameter. Figure 7b shows that HOSVD-S is less stable when compared with HOSVD-KSW, even though the Wiener filter step in HOSVD-W would improve this shortage in HOSVD-S.Fig. 8 Results of spill-over ratio (a) and its percentage standard deviation (b) for the cavities filled with water and air in NEMA phantom study Fig. 9 Counts profiles of the cross-section in water- and air-filled region in noisy and denoised images in the NEMA phantom study Figure 8 shows the results of the spill-over ratio and the corresponding percentage standard deviation for the cavities filled with water and air. As shown in Fig. 8a, the HOSVD-KSW decreased SORs both in water- and air-filled regions while other compared methods enlarged it besides the GF. However, compared with HOSVD-KSW, the improvement of GF to SOR in water- and air-filled regions was negligible. Again, the percentage standard deviation (%STD) of SOR in water- and air-filled regions resulting from the HOSVD-KSW are comparable to those from GF and BF, and the HOSVD-S and HOSVD-W resulted in a higher %STD than other algorithms although both of them reduced it when compare to the noisy image, this was shown in Fig. 8b. Figure 9 shows the profiles of the dotted lines drawn in the center of the water/air chambers region, a reduction of the normalized counts in cold chambers could be observed in HOSVD-KSW. Mini-Derenzo phantom Fig. 10 Transaxial slices that cut through the center of the mini-Derenzo phantom. a noisy image. b GF denoised image with FWHM = 1.07. c BF denoised image with σd=4.5 mm, σr=1.5. d ADF with λ = 0.15, κ = 4. e HOSVD-S, f HOSVD-W, g HOSVD-KSW with p = 9 The results of the mini-Derenzo phantom study are shown in Fig. 10. The noisy image is shown in Fig. 10a. The denoised image using GF with σd=1 is shown in Fig. 10b. As expected, the GF smoothed out signal and noise simultaneously. The BF denoised image demonstrated better preserved hot boundaries than the GF, but the cold boundaries were smoothed out, as denoted by the red arrow in Fig. 10c. ADF yielded a result between GF and BF as shown in Fig. 10d. HOSVD-based algorithms demonstrated better-preserved results of cold boundaries and hot regions. There is no evident difference in the denoised images resulting from HOSVD-S, HOSVD-W, and HOSVD-KSW. This may have benefit from the accurate structural information existing in the mini-Derenzo phantom, as the simple similarity measure used in HOSVD-S is just enough for similar patch selection for a certain reference patch. As shown in Fig. 10, the activity-filled area with the diameter of 1.6 mm can be distinguished clearly in all denoised images. However, the area with a diameter of 1.2 mm was smoothed in GF, BF, and ADF, while the HOSVD-based algorithms demonstrated better-preserved details in this area. It is clear that the HOSVD-based algorithms can preserve the spatial resolution of the PET system well. Nude mice study result Fig. 11 Transaxial slices that cut through the lesions in the mice study. a noisy image. Images filtered with GF (b) at FWHM = 1.07, BF (c) at σd = 3.0, σr = 10 and ADF (d) at λ = 0.15, κ = 2. e–g are the denoised images with HOSVD-S, and HOSVD-W and HOSVD-KSW respectively, with the patch size of p = 9 Fig. 12 Calculation results of contrast-to-noise ratio (CNR) and percentage recovered signal (PRS) in mice study. The VOIs are shown in a, where the background VOI is marked with a red circle and the lesion VOI with light blue circle. The CNR in noisy and denoised images are shown in b, and the PRS in images with different filters are shown in c Figure 11 presents coronal slices through lesions in the nude mice 18F-FDG PET study. PET image shown in (a) is a noisy reconstructed image, (b), (c), (d), (e), (f), and (g) are the images that are denoised by GF, BF, ADF, HOSVD-S, HOSVD-W, and HOSVD-KSW, respectively. Lesions are denoted by yellow arrows in these images. The lesion is ellipsoidal with its length in three axis are  1.1,  0.9 and  0.6 cm respectively, and the volume is  2.5 mL. It can be observed that GF suppressed noise as well as the lesion and organ boundaries. BF suppressed noise to a similar extent as GF in the background region, and yielded sharper boundaries of the lesion and organs, as shown in Fig. 11c. ADF preserved the high contrast objects, but low contrast details such as the lesion were smoothed, and there were stair-step artifacts in the boundaries of high contrast objects as denoted by red arrows shown in Fig. 11d. All of the HOSVD-based algorithms suppressed noise effectively, however, HOSVD-S yielded stair-step artifacts, as shown in Fig. 11e. Both HOSVD-W and HOSVD-KSW preserved the boundaries of the lesion and other high contrast objects, as can be seen in Fig. 11f, g. Quantitative results from the nude mice 18F-FDG PET study are shown in Fig. 12. VOIs in the lesion and background regions used for computing the CNR and PSR are delineated in Fig. 12a. It can be seen that all compared algorithms improved the CNR (Fig. 12b) in the lesion region in comparison to the noisy image. However, GF, BF, and ADF yielded worse results than HOSVD-based algorithms. GF improved CNR by 249 % (CNR = 18) compared to the noisy image (CNR = 5), and introduced 9 % (PSR = 91 %) reduction in average normalized counts in the lesion. BF and ADF improved CNR by 208 and 200 %, and yielded 8 % (PRS = 92 %) and 9 % (PNS = 91 %) reduction in the average normalized counts in the same region, respectively. These demonstrated that traditional edge preserving filters could not achieve better quantitative performance than GF when their smoothness of the background region was similar. However, HOSVD-based algorithms introduced a better CNR and yielded better PRS than other compared algorithms, among which HOSVD-W yielded the best performance, with a 387 % increase in CNR, while maintaining similar average normalized counts (PSR = 96 %) in the lesion when compare to the other two HOSVD-based methods. Compared to HOSVD-W, HOSVD-KSW yielded the second best CNR and PRS with CNR = 24 and PSR = 96 %, but its time consumption was dramatically lower than for the HOSVD-W. Discussion In this study, we investigated the use of a novel denoising algorithm based on high-order singular value decomposition in PET image denoising. This algorithm performs image denoising by exploiting the sparsity between similar patches by using the HOSVD transform, and offers a simple and elegant method for handling sparsity among similar patches. The HOSVD bases are learned from the image and thus are more adaptable to the image content and may achieve a much sparser representation than fixed bases such as wavelet and discrete cosine bases. To evaluate the performance of HOSVD-based denoising algorithms in PET image filtering, we performed three 18F-FDG PET studies with: a NEMA small animal image quality phantom, a mini-Derenzo phantom, and a nude mice with 4T1 tumors. The noisy reconstructed PET images were filtered by GF, BF, ADF, and HOSVD-based denoising methods. Our results demonstrated that HOSVD-based methods yielded consistently superior performance compared with GF, BF, and ADF, in terms of visual quality and quantitative metrics. HOSVD-W and HOSVD-KSW recovered the small hot region better, while preserving the spatial resolution and reducing the SOR of the background regions. On condition that the smoothness of the background region is similar with each other, both methods suppressed the noise in soft tissues effectively and preserved the organ boundaries. The straight usage of HOSVD-S might lead to stair-step artifacts in denoised PET images. This is because the patch similarity measure may collect patch pairs with quite different structures in the Anscombe transformed image with identical weight. If these patches are grouped into a 4D stack then the HOSVD bases that learned from them are not accurate, and lead to inappropriate information spreading into the final denoised PET image. While the Wiener filter step used in HOSVD-W can modify this shortcoming, the time consumption will dramatically increase to about twice as long as the HOSVD-S. Our results demonstrated that the proposed method HOSVD-KSW yielded similar performance compared with HOSVD-W, and the time-consumption was similar with HOSVD-S. The good denoising performance of the HOSVD-KSW is attributed to two main features. The first feature is that HOSVD-KSW groups the similar patches into a stack with a weighted manner, where the weight is calculated according to the similarity between reference patch and special patch in domain and statistics. This will make the HOSVD bases more accurate than those in HOSVD-S, and thus remove the stair-step artifacts existing in HOSVD-S denoised images. The second feature is that the similarity weight of similar patches is calculated by considering two factors. The first factor is the Euclidean distance between a similar patch and reference patch from the similarity measure process, and the second factor is the p value output of the K-S test of the difference patch between the reference patch and the similar patch. This two-part criterion enables the high accuracy of the HOSVD bases. The main shortcoming of the proposed HOSVD-KSW method for 3D PET image denoising is that we considered the noise in the reconstructed PET images as the Poisson distribution. It was not accurate enough, as well known, and a more accurate noise model and corresponding variance-stabilizing transformation should be used to improve the reliability of the results. Conclusions In summary, our study demonstrated that the HOSVD-W and the proposed HOSVD-KSW methods yield improved image quality while preserving the accuracy of lesion quantification. Considering that the time-consumption of the HOSVD-KSW is about half of the HOSVD-W, we believe that the HOSVD-KSW is more practical than the HOSVD-W at a very low performance compromise. Additional file 10.1186/s12938-016-0221-y Original 3D PET images data used in this work to generate the results. Abbreviations PETpositron emission tomography HOSVDhigher-order singular value decomposition AVSTanscombe variance-stabilizing transformation MLEMmaximum likelihood expectation maximization OSEMordered subset expectation maximization GFGaussian filter BFbilateral filter ADFanisotropic diffusion filters NLMnonlocal mean BM3Dblock-matching three dimension filter SNRsignal-to-noise ratio SVDsingular value decomposition PRSpercentage recovered signal VOIvolume of interest CNRcontrast-to-noise ratio HOSVDSstandard HOSVD HOSVD-Wwiener filter-augmented HOSVD K-S testKolmogorov-Smirnov test HOSVD-KSWweighted HOSVD PSFpoint spread function STDstandard deviation ROIregion of interest RCrecovery coefficient SORspill-over ratio Authors' contributions HBL contributed to the derivation and implementation of the algorithm. KW has completed the simulation and experiment together with HBL to verify the performance of proposed method. JT has provided a experimental platform, and made sure the reliablility of experiment data. All authors read and approved the final manuscript. Acknowledgements Authors thank Dr. Karen Milada Von Deneen in Xidian University for proofreading this manuscript. Competing interests The authors declare that they have no competing interests. Availability of data and materials The raw data used for analysis to draw the conclusion has been provided in the Additional file 1. Ethics approval The study was approved by the Animal Care and Use Committee of the School of Life Science and Technology, Xidian University. Funding This work is supported by the Program of the National Basic Research and Development Program of China (973) under Grant No. 2011CB707702, No. 2015CB755500, and the National Natural Science Foundation of China under Grant Nos. 81227901, 81090272, 81527805, 61231004, and 61401462, and the Scientific Research and Equipment Development Project of Chinese Academy of Sciences under Grant No. YZ201359. ==== Refs References 1. Zhao C Chen Z Ye X Zhang Y Zhan H Aburano T Tian M Zhang H Imaging a pancreatic carcinoma xenograft model with 11C-acetate: a comparison study with 18F-FDG Nucl Med Commun 2009 30 12 971 977 10.1097/MNM.0b013e328330adfc 19696689 2. Chen K Lawson M Reiman E Cooper A Feng D Huang S-C Bandy D Ho D Yun L-S Palant A Generalized linear least squares method for fast generation of myocardial blood flow parametric images with N-13 ammonia PET IEEE Trans Med Imaging 1998 17 2 236 243 10.1109/42.700735 9688155 3. Kwan RK-S Evans AC Pike GB MRI simulation-based evaluation of image-processing and classification methods IEEE Trans Med Imaging 1999 18 11 1085 1097 10.1109/42.816072 10661326 4. Silverman DH Small GW Chang CY Lu CS Kung De Aburto MA Chen W Czernin J Rapoport SI Pietrini P Alexander GE Schapiro MB Jagust WJ Hoffman JM Welsh-Bohmer KA Alavi A Clark CM Salmon E de Leon MJ Mielke R Cummings JL Kowell AP Gambhir SS Hoh CK Phelps ME Positron emission tomography in evaluation of dementia: regional brain metabolism and long-term outcome JAMA 2001 17 286 2120 2127 10.1001/jama.286.17.2120 11694153 5. Shepp LA Vardi Y Maximum likelihood reconstruction for emission tomography IEEE Trans Med Imaging 1982 1 2 113 122 10.1109/TMI.1982.4307558 18238264 6. Hudson HM Larkin RS Accelerated image reconstruction using ordered subsets of projection data IEEE Trans Med Imaging 1994 13 4 601 609 10.1109/42.363108 18218538 7. Ma J Hudson M Block-iterative fisher scoring algorithms for maximum penalized likelihood image reconstruction in emission tomography IEEE Trans Med Imaging 2008 27 8 1130 1142 10.1109/TMI.2008.918355 18672430 8. Sitek A Reconstruction of emission tomography data using origin ensembles IEEE Trans Med Imaging 2011 30 4 946 956 10.1109/TMI.2010.2098036 21147594 9. Szirmay-Kalos L Magdics M Toth B Bukki T Averaging and metropolis iterations for positron emission tomography IEEE Trans Med Imaging 2013 32 3 589 600 10.1109/TMI.2012.2231693 23221817 10. Vaissier PEB Goorden MC Taylor AB Beekman FJ Fast count-regulated OSEM reconstruction with adaptive resolution recovery IEEE Trans Med Imaging 2013 32 12 2250 2261 10.1109/TMI.2013.2279851 23996543 11. Nuyts J Fessler JA A penalized-likelihood image reconstruction method for emission tomography, compared to postsmoothed maximum-likelihood with matched spatial resolution IEEE Trans Med Imaging 2003 22 9 1042 1052 10.1109/TMI.2003.816960 12956260 12. Chen J Tu K Chen T Lu H Liu R Chou K Chen C Cross-reference maximum likelihood reconstruction for positron emission tomography with empiric studies Eur J Nucl Med 2001 28 8S 1073 13. Nuyts J Baete K Beque D Dupont P Comparison between MAP and postprocessed ML for image reconstruction in emission tomography when anatomical knowledge is available IEEE Trans Med Imaging 2005 24 5 667 675 10.1109/TMI.2005.846850 15889553 14. Qi J Leahy RM Iterative reconstruction techniques in emission computed tomography Phys Med Biol 2006 51 15 541 578 10.1088/0031-9155/51/15/R01 15. Wack DS Snyder KV Seals KF Siddiqui AH Masked smoothing using separable kernels for ct perfusion images BMC Med Imaging 2014 14 25 10.1186/1471-2342-14-28 25063729 16. Tomasi C, Manduchi R. Bilateral filtering for gray and color images. In: Alley G, editors. Sixth international conference on computer vision, 1998. Bombay: IEEE; 1998. p. 839–46. 17. Hofheinz F Langner J Beuthien-Baumann B Oehme L Steinbach J Kotzerke J van den Hoff J Suitability of bilateral filtering for edge-preserving noise reduction in PET EJNMMI Res 2011 1 1 23 10.1186/2191-219X-1-23 22214263 18. Perona P Malik J Scale-space and edge detection using anisotropic diffusion IEEE Trans Pattern Anal Mach Intell 1990 12 7 629 639 10.1109/34.56205 19. Chan C Fulton R Feng DD Meikle S Regularized image reconstruction with an anatomically adaptive prior for positron emission tomography Phys Med Biol 2009 54 24 7379 7400 10.1088/0031-9155/54/24/009 19934490 20. Mumcuoğlu EU Leahy RM Cherry SR Bayesian reconstruction of PET images: methodology and performance analysis Phys Med Biol 1996 41 9 1777 1807 10.1088/0031-9155/41/9/015 8884912 21. Chan C Fulton R Barnett R Feng DD Meikle S Postreconstruction nonlocal means filtering of whole-body PET with an anatomical prior IEEE Trans Med Imaging 2014 33 3 636 650 10.1109/TMI.2013.2292881 24595339 22. Nuyts J Michel C Brepoels L De Ceuninck L Deroose C Goffin K Mottaghy FM Stroobants S Van Riet J Verscuren R Performance of MAP reconstruction for hot lesion detection in whole-body PET/CT: an evaluation with human and numerical observers IEEE Trans Med Imaging 2009 28 1 67 73 10.1109/TMI.2008.927349 19116189 23. Vunckx K Atre A Baete K Reilhac A Deroose CM Van Laere K Nuyts J Evaluation of three MRI-based anatomical priors for quantitative PET brain imaging IEEE Trans Med Imaging 2012 31 3 599 612 10.1109/TMI.2011.2173766 22049363 24. Somayajula S Panagiotou C Rangarajan A Li Q Arridge SR Leahy RM PET image reconstruction using information theoretic anatomical priors IEEE Trans Med Imaging 2011 30 3 537 549 10.1109/TMI.2010.2076827 20851790 25. Comtat C Kinahan PE Fessler JA Beyer T Townsend DW Defrise M Michel C Clinically feasible reconstruction of 3D whole-body PET/CT data using blurred anatomical labels Phys Med Biol 2002 47 1 1 20 10.1088/0031-9155/47/1/301 11814220 26. Stefan W Chen K Guo H Renaut RA Roudenko S Wavelet-based denoising of positron emission tomography scans J Sci Comput 2012 50 3 665 677 10.1007/s10915-011-9529-8 27. Shidahara M Ikoma Y Kershaw J Kimura Y Naganawa M Watabe H PET kinetic analysis: wavelet denoising of dynamic PET data with application to parametric imaging Ann Nucl Med 2007 21 7 379 386 10.1007/s12149-007-0044-9 17876550 28. Le Pogam A Hanzouli H Hatt M Cheze Le Rest C Visvikis D Denoising of PET images by combining wavelets and curvelets for improved preservation of resolution and quantitation Med Image Anal 2013 17 8 877 891 10.1016/j.media.2013.05.005 23837964 29. Bagci U Mollura DJ Denoising PET images using singular value thresholding and stein’s unbiased risk estimate Med Image Comput Comput Assist Interv 2013 16 Pt 3 115 122 24505751 30. Buades A Coll B Morel JM A review of image denoising algorithms, with a new one Multiscale Model Simul 2005 4 2 490 530 10.1137/040616024 31. Dabov K Foi A Katkovnik V Egiazarian K Image denoising by sparse 3-D transform-domain collaborative filtering IEEE Trans Image Process 2007 16 8 2080 2095 10.1109/TIP.2007.901238 17688213 32. Manjón JV Coupé P Buades A Louis Collins D Robles M New methods for MRI denoising based on sparseness and self-similarity Med Image Anal 2012 16 1 18 27 10.1016/j.media.2011.04.003 21570894 33. Manjón JV Coupé P Martí-Bonmatí L Collins DL Robles M Adaptive non-local means denoising of MR images with spatially varying noise levels J Magn Reson Imaging 2010 31 1 192 203 10.1002/jmri.22003 20027588 34. Aksam Iftikhar M Jalil A Rathore S Hussain M Robust brain MRI denoising and segmentation using enhanced non-local means algorithm Int J Imaging Syst Technol 2014 24 1 52 66 10.1002/ima.22079 35. Maggioni M Katkovnik V Egiazarian K Foi A Nonlocal transform-domain filter for volumetric data denoising and reconstruction IEEE Trans Image Process 2013 22 1 119 133 10.1109/TIP.2012.2210725 22868570 36. Rajwade A Rangarajan A Banerjee A Image denoising using the higher order singular value decomposition IEEE Trans Pattern Anal Mach Intell 2013 35 4 849 862 10.1109/TPAMI.2012.140 22732660 37. Zhang X Xu Z Jia N Yang W Feng Q Chen W Feng Y Denoising of 3D magnetic resonance images by using higher-order singular value decomposition Med Image Anal 2015 19 1 75 86 10.1016/j.media.2014.08.004 25291148 38. Lathauwer LD Signal processing based on multilinear algebra 1997 Leuven Katholike Universiteit Leuven 39. Donoho DL Johnstone IM Ideal spatial adaptation by wavelet shrinkage Biometrika 1994 81 3 425 455 10.1093/biomet/81.3.425 40. Anscombe FJ The transformation of poisson, binomial and negative-binomial data Biometrika 1948 35 3/4 246 254 10.2307/2332343 41. Makitalo M Foi A Optimal inversion of the Anscombe transformation in low-count poisson image denoising IEEE Trans Image Process 2011 20 1 99 109 10.1109/TIP.2010.2056693 20615809 42. National Electrical Manufacturers Association. NEMA standard publication NU 4-2008: performance measurements of small animal positron emission tomographs. Rosslyn, VA: National Electrical Manufacturers Association; 2008.

PMC005xxxxxx/PMC5002342.txt

==== Front Biomed Res IntBiomed Res IntBMRIBioMed Research International2314-61332314-6141Hindawi Publishing Corporation 10.1155/2016/3219068Research ArticleLiver Tumor Segmentation from MR Images Using 3D Fast Marching Algorithm and Single Hidden Layer Feedforward Neural Network http://orcid.org/0000-0003-1457-1819Le Trong-Ngoc 1 2 http://orcid.org/0000-0002-4847-4366Bao Pham The 3 http://orcid.org/0000-0002-2097-0704Huynh Hieu Trung 1 * 1Faculty of Information Technology, Industrial University of Ho Chi Minh City, 12 Nguyen Van Bao, Go Vap District, Ho Chi Minh City, Vietnam2Faculty of Information Technology, University of Science, 227 Nguyen Van Cu, District 5, Ho Chi Minh City, Vietnam3Faculty of Mathematics and Computer Science, University of Science, 227 Nguyen Van Cu, District 5, Ho Chi Minh City, Vietnam*Hieu Trung Huynh: hthieu@ieee.orgAcademic Editor: Yudong Cai 2016 14 8 2016 2016 32190689 5 2016 14 7 2016 19 7 2016 Copyright © 2016 Trong-Ngoc Le et al.2016This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Objective. Our objective is to develop a computerized scheme for liver tumor segmentation in MR images. Materials and Methods. Our proposed scheme consists of four main stages. Firstly, the region of interest (ROI) image which contains the liver tumor region in the T1-weighted MR image series was extracted by using seed points. The noise in this ROI image was reduced and the boundaries were enhanced. A 3D fast marching algorithm was applied to generate the initial labeled regions which are considered as teacher regions. A single hidden layer feedforward neural network (SLFN), which was trained by a noniterative algorithm, was employed to classify the unlabeled voxels. Finally, the postprocessing stage was applied to extract and refine the liver tumor boundaries. The liver tumors determined by our scheme were compared with those manually traced by a radiologist, used as the “ground truth.” Results. The study was evaluated on two datasets of 25 tumors from 16 patients. The proposed scheme obtained the mean volumetric overlap error of 27.43% and the mean percentage volume error of 15.73%. The mean of the average surface distance, the root mean square surface distance, and the maximal surface distance were 0.58 mm, 1.20 mm, and 6.29 mm, respectively. Vietnam National Foundation for Science and Technology Development (NAFOSTED)102.01-2013.47 ==== Body 1. Introduction The hepatic cell carcinoma (HCC) is one of the most common cancers and rapidly growing worldwide. It was estimated that up to 90% of patients with liver cancer would be dead within five years of diagnosis [1]. The early detection and the treatment response evaluation of patients with the liver cancer are very important to improve the survival rate. The basic criterion to evaluate the tumor response to the treatment is the tumor size. There are some criteria for tumor size including one-, bi-, or tridimensional measurement [2–5]. However, clinical researches have shown that the volume measurement (3D) can give the best reflection of the tumor response [3–5]. The liver tumor volumetry requires the tumor segmentation. Traditionally, this task can be performed by manually tracing the tumor regions on slices. It is tedious and time-consuming. In addition, the volume of manual delineations is subjective; it was estimated that the intra- and interobserver variability are about 8% for liver tumors [6]. Hence, it is crucial to investigate in the computerized scheme for liver tumor segmentation. Some computerized schemes have been developed for the liver tumor segmentation on CT images. These methods include the watershed like the paintbrush algorithm [7], deformable models [8], and level-set and active contour techniques. Multiple thresholding and adaptive techniques for liver tumor segmentation from CT images also were proposed by researchers [9–11]. Wong et al. [12] proposed a region growing technique based on knowledge-constraints for liver tumor segmentation in each slice. This technique demonstrated an average volume difference of 24.2% in tumor segmentation. Ben-Dan and Shenhav [13] developed an approach for liver tumor segmentation based on the active contour using a weighted function of the probability of each pixel. An approach for liver tumor segmentation based on Bayesian classification and the active contour method was proposed by Taieb et al. [14]. This approach can obtain an average volume difference of 44.2%. Smeets et al. [15] proposed an approach based on the combination of the level-set method and the spiral-scanning technique. In this method, users place a point approximately in the middle of the liver tumor and specify a maximal radius by placing other points surrounding the tumor. Next, the spiral-scanning technique was applied to generate the initial surface for the level-set segmentation. A method for the liver tumor segmentation by applying the minimal surfaces and Markov random fields was developed by Stawiaski et al. [16]. Some techniques using machine learning methods also were developed [17–21]. Li et al. [17] developed a technique for liver tumor segmentation based on machine learning using the intensity profiles of the liver tumors. This method is less accurate for segmenting the tumors with irregular boundaries due to biasing to the blob-like ones. Zhou et al. [18] developed a scheme, in which the tumor region from a single slice was extracted by using support vector machine (SVM). This region was propagated and applied to classify voxels in other slices which contain the tumor. A scheme for 3D liver tumor segmentation based on the random walker has been proposed by Jolly and Grady [19]. The user-defined seed point is required, and the additional seed points are generated from 2D fuzzy-connectedness segmentation of a slice containing the seed point. Freiman et al. [20] developed an approach for liver tumor segmentation from CTA using SVM, in which the features include the mean, the standard deviation, and the minimum and the maximum intensity values in a predefined window surrounding the voxel. Huang et al. [21] demonstrated an approach for liver tumor segmentation based on extreme learning machine. The features are generated from the mean, variance, intensity, intensity power, entropy, intensity cooccurrence, Law's texture, and sum and difference histogram. In comparison with CT images, the number of approaches for liver tumor segmentation on MR images is limited, while the increasing use of liver MRI as a single exam for liver disease leads to imperative demands for investigating researches in automatic MRI liver tumor volumetry. Developing the computerized scheme for liver tumor segmentation in MR images is a challenging task. The images have a low gradient response. The liver tumors generally have different shapes. The gray values of tumor depend on several factors including the tumor type, the image acquisition, and the contrast injection. In this study, we focus on a very challenging task of the liver tumor segmentation in abdominal MR images. The proposed method utilizes the local information in the liver; it combines the 3D fast marching algorithm and the neural network. The region of interest (ROI) surrounding the tumor is established to improve the performance of generating the initial regions as well as reduce the computational requirements. 2. Materials and Methods Our proposed scheme for liver tumor segmentation is described in Figure 1. It consists of four main stages: preprocessing, generating the initial labeled regions, classifying the unlabeled voxels using the single hidden layer feedforward neural network (SLFN), and postprocessing. 2.1. Preprocessing Stage Firstly, users can choose the seed points inside and outside the tumor. The 3D region of interest (ROI), I R 0(x, y, z), involving the tumor is determined based on the seed points. This ROI is passed to reduce noise and enhance the liver tumor structures by using an anisotropic diffusion algorithm. This algorithm is handled by a modified curvature diffusion equation given by (1) ∂IR∂t=∇IR∇·c∇IR∇IR∇IR, where I R(·, t) is the image function with the initial image at t = 0 given by I R(·, t = 0) = I R 0(·) and c(·) is the diffusion coefficient which controls the sensitivity of the edge contrast. This algorithm smooths noise in the image while preserving the major structures including the boundaries. The noise-reduced image is then passed through a Gaussian gradient magnitude filter to generate I G which enhances the boundaries. The gradient magnitude image is determined by (2) IM=∂IG∂x2+∂IG∂y2+∂IG∂z2. The gradient magnitude image is then employed to produce the edge potential image by using a sigmoid function given by (3) IP=11+e−IM−β/α, where α and β are parameters specifying the range and center, respectively, of the intensity to be enhanced. In this scheme, α and β were −3.5 and 8.0, respectively. The edge potential image is used as a speed function for a fast marching algorithm to generate the initial labeled regions. 2.2. Generating the Initial Labeled Regions The main goal in this stage is to generate the labeled regions or teacher regions by using a fast marching algorithm. This algorithm is based on the numerical solution of the Eikonal equation given by (4) ∇TF=1, where T is an arrival time function and F is a speed function which is obtained from the potential image, I P. The algorithm requires the initial seed points which correspond to the initial location (T = 0). These seed points (1-2 points inside the tumor and 1–3 points outside the tumor) in our scheme are placed by a radiologist. In the first step of the algorithm, the grid points from the entire image are categorized into three classes. The initial seed points are classed as Known. The neighbors of the Known points are classed as Trial, and their arrival time is computed by using the first-order scheme (see (4)). Other points are classified as Far with the infinity arrival time. The algorithm propagates the information in one way from the smaller values to the larger values of T. The point q with the smallest arrival time in the Trial list is chosen and moved to the Known one. The neighbors of q are moved to the Trial list and the arrival time is recomputed by using the first-order scheme. This iterative process is terminated when the maximum number of iterations is met. The salient point of this algorithm is that the heap data structure is used to speed up locating the Trial point with the smallest T value. The output of the fast marching algorithm is a time-crossing map image which indicates the time travelling to each point. The labeled regions are generated by (5) where M F is the maximum number of iterations for the fast marching algorithm and I TM is the time-crossing map image. The points with the values of 1 or 2 in the image I S are considered to be labeled, otherwise considered to be unlabeled because they were not treated from the fast marching algorithm. Due to variations of intensity in tumors, the labeled regions could not form the rough shape of the liver tumor. However, they can be considered as teacher regions for training the neural network to classify the unlabeled points. 2.3. Classifying the Unlabeled Voxels The labeled points are categorized into two classes, tumor or nontumor corresponding to the labels of 1 or 2, respectively. Now, we have to classify the unlabeled points which are not treated by the fast marching algorithm. The neural network is one of the powerful tools in biology and medical data analysis applications [22, 23]. Hence, in this study we utilized the neural network to categorize the unlabeled points. Several network architectures have been developed; however a single hidden layer feedforward neural network (SLFN) can form the decision boundaries with arbitrary shapes if the activation function is chosen properly. A typical architecture of SLFN consists of an input layer, a hidden layer with K units, and an output layer with M units. In this study, the SLFN handles the input gray levels directly. The inputs to the SLFN comprise the normalized voxel in ROI image (I R) and its spatially adjacent normalized voxels within the local window L W. An input pattern corresponding to I R(x, y, z) is determined by (6) pxyz=−1+2IRx−i,y−j,z−k−IminImax−Imin ∣ i,j,k∈LW, where I min, I max are the minimum and maximum voxel values, respectively, in I R. The output of the SLFN which corresponds to the center voxel in the local window is the value indicating either tumor or nontumor, represented by (7) oxyz=∑k=1Kakφpxyz·wk+bk, where w k is the input weights connecting from the input layer to the kth hidden unit, b k is its bias, a k = [a k1, a k2,…,a kM]T is the output weights connecting from the kth hidden unit to the output layer, and φ(·) is the activation function of the hidden layer which was a sigmoidal function in this study. The activation function of the output layer was an identify function. One of the important issues in the SLFN is to train the network to determine the network weights w, b, and a. Traditionally, training the neural network could be performed by employing the backpropagation algorithm. This algorithm has some limitations such as slow convergence, local minima, overfitting, or improper learning rate. Although there are several improvements to overcome the problems of backpropagation algorithm, up to now, the training algorithms based on the gradient descent approach are still slow due to many iterative steps. In this study, we employed an effective training algorithm for SLFN called extreme learning machine (ELM) and its improvements [24–27]. The desired output corresponding to the input pattern p xyz is obtained from the teacher regions in I S, represented by (8) txyz=01Tif  ISx,y,z=110Tif  ISx,y,z=2. Assume that there are n labeled points in the teacher regions. The training set is given by S = {(p i, t i), i = 1,2,…, n}, where the input patterns p i are obtained from I R by using (6) and its corresponding desired output, t i, is obtained from the teacher regions in I S by using (8). The error function to be minimized by training process is defined by (9) E=∑i=1noi−ti, where o i is the actual output corresponding to the input pattern p i (by using (7)). In the extreme learning machine, the training process can be modeled by finding the solution of a linear model given by H A = T, where H is the hidden layer output matrix of SLFN defined by (10) H=φw1·p1+b1⋯φwK·p1+bK⋮⋱⋮φw1·pn+b1⋯φwK·pn+bK,T=t1t2⋯tnT,A=a1a2⋯aKT. In the extreme learning machine, the biases and input weights are assigned by the random values, and the output weights are found by using the Moore-Penrose generalized inverse: (11) A⌢=H†T, where H † is the pseudoinverse of H. Several improvements of ELM have been developed by researchers [26, 27]. In order to make the system more stable, the regularization approaches have been proposed, in which a coefficient λ is added and the solution of A is given by (12) A⌢=HTH+λI−1HTT. These training algorithms are fast and can offer a good performance in many applications including medical data analysis. 2.4. Postprocessing The classifying results from the SLFN for unlabeled regions are combined with the ones from the fast marching algorithm. A voxel belongs to tumor region if it is classified as tumor by either the fast marching algorithm or the SLFN. The combined results are then passed to the connected component and the relabeling filters. The region containing the tumor seed points is extracted and passed to remove small isolated artifacts and small holes by using the morphological operations. The liver tumor volume is calculated from the segmented tumor regions. 3. Experimental Results 3.1. Datasets This study was approved by the institutional review board (IRB) of the Medic Medical Center. The datasets consist of 25 liver tumors which were obtained from 16 patients. Fifteen tumors were obtained from 10 patients by using the 1.5 T magnetic resonance imaging (MRI) scanners (Avanto, Siemens) at the Medic Medical Center, which is one of the largest diagnostic imaging centers in Vietnam. Informed consent was obtained from all patients. Postcontrast MR images were obtained by using the T1-weight volumetric interpolated breath-hold examination (VIBE) sequence. A flip angle of 10 degrees was used with the TE and TR of 2.38 and 4.74, respectively. The scanning parameters included collimation and reconstruction intervals ranging from 3.5 to 4 mm. Each MRI slice had a matrix size of 230 × 320 pixels with an in-plane pixel size ranging from 1.18 to 1.4 mm. The number of slices in each case ranged from 44 to 56. Ten other tumors were obtained from 6 patient cases which were extracted from The Cancer Imaging Archive (TCIA) [28]. A board-certified abdominal radiologist carefully manually traced the tumor contours on each slice which contains the liver tumor. The liver tumor volume was calculated by multiplying the areas of the manually traced regions in each slice by the reconstruction interval. The total tumor liver volume in each case was determined by the summation of the volumes in all of the slices. The times required to complete the manual contour tracing were recorded. 3.2. Evaluation Criteria The liver tumor volumes obtained by our computerized approach were compared to the “ground-truth” manual volumes which were determined by the radiologist. The true positive (TP), false positive (FP), and false negative were calculated for detail analysis. The percentage volume error (E) for each computerized volume (V c) and the gold standard manual volume (V m) is determined by (13) E=Vc−VmVm. The volumetric overlap error representing the fraction of the overlapping volume and the volume of two segmentation methods is given by (14) VO=1−TPTP+FP+FN. Three surface distance metrics including the average symmetric surface distance (ASSD), the root mean square symmetric surface distance (RMSSD), and the maximal symmetric surface distance (MSSD) also were used to evaluate the global and the local disagreement between two segmentation methods. 3.3. Results and Discussions The intermediate results of the proposed scheme for an example case were shown in Figure 2. An original 3D MR image was shown in Figure 2(a). The 3D region of interest (ROI) which was extracted from the original 3D MR image was shown in Figure 2(b). This ROI image was passed to reduce the noise by using an anisotropic diffusion algorithm. A gradient magnitude filter was applied to the reduced-noise image to enhance the boundaries and create an edge image, as shown in Figure 2(c). The edge potential image was generated, and then the fast marching algorithm and thresholding filter were applied to generate the image with the labeled regions, as shown in Figure 2(d). The labeled regions were employed to train the SLFN; this trained network was then used to classify voxels in the unlabeled regions as shown in Figure 2(e). The classification result from SLFN was combined with the one from the fast marching algorithm. The connected component and relabeling operations were applied, and the region which contains the seed points inside tumor was filtered out as shown in Figure 2(f). A comparison between the computerized liver tumor segmentation (black contour) and the “ground-truth” manual liver tumor segmentation (white contour) was illustrated in Figure 2(g). The liver tumor volume was computed from the segmented regions. The average liver tumor volumes measured from two methods for two datasets, Medic Medical Center and TCIA, were shown in Table 1. For both datasets, the average tumor volume of the reference standard manual method was 15.36 ± 36.77 cm3 (range: 0.19–162.89 cm3), whereas the average tumor volume of the proposed computerized scheme was 13.88 ± 33.21 cm3 (range: 0.10–147.95 cm3). A comparison of the liver tumor segmentation on MR image between two methods was shown in Table 2. For both datasets, the overall mean of volumetric overlap error and the mean percentage volume error were 27.43% and 15.74%, respectively. The mean of average surface distance, root mean square surface distance, and maximal surface distance were 0.58 mm, 1.20 mm, and 6.29 mm, respectively. It is not easy to directly compare the proposed method with existing methods in literature because of using different databases and quality measurements. Freiman et al. [20] evaluated the performance of their proposed scheme on CTA images, and it could obtain a volume overlap error and a volume difference of 33.8% and 22.6%, respectively. The mean of average surface distance, the RMS surface distance, and the maximal surface distance were 1.76 mm, 2.62 mm, and 13.73 mm, respectively. Smeets et al. [15] obtained the overlap error and the volume difference of 32.64% and 17.91%, respectively. The average, the RMS, and the maximum surface distance were 1.97 mm, 2.64 mm, and 10.13 mm, respectively. The method proposed by Huang et al. [21] could achieve the volume overlap error, the volume difference, the average surface distance, the RMS surface distance, and the maximal surface distance of 67.15%, 14.16%, 2.27 mm, 2.47 mm, and 8.46 mm, respectively; the performance of this scheme was evaluated on CT images. There are several parameters to be adjusted in our proposed scheme. They were determined by empirical analysis. The number of hidden nodes in the SLFN was given by K = 200 and λ = 0.01; the local window size for extracting features (L W) was set to 3 × 3 × 3. The number of iterations for the fast marching algorithm was set to 15. All the parameters were fixed for all patients. Some advanced schemes have applied the statistical models, used the intensity profiles of liver tumors, or used filters for the tumor segmentation. However, the actual tumors often differ from the simple models. There are tumors of various shapes and inhomogeneous tumors. In this study, the artificial neural network (ANN) was developed for accommodating the task of distinguishing the tumor voxels from the nontumor voxels. The network was trained by the subregions of the 3D MR image and operates on the voxel data directly, which is similar to the MTANN (massive training artificial neural network) [29]. However, unlike the original MTANN, the ANN proposed in this study was trained by the data from the 3D local information of each tumor, and the network was trained by a noniterative training algorithm which can overcome problems of the traditional training algorithms such as local minima, learning rate, and epochs. This training algorithm can offer a good performance with high learning speed in many different applications. A limitation in this study is the number of patients. Our proposed scheme was evaluated in 16 patients with 25 tumors while other studies evaluated 5 patients with 10 tumors [20], 7 patients with 10 tumors [15], and 10 patients with 10 lesions [30]. In general, a small number of tumors and patients may limit the variations among them. In the future, we will need to increase the number of patients and tumors used for evaluation. 4. Conclusion The increasing use of liver MRI as a single exam leads to imperative demands for investigating researches in the computerized MRI liver tumor segmentation. However, few studies have been reported for this challenging task. In this study, we developed a computerized scheme for liver tumor segmentation on MR images by employing the fast marching algorithm and the neural network, in which the neural network was trained by an effectively noniterative algorithm. The performance was evaluated in 25 tumors of 16 patients. Our experimental results have shown that the proposed method is accurate and efficient when compared to the manual ground-truth segmentation. Its accuracy also is comparable or better than the existing semiautomatic methods. With our proposed scheme, the time required for liver tumor segmentation is reduced significantly. Hence, it can be useful for radiologists for the liver tumor analysis on MR images. Acknowledgments This research is funded by the Vietnam National Foundation for Science and Technology Development (NAFOSTED) under Grant no. 102.01-2013.47. The authors are grateful to Phan Thanh Hai, MD, and Nguyen Thanh Dang, MD, for preparing images and for their clinical advice. Ethical Approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This paper does not contain any studies with animals performed by any of the authors. Consent Informed consent was obtained from all individual participants included in the study. Competing Interests The authors declare that they have no competing interests. Figure 1 Overview of the proposed scheme for liver tumor segmentation. Figure 2 The intermediate results of the proposed scheme. (a) A slice of the original 3D image. (b) A slice of the 3D region of interest containing the liver tumor which was extracted from the original 3D MR image. (c) The edge image generated by applying the gradient magnitude filter. (d) The labeled regions generated by the fast marching algorithm and thresholding filter. (e) Unlabeled voxels were classified by using the SLFN. (f) The segmented liver tumor. (g) A comparison between the computerized liver tumor segmentation (black contour) and the “ground-truth” manual liver tumor segmentation (white contour). Table 1 Comparison of liver tumor volume between the computerized method and the manual method. Dataset Volume Manual method (cc) Computerized method (cc) Medic Medical Center Average 3.48 3.16 SD 4.00 3.97 Min 0.23 0.14 Max 16.10 16.18 TCIA Average 33.18 29.95 SD 54.69 49.38 Min 0.19 0.10 Max 162.89 147.95 Both Average 15.36 13.88 SD 36.77 33.21 Min 0.19 0.10 Max 162.89 147.95 Table 2 Summary of the comparison results. Dataset Evaluation measure Mean SD Min Max Medic Medical Center Volumetric overlap error (%) 26.66 7.06 15.70 39.47 Percentage volume error (%) 16.68 12.51 0.17 39.47 Average surface distance (mm) 0.44 0.47 0.21 2.12 RMS surface distance (mm) 0.97 0.97 0.53 4.40 Maximal surface distance (mm) 4.84 4.89 1.68 21.18 TCIA Volumetric overlap error (%) 28.57 10.89 13.71 47.89 Percentage volume error (%) 14.32 15.81 0.21 47.89 Average surface distance (mm) 0.79 0.73 0.14 2.66 RMS surface distance (mm) 1.55 1.06 0.35 3.90 Maximal surface distance (mm) 8.46 6.45 1.61 19.60 Both Volumetric overlap error (%) 27.43 8.63 13.71 47.89 Percentage volume error (%) 15.74 13.65 0.17 47.89 Average surface distance (mm) 0.58 0.60 0.14 2.66 RMS surface distance (mm) 1.20 1.03 0.35 4.40 Maximal surface distance (mm) 6.29 5.73 1.61 21.18 ==== Refs 1 Primary liver cancer, Celsion corporation, 2016, http://celsion.com/docs/pipeline_liver_cancer 2 Therasse P. Arbuck S. G. Eisenhauer E. A. New guidelines to evaluate the response to treatment in solid tumors Journal of the National Cancer Institute 2000 92 3 205 216 10.1093/jnci/92.3.205 2-s2.0-0034594628 10655437 3 Tran L. N. Brown M. S. Goldin J. G. Comparison of treatment response classifications between unidimensional, bidimensional, and volumetric measurements of metastatic lung lesions on chest computed tomography Academic Radiology 2004 11 12 1355 1360 10.1016/j.acra.2004.09.004 2-s2.0-10444278156 15596373 4 Tuma R. S. Sometimes size doesn't matter: reevaluating RECIST and tumor response rate endpoints Journal of the National Cancer Institute 2006 98 18 1272 1274 10.1093/jnci/djj403 2-s2.0-33749029255 16985244 5 Prasad S. R. Jhaveri K. S. Saini S. Hahn P. F. Halpern E. F. Sumner J. E. CT tumor measurement for therapeutic response assessment: comparison of unidimensional, bidimensional, and volumetric techniques—Initial observations Radiology 2002 225 2 416 419 10.1148/radiol.2252011604 2-s2.0-0036828529 12409574 6 Bellon E. Feron M. Maes F. Evaluation of manual vs. semi-automated delineation of liver lesions on CT images European Radiology 1997 7 3 432 438 10.1007/s003300050180 2-s2.0-0030633139 9087371 7 Maes F. Vandermeulen D. Suetens P. Marchal G. Automatic image partitioning for generic object segmentation in medical images Computational Imaging and Vision 1995 3 215 226 8 Yim P. J. Foran D. J. Volumetry of hepatic metastases in computed tomography using the watershed and active contour algorithms Proceedings of the 16th IEEE Symposium on Computer Based Medical Systems June 2003 New York, NY, USA 329 335 2-s2.0-0042090591 9 Moltz J. Bornemann L. Dicken V. Peitgen H. Segmentation of liver metastases in CT scans by adaptive thresholding and morphological processing The MIDAS Journal—Grand Challenge Liver Tumor Segmentation 2008 1 10 Proceedings of the 2008 MICCAI Workshop 10 Abdel-Massieh N. H. Hadhoud M. M. Amin K. M. Fully automatic liver tumor segmentation from abdominal CT scans Proceedings of the 6th International Conference on Computer Engineering and Systems (ICCES '10) December 2010 Cairo, Egypt IEEE 197 202 10.1109/icces.2010.5674853 2-s2.0-78751631373 11 Choudhary A. Moretto N. Ferrarese F. P. Zamboni G. A. An entropy based multi-thresholding method for semi-automatic segmentation of liver tumors The MIDAS Journal-Grand Challenge Liver Tumor Segmentation (2008 MICCAI Workshop) 2008 41 43 43 49 12 Wong D. Liu J. Yin F. A semi-automated method for liver tumor segmentation based on 2D region growing with knowledge-based constraints The MIDAS Journal-Grand Challenge Liver Tumor Segmentation (2008 MICCAI Workshop) 2008 41 43 p. 159 13 Ben-Dan I. Shenhav E. Liver tumor segmentation in CT images using probabilistic The MIDAS Journal—Grand Challenge Liver Tumor Segmentation 2008 1 6 Proceedings of the MICCAI Workshop, 2008 14 Taieb Y. Eliassaf O. Freiman M. Joskowicz L. Sosna J. An iterative Bayesian approach for liver analysis: tumor validation study The MIDAS Journal—Grand Challenge Liver Tumor Segmentation (2008 MICCAI Workshop) 2008 1 6 15 Smeets D. Loeckx D. Stijnen B. De Dobbelaer B. Vandermeulen D. Suetens P. Semi-automatic level set segmentation of liver tumors combining a spiral-scanning technique with supervised fuzzy pixel classification Medical Image Analysis 2010 14 1 13 20 10.1016/j.media.2009.09.002 2-s2.0-70549098227 19828356 16 Stawiaski J. Decenciere E. Bidault F. Interactive liver tumor segmentation using graph-cuts and watershed The MIDAS Journal—Grand Challenge Liver Tumor Segmentation (MICCAI Workshop) 2008 1 6 17 Li Y. Hara S. Shimura K. A machine learning approach for locating boundaries of liver tumors in CT images Proceedings of the 18th International Conference on Pattern Recognition (ICPR '06) August 2006 Hong Kong 400 403 10.1109/icpr.2006.93 2-s2.0-34047237346 18 Zhou J. Xiong W. Tian Q. Semi-automatic segmentation of 3D liver tumors from CT scans using voxel classification and propagational learning The MIDAS Journal—Grand Challenge Liver Tumor Segmentation (MICCAI Workshop) 2008 1 10 19 Jolly M.-P. Grady L. 3D general lesion segmentation in CT Proceedings of the 5th IEEE International Symposium on Biomedical Imaging: From Nano to Macro (ISBI '08) May 2008 Paris, France IEEE 796 799 10.1109/isbi.2008.4541116 2-s2.0-51049105255 20 Freiman M. Cooper O. Lischinski D. Joskowicz L. Liver tumors segmentation from CTA images using voxels classification and affinity constraint propagation International Journal of Computer Assisted Radiology and Surgery 2010 6 2 247 255 20574799 21 Huang W. Li N. Lin Z. Liver tumor detection and segmentation using kernel-based extreme learning machine Proceedings of the 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC '13) July 2013 Osaka, Japan 3662 3665 10.1109/embc.2013.6610337 2-s2.0-84886513922 22 Huynh H. T. Kim J.-J. Won Y. Performance comparison of SLFN training algorithms for DNA microarray classification Advances in Experimental Medicine and Biology 2011 696 135 143 10.1007/978-1-4419-7046-6_14 2-s2.0-79957979361 21431554 23 Huynh H. T. Won Y. Kim J. Hematocrit estimation using online sequential extreme learning machine Bio-Medical Materials and Engineering 2015 26 supplement 1 S2025 S2032 10.3233/bme-151507 26405979 24 Huang G.-B. Zhu Q.-Y. Siew C.-K. Extreme learning machine: theory and applications Neurocomputing 2006 70 1–3 489 501 10.1016/j.neucom.2005.12.126 2-s2.0-33745903481 25 Liang N.-Y. Huang G.-B. Saratchandran P. Sundararajan N. A fast and accurate online sequential learning algorithm for feedforward networks IEEE Transactions on Neural Networks 2006 17 6 1411 1423 10.1109/TNN.2006.880583 2-s2.0-34047174077 17131657 26 Huynh H. T. Won Y. Kim J.-J. An improvement of extreme learning machine for compact single-hidden-layer feedforward neural networks International Journal of Neural Systems 2008 18 5 433 441 10.1142/S0129065708001695 2-s2.0-55749087656 18991365 27 Huynh H. T. Won Y. Small number of hidden units for ELM with two-stage linear model IEICE Transactions on Information and Systems 2008 E91-D 4 1042 1049 10.1093/ietisy/e91-d.4.1042 2-s2.0-55749110008 28 National Cancer Institute Cancer Imaging Program, Frederick National Laboratory for Cancer Research the Cancer Imaging Archive (TCIA) 2013, http://cancerimagingarchive.net/ 29 Suzuki K. Doi K. How can a massive training artificial neural network (MTANN) be trained with a small number of cases in the distinction between nodules and vessels in thoracic CT? Academic Radiology 2005 12 10 1333 1341 10.1016/j.acra.2005.06.017 2-s2.0-25144467873 16179210 30 Kumar S. S. Moni R. S. Rajeesh J. Automatic liver and lesion segmentation: a primary step in diagnosis of liver diseases Signal, Image and Video Processing 2013 7 1 163 172 10.1007/s11760-011-0223-y 2-s2.0-84871921854

PMC005xxxxxx/PMC5002396.txt

==== Front Mol Cell PediatrMol Cell PediatrMolecular and Cellular Pediatrics2194-7791Springer Berlin Heidelberg Berlin/Heidelberg 275108966110.1186/s40348-016-0061-7Short CommunicationTransient spontaneous remission in congenital MLL-AF10 rearranged acute myeloid leukemia presenting with cardiorespiratory failure and meconium ileus Gyárfás Tobias tobias.gyarfas@med.uni-duesseldorf.de 1Wintgens Juergen juergen.wintgens@sk-mg.de 2Biskup Wolfgang wolfgang.biskup@sk-mg.de 2Oschlies Ilske ioschlies@path.uni-kiel.de 3Klapper Wolfram wklapper@path.uni-kiel.de 3Siebert Reiner rsiebert@medgen.uni-kiel.de 4Bens Susanne sbens@medgen.uni-kiel.de 4Haferlach Claudia claudia.haferlach@mll.com 5Meisel Roland roland.meisel@med.uni-duesseldorf.de 6Kuhlen Michaela +49 211 81 16491/04982michaela.kuhlen@med.uni-duesseldorf.de 6Borkhardt Arndt arndt.borkhardt@med.uni-duesseldorf.de 61 Medical Faculty, Department of General Pediatrics, Neonatology and Pediatric Cardiology, Centre for Child and Adolescent Health, University of Duesseldorf, Duesseldorf, Germany 2 Department of Neonatology, Staedtische Kliniken Moenchengladbach, Elisabeth Krankenhaus Rheydt, Rheydt, Germany 3 Department of Pathology, Christian-Albrechts-University Kiel, Kiel, Germany 4 Institute of Human Genetics, Christian-Albrechts-University Kiel, Kiel, Germany 5 MLL Munich Leukemia Laboratory, Munich, Germany 6 Department of Pediatric Oncology, Hematology and Clinical Immunology, University Children’s Hospital, Medical Faculty, Heinrich Heine University, Moorenstr. 5, 40225 Duesseldorf, Germany 29 8 2016 29 8 2016 12 2016 3 1 3027 3 2016 3 8 2016 © The Author(s). 2016 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.Background Neonatal leukemia is a rare disease with an estimated prevalence of about one to five in a million neonates. The majority being acute myeloid leukemia (AML), neonatal leukemia can present with a variety of symptoms including hyperleucocytosis, cytopenia, hepatosplenomegaly, and skin infiltrates. Chromosomal rearrangements including mixed lineage leukemia (MLL) translocations are common in neonatal AML. Case presentation A female neonate born at 34 weeks gestation presented with cardiorespiratory failure, hepatosplenomegaly, pancytopenia, and coagulopathy. She required intensive care treatment including mechanical ventilation, high-dose catecholamine therapy, and multiple transfusions. Small intestinal biopsy obtained during laparotomy for meconium ileus revealed an infiltrate by an undifferentiated monoblastic, MLL-rearranged leukemia. No other manifestations of leukemia could be detected. After spontaneous clinical remission, lasting 5 months without any specific treatment, the patient presented with leukemia cutis and full-blown monoblastic leukemia. MLL-AF10-rearranged AML could be re-diagnosed and successfully treated with chemotherapy and hematopoietic stem cell transplantation. Conclusions Our patient exhibited a unique manifestation of neonatal MLL-AF10 rearranged AML with cardiorespiratory failure and intestinal infiltration. It highlights the importance of leukemia in the differential diagnosis of neonatal distress, congenital hematological abnormalities, and skin lesions. Keywords Acute myeloid leukemiaCongenitalMeconium ileusCardiorespiratory failureSpontaneous remissionCase reportissue-copyright-statement© The Author(s) 2016 ==== Body Correspondence/findings Background Acute myeloid leukemia (AML) accounts for about 18 % of pediatric leukemias [1]. Disease onset during the neonatal period is extremely rare, with an estimated prevalence of one to five in a million neonates. The majority of neonatal leukemias are AML, and the most common type is monoblastic leukemia (FAB M5). Typical clinical features are skin infiltrations, hepatosplenomegaly, and cardiac failure. Prenatal symptoms including non-immune hydrops and polyhydramnios illustrate the in utero origin of the disease [2]. Like in infant leukemias, chromosomal rearrangements are regularly found in neonatal leukemias [2, 3] with chromosome 11q23/MLL rearrangements being by far the most common ones. [4]. Case report We present the case of a 14-month-old girl from birth until today. Pregnancy was uneventful until 34 weeks gestation, when polyhydramnios, pathological cerebral Doppler flow, pleural effusions, and pathological cardiotocography led to performing a cesarean section. The girl was born pale, with no muscle tone, no spontaneous breathing, and bradycardia. The APGAR score was 2/6/7 at 1/5/10 min. Persisting respiratory insufficiency required maximally invasive treatment including the administration of endotracheal surfactant, the use of a high-frequency oscillation ventilator and nitrogen monoxide to maintain sufficient oxygenation. Continuous dobutamine and adrenaline infusions were necessary due to severe circulatory deterioration starting immediately after birth. Echocardiogram revealed pulmonary hypertension and compromised right ventricular function. Massive hepatosplenomegaly was detectable on abdominal ultrasound. An initial hemoglobin value of 6.3 g/dl led to an emergency erythrocyte transfusion, and further erythrocyte transfusions were necessary in the first weeks of life. Multiple petechiae and suggilations were present at birth, thrombocytopenia and disseminated intravascular coagulopathy with unmeasurable thrombin time repeatedly required multiple thrombocyte concentrate and fresh-frozen plasma transfusions. An initial leukocyte count of 45.000/μl rapidly dropped to a minimum of 600/μl. The etiology of the clinical condition with cardiorespiratory failure, pancytopenia, and disseminated intravascular coagulopathy together with skin hemorrhages and hepatosplenomegaly was at first elusive. Ganciclovir was initiated for suspected cytomegalovirus infection but withdrawn following negative results. Given the possibility of neonatal sepsis, various antibiotics and fluconazole were administered, but inflammatory values did not markedly increase and microbiological cultures and extensive viral PCR studies yielded exclusively negative results. Peripheral blood smears from the fourth day of life revealed immature monocytic cells, and bone marrow puncture was found unrepresentative without evidence of blastic cells. Additional flow cytometry was unremarkable. Genetic predisposition for hemophagocytic lymphohistiocytosis was excluded. The girl’s clinical condition remained highly instable, when on the tenth day of life the assessment of acute abdomen required explorative laparotomy. During surgery, a meconium ileus was found that could only be resolved by establishment of an ileostomy. Because of a massively distended abdomen, abdominal closure could only be realized by placement of a patch. Buccal mucosa and ileum biopsies were undertaken to exclude neonatal hemochromatosis. Unexpectedly, pathological examination revealed diffuse submucosal infiltration by polymorphous, atypical cells suspicious of Langerhans cell histiocytosis (LCH). Single-shot treatment with vincristine (0.05 mg/kg) and prednisolone (1.5 mg/kg/d) for suspected LCH was immediately commenced. Five days later, the LCH diagnosis was rejected by reference pathology evaluation and reclassified as completely undifferentiated blastic transmural infiltration with diffuse growth pattern, positive staining for CD68, partial expression of S100, negative staining for langerin, lysozyme, CD1a, myeloperoxidase, CD34 ,and CD30, and an increased proliferation rate of 80 % (MIB1) (Fig. 1).Fig. 1 Submucosal infiltrate of polymorphic, atypical cells in the terminal ileum. Top row: ×25 magnification, bottom row: ×400 magnification. Left: hematoxylin and eosin (H&E) stain. Middle: KI67 staining shows a proliferation rate of 80 %. Right: the majority of cells display positive staining for the monocyctic marker CD68. A clonal MLL breakpoint in 11q23 was detected by FISH on paraffin embedded tissue-slides in the cells of the intestinal infiltrate. The population was classified as an extramedullary manifestation of monoblastic acute myeloid leukemia (AML FAB M5) To further classify the blastic infiltration, additional immuno-histochemical staining was performed and revealed partial expression of CD163, negativity for CD56, CD117, CD3, CD79a, desmin, NB84, pan-cytokeratin, synaptophysin, and TdT, and preserved nuclear expression of INI1, thus being consistent with an undifferentiated leukemia with partial monoblastic differentiation. Further, a chromosomal breakpoint at the mixed lineage leukemia (MLL) locus 11q23 could be detected by fluorescence in situ hybridization (FISH, LSI MLL Dual Color, Break Apart Probe, Abbott) substantiating the diagnosis of an undifferentiated, MLL positive leukemia. Two subsequent bone marrow punctures at the age of 3 and 5 weeks did not show any signs of leukemia, and no MLL rearrangement could be detected by FISH (Table 1).Table 1 Results of molecular studies on MLL rearrangement Sample type Obtained at age Method Analysis Result Ileum biopsy 10 days FISH MLL rearrangement Positive Bone marrow 3 weeks FISH MLL rearrangement Not feasible Nested PCR MLL-AF10 Positive (in retrospect) Bone marrow 5 weeks FISH MLL rearrangement Negative Nested PCR MLL-AF10 Positive (in retrospect) Peripheral blood 5 months FISH MLL rearrangement Positive RT-PCRa MLL-AF10 Positive Bone marrow 5 months RT-PCRa MLL-AF10 Positive aQuantitative real time PCR Subsequently, leukocyte, erythrocyte, and thrombocyte counts recovered spontaneously, accompanied by gradual clinical improvement including complete weaning from ventilator support and reversal of ileostomy. The patient was discharged at 6 weeks of age and a weight of 2900 g in excellent clinical condition. Subsequent clinical follow-up displayed regular weight gain (weight at birth 3000 g, at the age of 6 weeks 2900 g, at 3 months 4300 g) and development and routine hematological controls remained unremarkable. Five months later, while on treatment for hemangioma with propranolol, suspicious, blue-berry muffin-like skin lesions were noted on clinical examination. Leukocyte count was 28.500/μl, and thrombocyte and erythrocyte counts were normal. Monoblastic leukemia (AML FAB M5) was diagnosed from peripheral blood and bone marrow (Fig. 2). An MLL-AF10 fusion transcript was detected by real-time and nested PCR. Retrospective analyses (by nested PCR) at the time point, when bone marrow sampled in the third week of life was MLL-rearrangement-negative by FISH, showed that the MLL-AF10 fusion transcript was already detectable back then, 4 months prior to the definitive diagnosis of AML in the bone marrow (Table 1).Fig. 2 Monoblastic leukemia (AML FAB M5) (×63 magnification) diagnosed from peripheral blood (left) and bone marrow (right) at the age of 5 months Chemotherapy was initiated according to the German AML-BFM 2012 study protocol [5]. Due to the MLL-AF10 rearrangement, the patient was classified as “high risk” and consequently qualified for hematopoietic stem cell transplantation (hSCT). The patient received one block of induction and two blocks of consolidation chemotherapy. Subsequently, the patient underwent hSCT from an unrelated, HLA-matched donor. The conditioning regimen consisted of busilvex, cyclophosphamide, melphalan, and anti-thymocyte globuline (ATG). Cyclosporin A and methotrexate were administered as prophylaxis against Graft versus Host Disease (GvHD). Leukocyte engraftment was detected at day 19, granulocyte engraftment at day 23 post transplantation. Chemotherapy and stem cell transplantation were well tolerated under supportive care. The girl was discharged at day +60. Apart from partial compromise of renal function, no significant complications occurred. Until now, the patient is well, relapse-free, and showing no signs of GvHD 9 months after transplantation. Donor chimerism is 100 %. Discussion Neonatal leukemia is extremely rare. It accounts for about 1 % of pediatric leukemias and has an estimated prevalence of one to five in a million neonates. The disease was already known in the early twentieth century [6]. Neonatal AML shows similarities to transient myeloproliferative disorder (TMD), an abnormal proliferation of myeloid blasts observed in Down’s syndrome that usually resolves without therapy [7]. A few cases with spontaneous remission, which in one case was transient, mimicking TMD of Down syndrome, have been described in neonates with normal karyotypes [2, 8, 9]. Isolated leukemia cutis has also been described to disappear spontaneously [10, 11]. However, to the best of our knowledge, no case of a preterm baby with persistent pulmonary hypertension (PPHN) and cardiorespiratory failure requiring NO-based artificial ventilation and high-dose catecholamines could successfully be rescued by combined efforts of intensive care, chemotherapy, and stem cell transplantation. The clinical picture of neonatal leukemia is variable. Common symptoms include leukocytosis, cytopenia, hepatosplenomegaly, and skin infiltrations. The latter are known as leukemia cutis, appear in about 60 %, and are the first symptom in about one half of neonatal leukemia cases. Leukemia cutis, presenting as firm nodules, papules or plaques of red, blue or purple color, is a differential diagnosis of the so-called blueberry muffin baby. This syndrome has a diversified differential, ranging from hemato-oncological disorders like Langerhans cell histiocytosis, neuroblastoma, or rhabdomyosarcoma to infectious diseases like rubella, cytomegalovirus infection, or toxoplasmosis to hemolytic disease and blue rubber bleb nevus syndrome [12, 13]. Leukemia cutis has been described as an exclusive manifestation of AML without bone marrow disease [14], as well as appearing prior to or simultaneously with bone marrow involvement [13, 15]. Our patient presented with severe bruising at birth, probably due to thrombocytopenia and coagulopathy. Months later after presumed relapse of the so far not unequivocally diagnosed AML, she primarily presented with skin infiltrations. As a hematological malignancy, leukemia is essentially a distributed disease, but primary presentation as extramedullary infiltration, also called myeloid sarcoma, is a rarity in AML. Pediatric myeloid sarcoma has a prevalence of about 0.7 in one million children and can progress to AML [16, 17]. Infiltration of the gastrointestinal tract, including the ileum, has been reported in AML [18, 19]. In our patient, acute meconium ileus appeared on the tenth day of life. An extensive infiltration of monoblasts was found as the probable cause of the mechanical ileus. To our knowledge, such a condition has also not yet been described in neonatal leukemia. Chromosome 11q23 rearrangements involving the MLL gene locus are common in infant leukemia; they are the most frequent finding in neonatal AML [2]. Specifically, our patient had an MLL-AF10 fusion transcript, which is rather rarely found among the large variety of the various MLL rearrangements [20]. In the AML-BFM 2012 protocol, MLL-AF10 positive patients are considered “high risk” and recruited to a treatment protocol including hSCT. In our patient, a MLL rearrangement was first diagnosed from the intestinal infiltrate found in the terminal ileum and a MLL-AF10 fusion transcript was first detected at 5 months of age. In retrospect, its presence could then be verified from the bone marrow obtained at the age of 3 and 5 weeks by a sensitive PCR assay. As from the first days of life there is no more material available, it remains somehow unclear of what extent the leukemia initially presented and whether the cytopenia was due to bone marrow involvement or whether it was associated to cytokine release and/or coagulopathy. In the following weeks, the clinical condition improved and no signs of leukemia could be detected. It cannot be ruled out with certainty that regression of the leukemia was not spontaneous but caused by a single dose of vincristine and prednisolone treatment; however, given the prolonged recovery, the latter seems unlikely. Thus, other reasons for spontaneous regression of this aggressive type of leukemia, such as infection or disease related cytokine release, should be considered. Taken together, these findings indicate that our patient had a neonatal, MLL-AF10 positive AML with spontaneous transient regression that later reappeared as leukemia cutis and bone marrow disease. Conclusion We present a unique manifestation of neonatal MLL-AF10 positive AML with cardiorespiratory failure at birth, intestinal infiltration, and skin involvement. After a transient spontaneous regression, the patient was successfully treated by chemotherapy and hSCT and is alive and well. Abbreviations AF10, ALL1-fused gene from chromosome 10 protein; AML, acute myeloid leukemia; ATG, anti-thymocyte globuline; BFM, Berlin Frankfurt Münster; CD, cluster of differentiation; FAB, French-American-British; FISH, fluorescence in situ hybridization; GvHD, Graft versus Host Disease; hSCT, hematopoietic stem cell transplantation; LCH, Langerhans cell histiocytosis; MLL, mixed lineage leukemia; PCR, polymerase chain reaction; PPNH, persistent pulmonary hypertension; RT-PCR, real-time PCR; TMD, transient myeloproliferative disorder Funding No funding. Authors’ contributions TG first drafted the manuscript. JW and WB cared for the preterm and critically reviewed the manuscript for important intellectual content. WK and IO performed the pathological analysis. RS, SB, and CH performed the molecular analyses. RM critically revised the manuscript for important intellectual content. MK and AB supervised the project and reviewed and revised the manuscript for important intellectual content. All authors approved the final manuscript as submitted. Competing interests The authors declare that they have no competing interests. Consent for publication Written informed consent for the publication was given by both parents. ==== Refs References 1. Puumala SE Ross JA Aplenc R Spector LG Epidemiology of childhood acute myeloid leukemia Pediatric Blood & Cancer 2013 60 5 728 733 10.1002/pbc.24464 23303597 2. van der Linden MH Creemers S Pieters R Diagnosis and management of neonatal leukaemia Seminars in Fetal & Neonatal Medicine 2012 17 4 192 195 10.1016/j.siny.2012.03.003 22510298 3. Sansone R Negri D Cytogenetic features of neonatal leukemias Cancer Genetics and Cytogenetics 1992 63 1 56 61 10.1016/0165-4608(92)90065-G 1423228 4. Slany RK The molecular biology of mixed lineage leukemia Haematologica 2009 94 7 984 993 10.3324/haematol.2008.002436 19535349 5. Creutzig U Zimmermann M Dworzak MN Ritter J Schellong G Reinhardt D Development of a curative treatment within the AML-BFM studies Klinische Padiatrie 2013 225 Suppl 1 S79 86 23700063 6. Bernhard WG Gore I Kilby RA Congenital leukemia Blood 1951 6 11 990 1001 14869361 7. Karandikar NJ Aquino DB McKenna RW Kroft SH Transient myeloproliferative disorder and acute myeloid leukemia in Down syndrome. An immunophenotypic analysis American Journal of Clinical Pathology 2001 116 2 204 210 10.1309/XREF-C9T2-6U0A-4EDT 11488066 8. Isaacs H Jr Fetal and neonatal leukemia Journal of Pediatric Hematology/Oncology 2003 25 5 348 361 10.1097/00043426-200305000-00002 12759620 9. Chu JY O’Connor DM Gale GB Silberstein MJ Congenital leukemia: two transient regressions without treatment in one patient Pediatrics 1983 71 2 277 279 6823435 10. D’Orazio JA Pulliam JF Moscow JA Spontaneous resolution of a single lesion of myeloid leukemia cutis in an infant: case report and discussion Pediatric Hematology and Oncology 2008 25 5 457 468 10.1080/08880010802104494 18569848 11. Gottesfeld E Silverman RA Coccia PF Jacobs G Zaim MT Transient blueberry muffin appearance of a newborn with congenital monoblastic leukemia Journal of the American Academy of Dermatology 1989 21 2 Pt 2 347 351 10.1016/S0190-9622(89)80032-5 2754068 12. Fender AB Gust A Wang N Scott GA Mercurio MG Congenital leukemia cutis Pediatr Dermatol 2008 25 1 34 37 10.1111/j.1525-1470.2007.00578.x 18304150 13. Isaacs H Jr Cutaneous metastases in neonates: a review Pediatr Dermatol 2011 28 2 85 93 10.1111/j.1525-1470.2011.01372.x 21504442 14. Torrelo A Madero L Mediero IG Bano A Zambrano A Aleukemic congenital leukemia cutis Pediatr Dermatol 2004 21 4 458 461 10.1111/j.0736-8046.2004.21408.x 15283790 15. Zhang IH Zane LT Braun BS Maize J Jr Zoger S Loh ML Congenital leukemia cutis with subsequent development of leukemia Journal of the American Academy of Dermatology 2006 54 2 Suppl S22 27 10.1016/j.jaad.2005.04.038 16427986 16. Antic D Elezovic I Bogdanovic A Vukovic NS Pavlovic A Jovanovic MP Jakovic L Kraguljac N Isolated myeloid sarcoma of the gastrointestinal tract Internal Medicine 2010 49 9 853 856 10.2169/internalmedicine.49.2874 20453407 17. Reinhardt D Creutzig U Isolated myelosarcoma in children—update and review Leukemia & Lymphoma 2002 43 3 565 574 10.1080/10428190290012056 12002760 18. Cornes JS Jones TG Leukaemic lesions of the gastrointestinal tract Journal of Clinical Pathology 1962 15 305 313 10.1136/jcp.15.4.305 13881389 19. Domingo-Domenech E Boque C Narvaez JA Romagosa V Domingo-Claros A Granena A Acute monocytic leukemia in the adult presenting with associated extramedullary gastric infiltration and ascites Haematologica 2000 85 8 875 877 10942941 20. Meyer C Hofmann J Burmeister T Groger D Park TS Emerenciano M Pombo de Oliveira M Renneville A Villarese P Macintyre E Cave H Clappier E Mass-Malo K Zuna J Trka J De Braekeleer E De Braekeleer M Oh SH Tsaur G Fechina L van der Velden VH van Dongen JJ Delabesse E Binato R Silva ML Kustanovich A Aleinikova O Harris MH Lund-Aho T Juvonen V Heidenreich O Vormoor J Choi WW Jarosova M Kolenova A Bueno C Menendez P Wehner S Eckert C Talmant P Tondeur S Lippert E Launay E Henry C Ballerini P Lapillone H Callanan MB Cayuela JM Herbaux C Cazzaniga G Kakadiya PM Bohlander S Ahlmann M Choi JR Gameiro P Lee DS Krauter J Cornillet-Lefebvre P Te Kronnie G Schafer BW Kubetzko S Alonso CN zur Stadt U Sutton R Venn NC Izraeli S Trakhtenbrot L Madsen HO Archer P Hancock J Cerveira N Teixeira MR Lo Nigro L Moricke A Stanulla M Schrappe M Sedek L Szczepanski T Zwaan CM Coenen EA van den Heuvel-Eibrink MM Strehl S Dworzak M Panzer-Grumayer R Dingermann T Klingebiel T Marschalek R The MLL recombinome of acute leukemias in 2013 Leukemia 2013 27 11 2165 2176 10.1038/leu.2013.135 23628958

PMC005xxxxxx/PMC5002397.txt

==== Front Glob Health ActionGlob Health ActionGHAGlobal Health Action1654-97161654-9880Co-Action Publishing 3154210.3402/gha.v9.31542Original ArticleAccess to rehabilitation: patient perceptions of inequalities in access to specialty pain rehabilitation from a gender and intersectional perspective Wiklund Maria 123*Fjellman-Wiklund Anncristine 12Stålnacke Britt-Marie 24Hammarström Anne 23Lehti Arja 251 Department of Community Medicine and Rehabilitation, Physiotherapy, Umeå University, Umeå, Sweden2 Umeå Centre for Gender Studies in Medicine, Umeå University, Umeå, Sweden3 Department of Public Health and Clinical Medicine, Social Medicine, Umeå University, Umeå, Sweden4 Department of Community Medicine and Rehabilitation, Rehabilitation Medicine, Umeå University, Umeå, Sweden5 Department of Clinical Sciences, Professional Development, Umeå University, Umeå, Sweden* Correspondence to: Maria Wiklund, Department of Community Medicine and Rehabilitation, Physiotherapy, Umeå University, SE-901 87 Umeå, Sweden, Email: maria.e.wiklund@umu.seResponsible Editor: Carmen Vives Cases, Alicante University, Spain. 26 8 2016 2016 9 10.3402/gha.v9.3154207 3 2016 12 7 2016 15 7 2016 © 2016 Maria Wiklund et al.2016This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License, allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material for any purpose, even commercially, provided the original work is properly cited and states its license.Background Long-term musculoskeletal pain is common, particularly among women. Pain conditions are a concern in primary health care, and people with severe and complex pain are referred to specialty health care. There is gender bias in access, counselling, assessment, and treatment of long-term pain. Objective This study explores patient accounts and perceptions about important (social) factors for accessing specialised pain rehabilitation from gender and intersectional equality perspectives. We aimed to identify potential biases and inequalities in accessing rehabilitation resources at a specialised rehabilitation clinic. Design Individual semi-structured interviews were conducted with 10 adults after an assessment or completion of a specialised rehabilitation programme in northern Sweden. Qualitative content analysis was used to explore patients’ perceptions of important factors for accessing rehabilitation. Results One main theme was formulated as Access to rehab – not a given. Three categories of perceived inequality were demonstrated: power of gender, power of social status, and power of diagnosis. Participants perceived rehabilitation as a resource that is not equally available, but dependent on factors such as gender, socio-economic status, ability to work, ethnicity, or age, and more subtle aspects of social status and habitus (e.g. appearance, fitness, and weight). The character of diagnosis received (medical versus psychiatric or social) was also noted. Conclusions It is crucial that professionals are aware of how potential inequalities related to gender, social status, and diagnosis, and their intersections, can be created, perceived, and have influence on the processes of assessment and treatment. Reduction of social determinants of health and biases remain important within global, national, and local contexts. chronic paintreatment of painmultimodal rehabilitationgender biasequity in healthintersectionalityqualitative interviewsSwedenThis paper is part of the Special Issue: Gender and Health Inequality - intersections with other relevant axes of oppression. More papers from this issue can be found at www.globalhealthaction.net ==== Body Introduction Long-term musculoskeletal pain is common worldwide, occurring in approximately 19% of the adult European population. Women seem to be affected more frequently (1). Long-lasting, benign pain is influenced by, and interacts with, physical, emotional, psychological, and social factors. Therefore, a biopsychosocial perspective is required in order to understand and manage long-term pain (2, 3). Patients describe a range of negative consequences of pain, including significant pain intensity, psychological distress, anxiety and depression, insomnia, reduced ability to work (sick leave), ill health, and low quality of life (4–6). Individuals suffering from long-term pain may experience mental distress and may encounter social stigma, shame, or discrimination because of their condition (7–12). In addition, such pain is occasionally referred to as a ‘medically unexplained’ condition. This may be problematic for those suffering from pain (7, 13). Having a medically unexplained illness may be particularly troublesome in a health insurance system where having a specific medical diagnosis is an advantage in qualifying for sickness benefits. Public health insurance was a former cornerstone of the Swedish welfare state. In recent years, the phenomenon of individuals being excluded from sickness benefit has been the subject of heated debate among both lay people and politicians. In welfare states, such as in Sweden, the primary healthcare system is often primarily responsible for treatment and decisions about eventual sick leave (14). Pain conditions are a primary healthcare concern in Sweden, because 20–40% of people seeking primary health care suffer from pain conditions (15). Patients with recurrent, widespread, or severe pain can be referred to multidisciplinary assessment and multimodal rehabilitation in specialty health care. In terms of treatment models, multimodal pain rehabilitation has proven successful for patients with complex long-term pain (12, 14, 16). However, not all individuals who undergo assessment by a specialty team are selected for the multimodal rehabilitation programme, because they may need further investigation or benefit from unimodal treatment. Those who are not selected may find this hard to understand and accept. Medical researchers point to gender bias in access, counselling, assessment, and treatment of long-term pain in primary and specialty care (17–20). For instance, women with lower levels of education are less often selected for rehabilitation programmes (19, 20). Bias in health care is defined as unintended or systematic neglect of certain groups. This could be because of socio-economic position, educational level, sex/gender, or age (17, 21–23). Intersectional perspectives are therefore crucial to a better understanding of people with conditions such as musculoskeletal pain (8, 24, 25). Feminist perspectives point out the impact of social and cultural variables on mental health and emphasise gender inequalities and intersections of power differentials as key factors in women's emotional distress (26) and pain (19). Our theoretical frame is informed by the view of gender as a central social category and a social construction that is present in all social relationships, social structures, and power hierarchies, as well as in various health and illness processes (27–29). Gender is defined as a social determinant of health (30), which intersects with other aspects of oppression and social stratification (8, 31). Using Bourdieu's conceptualisation (32, 33), health care and rehabilitation are viewed as ‘fields’ where gendered and intersectional processes come into play, and where social status is continuously created by professionals and patients, depending on their social and symbolic ‘capital’ and ‘habitus’ (34). From the perspective of Connell (28), health care as societal institution may reflect gender orders and gender relations in the overall society. We may also find certain ‘gender regimens’ within this specific sector, for example, connected to rehabilitation practices (34). Little is currently known about the selection processes for specialised pain rehabilitation from an intersectional equality and gender perspective. To our knowledge, few studies have explored inequalities in assessment or pain rehabilitation from the patient's perspective. In a research project about equal care in pain rehabilitation, we investigated various aspects of gendered and social processes (19, 20, 35–37). Further identification of potential biases and inequalities in accessing rehabilitation resources, as perceived by individuals seeking help for long-term pain at a specialised rehabilitation clinic, is crucial. In this interview study, we aim to explore patients’ accounts and perceptions about important (social) factors in accessing specialty pain rehabilitation from a gender and intersectional equality perspective. Methods Setting and design This study was part of an overarching research project in northern Sweden named Equal care in rehabilitation, which combines quantitative (19, 20, 36) and qualitative methods (35, 37). In this substudy, we used a qualitative approach (38, 39) to focus on and deepen our understanding of patients’ perceptions and experiences. This specific focus distinguishes the present analysis from a previous one in which patients’ voices were merged with the views of professionals to illustrate the path of suffering (Via Dolorosa) along the normative chain of rehabilitation from primary care to the specialty pain clinic (37). Individual interviews were analysed by qualitative content analysis as applied in health sciences research (40). Qualitative content analysis is a suitable method for analysing human communication and individuals’ experiences in a systematic way (41) and has been used to explore the meanings of gender and ethnicity (42–44). The interview study was conducted in a clinical setting. Patients were referred by general practitioners for assessment (and eventual participation in a multimodal rehabilitation programme) at a specialty pain rehabilitation clinic in northern Sweden. Inclusion criteria for the multimodal programme were: the presence of disabling, non-malignant, long-lasting, and complex musculoskeletal pain (causing absenteeism due to sickness or major interference in daily life); aged 18–65 years; and not in need of further medical investigation. Thus, not every patient was selected for the programme. Those who were not selected were referred back to primary care with a rehabilitation plan. Procedures and participants Participants were purposively sampled from among individuals assessed recently at the pain rehabilitation clinic. They could have been selected for the multimodal programme or referred back to primary care. The sampling strategy aimed for variation in perceptions and experiences, and therefore, patients of different ages, sexes/genders, socio-economic, and ethnic backgrounds were invited to take part. The inclusion criterion for this study was suffering from severe and complex benign, long-term musculoskeletal pain. Fluency in Swedish was an inclusion criterion for attending the multimodal rehabilitation programme (partly based on group sessions), and therefore also for being invited to this study. Exclusion criteria were diagnosis of other pain than benign musculoskeletal pain. One of the authors (B-MS, the responsible physician) made the selections to attain diversity of ages, gender, and country of birth. The participants were first contacted with a letter signed by the responsible physician, and then a research assistant contacted them by telephone. Between December 2010 and February 2011, 10 patients (5 women and 5 men) aged 35–65 were eligible and interviewed. Of these, four were born in Sweden, three in Finland, and three outside Europe (Middle East or Latin America). Half of the participants had a mother tongue other than Swedish, but all spoke Swedish fluently. Five participants had higher education, and five did not have education beyond compulsory school. Four were working full-time and six were on sick leave or unemployed. Eight participants had undergone assessment at the rehabilitation clinic but not been selected to participate in the multimodal programme. Two had undergone assessment and had been selected to participate in the rehabilitation programme, which they had done. Data collection Individual, semi-structured, qualitative interviews were conducted (38, 39). An interview guide was used to support the interviews. The guide was based on open-ended questions about patient perceptions and experiences of assessment at the rehabilitation clinic, and participation in the rehabilitation programme. Aspects of gender, ethnicity, and class were of particular interest, but other aspects of potential perceived inequality were also explored with probing and follow-up questions. Examples of questions in the interview guide are: ‘How was your opportunity to make your voice heard in the meeting with the professionals at the specialty rehab?’ ‘What are your associations when you hear the expression “care on equal terms” [as it relates to your personal experience]?’, ‘Did you perceive that the assessment [at the specialty clinic] was influenced by your sex/gender, education, country of birth, or other characteristics?,’ and ‘What happened after the assessment?’ All interviews were digitally recorded and transcribed verbatim. Data analyses Data analyses were conducted using qualitative content analysis according to Graneheim and Lundman (40). Our interpretative frame was informed by our specific focus on aspects of gender and equality. The initial inductive phase kept relatively close to patients’ own accounts and the manifest content (45). The later steps of abstraction were guided by our clinical and theoretical awareness of potential inequalities and biases linked to gender, class, age, or ethnicity. First, the interviews were read by each author individually (MW, AF-W, AL) with respect to our naïve understanding and a sense for the whole. Preliminary aspects that met the research aim were thoroughly explored and discussed. Initially, the whole material was divided into two content areas: experiences from primary care versus experiences from specialty rehab. The present analysis is limited to experiences from the specialty rehab. Experiences from primary care have been published elsewhere (37). The analysis was followed by detailed reading and coding on a slightly higher level of abstraction; this stage of the coding was done using qualitative software. Codes similar in content were then grouped into categories that corresponded to the research aim and our overall interpretation of the interviews. Finally, one main theme was formulated that captured the latent meaning, running as a red thread, throughout the codes and categories. During the final steps of the analytical procedure, triangulation between researchers (all five authors) with different clinical or theoretical perspectives was used as a method to increase trustworthiness (38, 46). The iteration between codes and interviews was another way to ensure trustworthiness. Results The results are presented as one theme: Access to rehab – not a given, which runs as a thread through the three categories: the power of gender, the power of social status, and the power of diagnosis (Table 1). Each category, with its interrelated subcategories, represents a factor of importance for accessing rehabilitation resources, as perceived by patients. Table 1 Patient perceptions of access to rehabilitation presented as subcategories, categories and the main theme Subcategories Categories Theme Women and men are (not) different The power of gender Men's voices are heard and get through Women with pain are drowned out in the crowd Patients are treated differently The power of social status Access to rehab – not a given Suit and tie helps Money talks Diagnosis with (non)status The power of diagnosis Diagnosis and ability to work as ‘entrance ticket’ Unfair assignment of diagnoses Access to rehab – not a given The theme, Access to rehab – not a given, captures the view of rehabilitation as a resource that is not equally available to all who need it. The participants were aware that they would not automatically qualify for rehabilitation and support for coping with pain. They articulated their own or others’ experiences of inequality, either in connection to visiting the pain clinic or in a more general sense. The type of problem, diagnosis, ability to work, gender, ethnicity, and social status were factors perceived as either facilitating or hindering access to rehabilitation resources and professional support. Notably, overlapping category content points to intersections between social aspects. The power of gender The category ‘the power of gender’ consists of the subcategories: women and men are (not) different, men's voices are heard and get through, and women with pain are drowned out in the crowd. The category demonstrates how gender perceptions permeated participant beliefs about treatment and access to rehabilitation, although not every participant had previously reflected upon these issues. Women and men are (not) different Some participants stated that women and men are different in terms of genetics, life-situations, and symptoms. Consequently, they thought these factors should be taken into account during assessment and treatment, as well as during decisions about rehabilitation programme admission. One woman stated that men's and women's situations are different because a woman is expected to be an ‘octopus’ with four pairs of hands. She was disappointed and thought she had been treated unfairly because she presented herself as ‘capable and quick’ during the assessment. Despite many years of suffering with pain, psychological trauma, and social pressures, this was how she behaved. And in her view, the facade of ‘capability’ was both part of her coping strategy and at the heart of her problem. She thought this (presenting appearance) should have been a strong reason for inclusion in the programme; whereas she thought the professionals saw it as a reason for exclusion. Others did not relate their experiences of being poorly treated at the rehabilitation clinic as an issue of gender or gender inequality; rather, they framed the issue as a question of being able to demonstrate strength and to provide answers:Researcher: You said [earlier] that you are very ‘thick-skinned’ and would speak up if you felt badly treated. Do you think that this is related to your sex, being a man? Participant: Hmm … I don't know. There are many men with low self-esteem, so … perhaps there are a greater percentage of men with thick skin, but there are also many tough girls who dare to speak their minds, so I don't know …. Men's voices are heard and get through Being seen, heard, and respected was an important issue to participants. Some expressed the view that it was easier for men to make their voices heard and have smoother passage through the healthcare system because they were regarded as more assertive. Women were regarded as being more in need of the support and advocacy of someone who ‘brought action’, and this was preferably a man.Researcher: Do you think that people receive different care? Participant: Hmmm, depending on the age, and sex, and what sort of person one is.… Researcher: Did you feel that the response you got was affected by you being a woman? Participant: Maybe not just because I'm a woman … but maybe if you aren't the type of person who just walks in there and claims that ‘I need this’, but you are probably the type of person that says ‘no, but this will pass’ … It's almost as if you need to bring someone else (with you). Researcher: In order to be taken seriously? Participant: Yes, and then most preferably [that person will be] a man, or someone who can speak well. On the question of whether she thought she would have got different treatment if she had been a man, the participant emphasised that women are expected to take more responsibility, and to manage on their own:Participant: I actually think that if it had been a man who had pain, workwise, he probably would not have been expected to arrange everything himself. I arranged everything myself. The perceptions that men have easier access to rehabilitation resources, receive more support, or have less responsibility may reflect dimensions of general bias and inequity related to gender. Other aspects of ‘having a voice’ or ‘lacking a voice’ such as not being a native Swedish speaker were also raised. Women with pain are drowned out in the crowd This subcategory of women with pain are drowned out in the crowd reflects the women's sense of invisibility and having less value or recognition because they were women. The subcategory also refers to intersects between the power of (male) sex/gender and power of certain (medical and specific) diagnoses. Other intersects were apparent, for instance, intersections of ethnicity (Swedish), sex/gender (woman), and diagnosis (fibromyalgia). The experience of marginalisation and low status throughout the rehabilitation process was expressed as ‘Swedish women with pain are treated like they are on an assembly line’. In contrast, another woman, whose pain started in connection to an accident, had a more positive experience during her rehabilitation programme treatment. She felt her voice was heard, and that the rehabilitation personnel were on ‘her side’. Generally, participants regarded women patients as less valued and holding a lower status than men within the healthcare system. The power of social status ‘The power of social status’ category is based upon the subcategories: patients are treated differently, suit and tie helps, and money talks. The category reflects perceptions of inequities related to socio-economic status, class, gender, and ethnicity. To some extent, this category represents general perceptions and is not related to the specific clinic, assessment, or programme where the participants received care. Patients are treated differently A common opinion was that patients are treated differently by health professionals, based on factors such as economic status, profession, age, nationality, and sex/gender. Aspects of appearance, weight, and fitness were mentioned by some participants. Perceived unfairness was commonly related as others’ experiences. For example, one woman said that age and ethnicity matters with reference to her parents-in-law, who were born in Finland: ‘They do not ask questions of those with higher education because the language makes it difficult to express themselves’. The same woman stated that, ‘Everything costs money in health care, even referral to radiology’. She thought this was an explanation for why treatments were not made available to everyone (i.e. the cost made some people less worthy of treatment). She was not accepted in the rehabilitation programme and tried to cope with her pain on her own (using walks, warm pillows, and an acupressure mat). Another participant, who had a low-status job in personal care, thought that people were treated differently because of their age, sex, and societal status. This was exemplified by her belief that her friend (a top administrator in the healthcare sector) received better treatment than she had. She also thought that her husband received a better response and more respect from the healthcare personnel than she did. Aspects connected to ethnicity, the advantage or disadvantage of being born in or outside Sweden were also expressed:Participant: I can imagine that there are some cultures where the difference is very large [comparing Swedes and other ethnicities in terms of behaviour and attitudes], and there is a lot of ignorance among Swedes that affects healthcare encounters. To me, they [healthcare professionals] start to speak English at once, even though I've lived here since I was 17, and I'm 48 years old tomorrow. Suit and tie helps (to access rehabilitation) Participants remarked that men with higher social status (white-collar workers) had advantages in accessing assessment and rehabilitation, and were given more respect. Social constructions of gender, for example, being a man or a woman, also were important for patients. As one participant expressed it, ‘a suit and tie helps’ in contacts with the rehabilitation team.Participant: One may be treated differently if you come in a suit and tie, so it matters if I arrive scruffy, or if I dress in a tie. Women who dressed in ‘make-up and high heels’ were seen to have an advantage. However, this dress code was not always seen as a guarantee for being respected or having one's pain taken seriously:Participant: Sometimes you go to a doctor when you have put on a little extra makeup and done your hair nicely. Then they [doctors] think you look great, and think ‘How can you have so much pain?’ – it isn't visible. This subcategory points to participants’ awareness of the importance of being an ‘ideal’ and ‘respectable’ patient with some social power and status (habitus), expressed through status markers and attributes such as appearance, proper clothing, and moderate weight. This is the power of a fit and young body. Money talks The above metaphor of suit and tie can also represent differences in socio-economic status. Certain, often expensive, treatments were talked about as not always being available to those with less money, or as available only when working for a large company with generous occupational health insurance:Participant: It is clear that those who have good jobs and earn good money … receive better care. Participant: There are some physiotherapists who are good at treating fibromyalgia, but it costs so damn much, you cannot afford it. The power of social status, including the metaphors of ‘suit and tie’ and ‘make-up and high heels’ that work as ‘entrance tickets’ to rehabilitation points to an intersection of gender, socio-economic, and educational status, and their impact on perceived inequalities in access to health care. The power of social status also points to subtle biases that are tied to appearance, including the power of a fit and young body. The power of diagnosis ‘The power of diagnosis’ category consists of the subcategories: diagnoses with (non)status, diagnosis and ability to work as ‘entrance ticket’, and unfair assignment of diagnoses. Participants recognised certain diagnoses with interrelated aetiologies, preferably specific medical diagnoses, as crucial ‘entrance tickets’ to care and rehabilitation. They also viewed ability to work, that is, a good prognosis for rehabilitation back into working life, as facilitating access to support and care. Perceived inequalities in assessment and treatment, or unclear and unfair distribution of diagnoses, generated feelings of powerlessness and frustration. Diagnosis with (non)status The subcategory diagnosis with (non)status points to the participants’ awareness of differences in the status of certain diagnoses, and how this status moderated access to rehabilitation. Musculoskeletal diagnoses were generally preferred. Diagnoses connected to the ‘psyche’ or ‘social’, along with non-specific, diffuse symptoms or long-term pain, were thought to have a lower status. A treatment recommendation of ‘psych drugs’ was seen as demeaning. The experience of being demeaned and ignored because of non-specific or non-measurable pain problems was expressed as: ‘Who cares about the ones with pain?’ or ‘pain cannot be measured, and pain problems are disregarded’. Biomedical health problems and diseases, which are possible to measure and treat, were felt to have a higher status. Fibromyalgia was one example of a ‘psychosomatic disease’ that was viewed as getting less attention and having a low reputation in health care. This view was put forward by a woman born in Latin America. The psychosocial-oriented approach used by the rehabilitation team was questioned. A man who suffered from pain and had a very high intake of pain medications expressed his disappointment and powerlessness:Participant: The social parts [of assessment in rehabiliation] were alright – it was not that. I should not complain, but I did not come here to get the ‘socialization’ [Swedish: social träning] or something … I had hoped when I was sent here that they would do something instead of just talking … I am still on so much medication. As seen above, those who were frustrated had expected more help than ‘just talk’. Consequently, some had difficulty accepting the assessment outcome, particularly if they were not offered participation in the rehabilitation programme. Feelings of not being heard, believed, or minimised when meeting the interdisciplinary team were expressed by the men who were not selected for the programme, and who therefore had to continue rehabilitation training on their own:Participant: The individual meetings with the physiotherapist, doctor, and social worker were positive, but when all four of us sat together I was disappointed. You get the feeling that they sit there and think that you are making things up and lying. I left not much the wiser. They could not really say what it was. I felt that I was the one who was leading the whole discussion, and it should not be that way … They said: ‘Keep working out, but remember to cut down a bit, perhaps’. They shouldn't just send me home to ‘try and rehab yourself’. Instead, they should either send me to a real specialist or they [the rehab clinic] should continue (working with me). Participant: I felt that one of the team tried to minimize my problems. I thought that was really bad. His assessment was a bit different [compared to the others in the team]. But it has not had any [negative] consequences for me; it's Neurorehabilitation's assessment that is decisive anyway. We were not on the same wavelength; I didn't really understand. He wrote out the medication prescription and told me to rest while I exercise. You have to experiment, yourself, with your medicine to find some kind of acceptable level. Others expressed how pleased they were with their treatment and emphasised that the team was ‘very professional’. One of the men from the Middle East, who participated in the programme, was very pleased because ‘… they made me relax. Even the psychologist and social worker were professional. They wanted to look at it [the problem] from all sides and angles, investigate my real concerns. It was really good’. Diagnosis and ability to work as ‘entrance ticket’ Besides the importance of receiving a relevant explanation of symptoms and problems, receiving a diagnosis was seen as facilitating further access to specific examinations, radiology, or surgery, as well as to wider societal welfare resources such as sick leave and sickness allowance. Ability to work was seen as a prerequisite for rehabilitation and support. In those cases, the visit to the rehabilitation clinic did not lead to the expected diagnosis, dashed hopes, disappointment, and harsh consequences were described:Researcher: … So you don't believe that the visit to pain-rehab helped, or was consistent with your expectations? Participant: No, it is the opposite … in all the other places I have been, I have had hopes of getting help to find a way to feel better – and that possibility was shut down at pain-rehab because their focus had changed and become something entirely different. So it felt bad. You felt terribly alone in the situation, which is similar as it is in society now – everywhere it's just up to you to try and feel better. There is no one that can help me, so it feels lonely. As seen in the quote above, the participant compared her experiences to her perception of managing alone in the wider society. On the question of whether there was something with which she was particularly dissatisfied, she answered:Yes, well, with the decisions they made, after this visit I am now shut out from the health insurance system. Thus, the decision that they made, or the assessment they made, has not actually given me something, but it has instead made my situation worse. Economically … so that's … what I think is negative. In other words, … I actually went there in the hope of getting better and in the hope there was a program that would help. And instead they said ‘but we have no role here other than focusing on the labour market’. Oh, and it was as if to say ‘I'm not ready for that’ … and then end up in an investigation at the primary care health centre, and then one is excluded from sickness benefit. So their judgement/assessment of me has clearly influenced me negatively. The disappointed participant understands the decision as related to recent changes in the ‘political system’ (the Swedish social security net), where the ability and prognosis of being able to go back into work is a stimulus for access to social support; or, as she puts it, ‘there is nowhere I fit in without the ability to work’. Another woman with a high educational level, who was born in Finland, was also disappointed and dissatisfied because she did not receive a diagnosis. Because of this, she was excluded from sickness benefit by the health insurance office. Unfair assignment of diagnoses Thus, the actual assignment of diagnoses and treatment was understood by the participants to sometimes be unfair and to result in inequity. For instance, one man from the Middle East felt powerless about his situation and questioned why he was not offered surgery for his intervertebral disk displacement. He felt it was strange and unfair when he compared his situation and symptoms to those of a female Swedish colleague who got surgery and pain relief:Researcher: What could the best medical care do for you? Participant: A work colleague, she got surgery, and she had exactly the same problem. Why could they not give me surgery? I never really got any answer. Researcher: Was it something you were talking about? Participant: Yes, I said I wanted surgery. They said ‘Yes, but we do not think surgery is needed’. I think they mentioned something about it being hereditary or that my back is worn out … but I do not believe in heredity. Researcher: Is there anything that could help to get rid of or reduce the pain? Participant: I hope that there is a way, whether it is surgery or blocking nerves or something that can be done to at least reduce the pain. So that I can use less analgesics, because I am increasing the dose all the time. Where will it end otherwise? The above quote shows that the man had not understood or accepted the explanations about the nature of his pain that were given to him by the rehabilitation clinic team, and that he felt he was left alone with his questions, overwhelming pain, and need of medication. From his viewpoint, it was problematic and frustrating not to be selected for surgery or guided in pain management and medication adjustment. As participants viewed it, receiving a specific diagnosis represented an ‘entrance ticket’ to care, rehabilitation, and welfare, as well as to social acceptance, confirmation, and social legitimacy. The diagnosis and rehabilitation procedure can be understood as both gendered and classed, and as such intersects with social aspects of gender, class, age, and ability to work. Participants’ narratives point to the importance of getting a relevant and creditable diagnosis as a way to understand the pain and its causes, get relief from pain and suffering, be socially accepted and confirmed, access other examinations or treatments, and as a means to access support from the social health insurance system. Discussion According to our participants, access to rehabilitation depends on the power of social factors such as gender, ethnicity, social status, and the variable status of certain diagnoses. The ability to work was seen as a prerequisite for rehabilitation. Subtler aspects were also revealed. Examples of these are appearance, fitness, and weight. The theme ‘Access to rehab – not a given’ points to participants’ perceptions of rehabilitation as a resource that is not equally available to everyone. This result is congruent with our earlier findings from a register study in the same setting (19, 20, 36). Participants generally regarded women patients as being less valued and holding a lower status in the healthcare system than men. Women were regarded as being ‘in need of someone [preferably a man] who brought their case’, whereas men were regarded as having a stronger voice and the power to access more resources and support. Women were also viewed as having to take on greater responsibility, receiving less support, and being somewhat invisible in the system. This was formulated as women with pain are drowned out in the crowd. A number of studies suggest that physicians’ gendered expectations create gender differences in medicine. An example of this is found in neck pain rehabilitation (17, 47). Another example is found in a study of gender bias in general physicians’ initial examination of chest pain in men and women. The study found an overuse of services such as exercise testing and hospital care by men compared to women (23). In a recent study, Hammarström et al. (19) found that interdisciplinary teams in specialty health care may discriminate against poorly educated women with long-lasting disabling pain, but further research is still needed. Congruent with our view, Ahlsen et al. (48) argue that professionals in pain rehabilitation need to understand how constructs of gender interact with the pain story. With a better understanding of gender, a professional may help patients improve their health. Receiving a diagnosis or explanation was important to participants in making sense of their pain. This finding is supported by other studies (49, 50). Not receiving a diagnosis or an understandable explanation (particularly if not selected for the rehabilitation programme) was tied to perceptions of inequality or unfair treatment. Receiving a diagnosis can enhance a patient's sense of legitimacy when suffering from an illness that can otherwise be viewed as non-medical, psychiatric, or social and thus stigmatised. Within a strict biomedical discourse, the ‘legitimate user’ of health care is defined as one who has a ‘genuine condition’ (21). In such a context, it is understandable that a ‘proper’ diagnosis (i.e. medical and specific) that proves legitimacy is crucial to patients (7, 9, 10, 49). Participants felt that receiving a diagnosis facilitated access to rehabilitation resources, as well as to wider societal resources such as sickness benefit. This supports the idea of diagnoses as both categories and processes, and points to the multilayered complexities of diagnosis (51). The social construction of a ‘legitimate’ healthcare user, in this case of rehabilitation, can be defined as gendered, with the ideal masculine subject position constructed as an ‘infrequent-user’ and the feminine subject position as a ‘regular-user’ (21). By framing their condition within a biomedical discourse as opposed to ‘psychological’ or ‘social’, men (and perhaps women) can maintain a socially accepted identity (21). These gendered subject positions would be of great interest for further investigation in relation to rehabilitation and help-seeking. Such investigation should include patient experiences of legitimacy and respect because suffering from long-term pain is tied to complex gendered positions and expectations (7, 52). From a theoretical standpoint, our results show perceived inequalities and a biased social and gendered selection process that can be understood with the help of Bourdieu's concepts of social and cultural capital that are reflected in individuals’ (gendered) habitus and expressed through appearance, habits, and behaviours (32, 33). The category the power of social status captures the intersection between gender, socio-economic position, body shape (weight and fitness), appearance, and in some respects age and ethnicity. The category points to how participants wish to be seen as credible. When identifying inequities in health care and rehabilitation, it is thus crucial to be aware of the intersection of social categories and constructions. Using a tool that addresses gender inequality, and intersecting social aspects, in daily clinical work may be one way of raising awareness (36). The tool can be included in clinical assessment of patients and as an instrument for critical reflection on gender bias during health professional team education. Our intersectional study approach makes it clear that interplays between various social factors were created in a specific context and situation (53), and possibly a specific gender regimen (28). In our case, the interplay occurred during the process of selection for specialty rehabilitation within the powerful healthcare institution, which in turn is connected to the welfare system and conditions in the labour market. The ability to work as a prerequisite for rehabilitation and support is such an example.Within the context of our study, there were different ‘entrance tickets’ (related to social capital and habitus) that worked together to enhance or hinder selection and access to rehabilitation resources. The consequences of an individual being regarded as an ‘ideal patient’ (with access to specialty rehabilitation resources) or as a non-ideal patient (with access denied) were great. Outcomes were tangible for participants as they affected access to ways of coping with non-reversible pain, support, and the social health insurance system. Equity in health is important to consider in an individual's access to healthcare and rehabilitation resources (25). According to the World Health Organization, social determinants of health are ‘mostly responsible for health inequities – the unfair and avoidable differences in health status seen within and between countries’ (54). Therefore, it is crucial to reduce inequities in global, national, and local contexts (30, 54). Swedish national public health policies emphasise equal access to health care and the decrease of inequality and inequity in the population (55). This is why it is important to scrutinise and address equal access to rehabilitation resources. Our results highlight problems and barriers in the Swedish welfare model that result in tangible consequences for some individuals. Methodological considerations This study has strengths and limitations that are relevant to interpretation of the results. The strategic sample with variations in sex/gender, ages, and ethnicities is a strength (38). The fact that participants younger than 35 years were missing is a weakness. Moreover, the results may have been different if interviews were carried out separately for patients who participated in the multimodal rehabilitation programme, because they were more satisfied than the patients who were not selected to participate in the programme. Qualitative content analysis proved to be a suitable way of exploring the wide range of perceptions, as the method keeps to a descriptive level that is close to the participants’ own words and allows the display of diversity in categories (40). Trustworthiness was obtained through thick descriptions in quotations, and interdisciplinary triangulation during the analysis (public health, physiotherapy, rehabilitation medicine, family medicine, and gender studies) (46). The fact that some of the authors had field experiences in pain and rehabilitation can be an advantage and a disadvantage. Being familiar with pain, treatment, and rehabilitation made it easier to understand the meaning and context of the interviews. On the other hand, expectations derived from pre-existing knowledge could result in overlooking descriptions. Because the authors were not acquainted with the participants, there was no possibility of their relationships influencing the interviews. The equality perspective and gender-theoretical frame functioned as a critical lens that brought additional perspectives into the analyses and data interpretation (25). It is of note that in our analysis, gender appeared to be (and was treated as) the ‘master category’. From an intersectional perspective, this is not always the case (8, 25, 31). Despite a relatively small clinical sample with high within-group variability (e.g. gender, age, and ethnicity) and a complex research question, we believe that the study results reflect crucial aspects of perceived inequities. The results can likely be transferred to similar social contexts in terms of theoretical and analytical generalisations (46). However, there may also be diverging perceptions that were not captured in this study. Future studies need to further deepen and discuss these in relation to other rehabilitation contexts. Conclusion The participants’ narrated experiences of assessment or treatment at a pain rehabilitation clinic point to perceptions of inequalities. They are formulated as the theme Accessing rehab – not a given. Potential inequalities in access to rehabilitation resources were related to gender, diagnoses, and social status. It is crucial that healthcare professionals be aware of how potential inequalities can be created during assessment or treatment. Further research is needed in this field. Knowledge of biases and inequity in the rehabilitation for long-term pain is still lacking. There is also a need for studies on how other social categories intersect and have an impact on individual's access to rehabilitation resources. Ethics The study was approved by the Regional Ethics Vetting Board in Umeå, Sweden (Dnr 2010-44-31). Participants gave informed consent and were guaranteed confidentiality throughout the research process. Acknowledgements The authors thank the study participants for sharing their experiences and thoughts. Authors’ contributions AH was responsible for major study planning. B-MS contributed to recruitment of participants at the rehabilitation pain clinic. Research assistants conducted the interviews. MW, AF-W, and AL conducted the analyses. MW drafted the initial manuscript. AF-W and AL contributed to the writing process, and all authors contributed to revisions. Each author read and approved the final manuscript. Conflict of interest and funding This work was supported by the Swedish Research Council (grant nos 344-2009-5839 and 344-2011-5478). The authors report no declarations of interest. The authors alone are responsible for the content and writing of this article. Paper context Long-term pain is common worldwide. Gender bias and inequalities in access to rehabilitation and treatment are found because of socio-economic position, age, or gender. Application of intersecting perspectives is therefore crucial. To identify potential biases in accessing rehabilitation resources, we explored patients’ perceptions from a gender and intersectional equality perspective. Access to pain rehabilitation is not perceived as a given. How to provide equal access to rehabilitation resources needs to be further scrutinised and implemented. ==== Refs References 1 Breivik H Collett B Ventafridda V Cohen R Gallacher D Survey of chronic pain in Europe: prevalence, impact on daily life, and treatment Eur J Pain 2006 10 287 333 16095934 2 Nordin C Fjellman-Wiklund A Gard G In search of recognition – patients’ experiences of patient participation prior to multimodal pain rehabilitation Eur J Physiother 2014 16 49 57 3 Merrick D Sundelin G Stålnacke BM One-year follow-up of two different rehabilitation strategies for patients with chronic pain J Rehabil Med 2012 44 764 73 22915044 4 de Heer EW Gerrits MMJG Beekman ATF Dekker J van Marwijk HWJ de Waal MWM The association of depression and anxiety with pain: a study from NESDA PLoS One 2014 9 e106907 25330004 5 Bair MJ Robinson RL Katon W Kroenke K Depression and pain comorbidity: a literature review Arch Intern Med 2003 163 2433 45 14609780 6 Börsbo B Gerdle B Peolsson M Impact of the interaction between self-efficacy, symptoms and catastrophising on disability, quality of life and health in with chronic pain patients Disabil Rehabil 2010 32 1387 96 20513205 7 Werner A Malterud K It is hard work behaving as a credible patient: encounters between women with chronic pain and their doctors Soc Sci Med 2003 57 1409 19 12927471 8 Hankivsky O Women's health, men's health, and gender and health: implications of intersectionality Soc Sci Med 2012 74 1712 20 22361090 9 Album D Westin S Do diseases have a prestige hierarchy? A survey among physicians and medical students Soc Sci Med 2008 66 182 8 17850944 10 Grue J Johannessen LEF Rasmussen EF Prestige rankings of chronic diseases and disabilities. A survey among professionals in the disability field Soc Sci Med 2015 124 180 6 25461875 11 Holloway I Sofaer-Bennett B Walker J The stigmatisation of people with chronic back pain Disabil Rehabil 2007 29 1456 64 17729093 12 Gustafsson M Ekholm J Öhman A From shame to respect: musculoskeletal pain patients’ experience of a rehabilitation programme, a qualitative study J Rehabil Med 2004 36 97 103 15209451 13 Nettleton S ‘I just want permission to be ill’: towards a sociology of medically unexplained symptoms Soc Sci Med 2006 62 1167 78 16135395 14 Scascighini L Toma V Dober-Spielmann S Sprott H Multidisciplinary treatment for chronic pain: a systematic review of interventions and outcomes Rheumatology 2008 47 670 8 18375406 15 Hasselström J Liu-Palmgren J Rasjö-Wrååk G Prevalence of pain in general practice Eur J Pain 2002 6 375 85 12160512 16 SBU Chronic pain rehabilitation 2010 Stockholm Swedish Council of Health Technology Assessment 17 Hamberg K Risberg G Johansson EE Westman G Gender bias in physicians’ management of neck pain: a study of the answers in a Swedish national examination J Womens Health Gend Based Med 2002 11 653 66 12396897 18 Raine R Does gender bias exist in the use of specialist health care? J Health Serv Res Policy 2000 5 237 49 11184961 19 Hammarström A Haukenes I Fjellman Wiklund A Lehti A Wiklund M Evengård B Low-educated women with chronic pain were less often selected to multidisciplinary rehabilitation programs PLoS One 2014 9 e97134 24849625 20 Stålnacke B Haukenes I Lehti A Fjellman Wiklund A Wiklund M Hammarström A Is there a gender bias in the recommendations for further rehabilitation in primary care of patients with chronic pain after an interdisciplinary team assessment? J Rehabil Med 2015 47 365 71 25587848 21 Noone JH Stephens C Men, masculine identities, and health care utilisation Sociol Health Illn 2008 30 711 25 18564976 22 Ruiz-Cantero MT Vives-Cases C Artazcoz L Delgado A García Calvente MM Miqueo C A framework to analyse gender bias in epidemiological research J Epidemiol Community Health 2007 61 Suppl 2 46 53 23 Bosner S Haasenritter J Hani M Keller H Sonnichsen A Karatolios K Gender bias revisited: new insights on the differential management of chest pain BMC Fam Pract 2011 12 45 21645336 24 Hallett K Intersectionality and serious mental illness – a case study and recommendations for practice Women Ther 2015 38 156 74 25 Hammarström A Johansson K Annandale E Ahlgren C Aléx L Christianson M Central gender theoretical concepts in health research: the state of the art J Epidemiol Community Health 2014 68 185 90 24265394 26 Muenzenmaier K Margolis F Langdon GS Rhodes D Kobayashi T Rifkin L Transcending bias in diagnosis and treatment for women with serious mental illness Women Ther 2015 38 141 55 27 Connell R Gender, health and theory: conceptualizing the issue, in local and world perspective Soc Sci Med 2012 74 1675 83 21764489 28 Connell RW Gender and power: society, the person and sexual politics 1987 Stanford, CA Stanford University Press 29 Springer KW Hankivsky O Bates LM Gender and health: relational, intersectional, and biosocial approaches Soc Sci Med 2012 74 1661 6 22497844 30 CSDH Closing the gap in a generation: health equity through action on the social determinants of health Final Report of the Commission on Social Determinants of Health 2008 Geneva World Health Organization 31 Bates LM Hankivsky O Springer KW Gender and health inequities: a comment on the final report of the WHO Commission on the Social Determinants of Health Soc Sci Med 2009 69 1002 4 19665829 32 Bourdieu P The logic of practice 1990 Cambridge Polity Press 33 Adkins L Skeggs B Feminism after Bourdieu 2004 Oxford Blackwell 34 Öhman A Profession on the move: changing conditions and gendered development in physiotherapy 2001 Umeå Umeå University 35 Hammarström A Wiklund M Stålnacke B-M Lehti A Haukenes I Fjellman-Wiklund A Developing a tool for increasing the awareness about gendered and intersectional processes in the clinical assessment of patients – A study of pain rehabilitation PLoS One 2016 11 e0152735 27055029 36 Haukenes I Hensing G Stalnacke BM Hammarstrom A Does pain severity guide selection to multimodal pain rehabilitation across gender? Eur J Pain 2015 19 826 33 25366906 37 Lehti A Fjellman-Wiklund A Stålnacke B-M Hammarström A Wiklund M Walking down ‘Via Dolorosa’ from primary health care to the specialty pain clinic – patient and professional perceptions of inequity in rehabilitation of chronic pain Scand J Caring Sci 2016 doi: http://dx.doi.org/10.1111/scs.12312 38 Dahlgren L Emmelin M Winkvist A Qualitative methodology for international public health 2004 Umeå Epidemiology and Public Health Sciences, Department of Public Health and Clinical Medicine, Umeå University 39 Kvale S Interviews. An introduction to qualitative research interviewing 1996 London Sage 40 Graneheim UH Lundman B Qualitative content analysis in nursing research: concepts, procedures and measures to achieve trustworthiness Nurse Educ Today 2004 24 105 12 14769454 41 Downe-Wamboldt B Content analysis: method, applications, and issues Health Care Women Int 1992 13 313 21 1399871 42 Lehti AH Johansson EE Bengs C Danielsson U Hammarström A ‘The Western Gaze’ – An analysis of medical research publications concerning the expressions of depression, focusing on ethnicity and gender Health Care Women Int 2010 31 100 12 20390640 43 Alex L Fjellman Wiklund A Lundman B Christianson M Hammarström A Beyond a dichotomous view of the concepts of ‘sex’ and ‘gender’ focus group discussions among gender researchers at a medical faculty PLoS One 2012 7 e50275 23185593 44 Wiklund M Bengs C Malmgren-Olsson E-B Öhman A Young women facing multiple and intersecting stressors of modernity, gender orders and youth Soc Sci Med 2010 71 1567 75 20846769 45 Hsieh H-F Shannon SE Three approaches to qualitative content analysis Qual Health Res 2005 15 1277 88 16204405 46 Lincoln YS Guba EG Naturalistic inquiry 1985 London Sage 47 Hamberg K Gender bias in medicine Womens Health 2008 4 237 43 48 Ahlsen B Bondevik H Mengshoel AM Solbrække KN (Un)doing gender in a rehabilitation context: a narrative analysis of gender and self in stories of chronic muscle pain Disabil Rehabil 2014 36 359 66 23688293 49 Toye F Seers K Allcock N Briggs M Carr E Andrews J Patients’ experiences of chronic non-malignant musculoskeletal pain: a qualitative systematic review Br J Gen Pract 2013 63 e829 41 24351499 50 Skuladottir H Halldorsdottir S The quest for well-being: self-identified needs of women in chronic pain Scand J Caring Sci 2011 25 81 91 20409049 51 Jutel A Nettleton S Towards a sociology of diagnosis: reflections and opportunities Soc Sci Med 2011 73 793 800 21868144 52 Ahlsen B Mengshoel AM Solbrække KN Troubled bodies – troubled men: a narrative analysis of men's stories of chronic muscle pain Disabil Rehabil 2012 34 1765 73 22394105 53 Haraway D Simians, cyborgs, and women. The reinvention of nature 1991 New York Routledge 54 WHO Social determinants of health World Health Organization 2015 Available from: http://www.who.int/social_determinants/sdh_definition/en/ [cited 16 March 2016] 55 Socialdepartementet Svensk författningssamling (SFS) [Health Care Law] 1982 Stockholm Socialdepartementet

PMC005xxxxxx/PMC5002398.txt

==== Front Glob Health ActionGlob Health ActionGHAGlobal Health Action1654-97161654-9880Co-Action Publishing 3239010.3402/gha.v9.32390Original ArticleTuberculosis screening in patients with HIV: use of audit and feedback to improve quality of care in Ghana Bjerrum Stephanie 12*Bonsu Frank 3Hanson-Nortey Nii Nortey 3Kenu Ernest 45Johansen Isik Somuncu 1Andersen Aase Bengaard 26Bjerrum Lars 7Jarbøl Dorte 8Munck Anders 81 Department of Infectious Diseases, Odense University Hospital, Odense, Denmark2 Institute of Clinical Research, Faculty of Health Sciences, University of Southern Denmark, Odense, Denmark3 National Tuberculosis Control Programme, Disease Control and Prevention Department, Ghana Health Services, Korle-Bu Teaching Hospital, Accra, Ghana4 Department of Medicine–Fevers Unit, Korle-Bu Teaching Hospital, Accra, Ghana5 School of Public Health, University of Ghana, Legon, Accra, Ghana6 Department of Infectious Diseases, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark7 Department of Public Health, University of Copenhagen, Copenhagen, Denmark8 Department of Public Health, University of Southern Denmark, Odense, Denmark* Correspondence to: Stephanie Bjerrum, Department of Infectious Diseases, Odense University Hospital, Søndre Boulevard 29, DK-5000 Odense C, Denmark, Email: steph@medicinsk.dk26 8 2016 2016 9 10.3402/gha.v9.3239025 5 2016 01 8 2016 02 8 2016 © 2016 Stephanie Bjerrum et al.2016This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License, allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material for any purpose, even commercially, provided the original work is properly cited and states its license.Background Tuberculosis screening of people living with HIV (PLHIV) can contribute to early tuberculosis diagnosis and improved patient outcomes. Evidence-based guidelines for tuberculosis screening are available, but literature assessing their implementation and the quality of clinical practice is scarce. Objectives To assess tuberculosis screening practices and the effectiveness of audit and performance feedback to improve quality of tuberculosis screening at HIV care clinics in Ghana. Design Healthcare providers at 10 large HIV care clinics prospectively registered patient consultations during May and October 2014, before and after a performance feedback intervention in August 2014. The outcomes of interest were overall tuberculosis suspicion rate during consultations and provider adherence to the International Standards for Tuberculosis Care and the World Health Organizations’ guidelines for symptom-based tuberculosis screening among PLHIV. Results Twenty-one healthcare providers registered a total of 2,666 consultations; 1,368 consultations before and 1,298 consultations after the feedback intervention. Tuberculosis suspicion rate during consultation increased from 12.6 to 20.9% after feedback (odds ratio, OR 1.83; 95% confidence interval, CI: 1.09–3.09). Before feedback, sputum smear microscopy was requested for 58.7% of patients with suspected tuberculosis, for 47.2% of patients with cough ≥2 weeks, and for 27.5% of patients with a positive World Health Organization (WHO) symptom screen (any of current cough, fever, weight loss or night sweats). After feedback, patients with a positive WHO symptom screen were more likely to be suspected of tuberculosis (OR 2.21; 95% CI: 1.19–4.09) and referred for microscopy (OR 2.71; 95% CI: 1.25–5.86). Conclusions A simple prospective audit tool identified flaws in clinical practices for tuberculosis screening of PLHIV. There was no systematic identification of people with suspected active tuberculosis. We found low initial tuberculosis suspicion rate compounded by low referral rates of relevant patients for sputum smear microscopy. Adherence to recommended standards and guidelines for tuberculosis screening improved after performance feedback. HIVtuberculosisscreeningquality of healthcareclinical practiceauditfeedback ==== Body Introduction Tuberculosis remains a leading cause of morbidity and mortality among people living with HIV (PLHIV) (1). A review of post-mortem studies among HIV-infected individuals showed that tuberculosis was implicated in 33% of all HIV-associated adult deaths in sub-Saharan Africa, with almost half of the fatal cases undiagnosed in life (2). In Ghana, the HIV epidemic is moderate with a prevalence of 1.4% (3). The overall tuberculosis prevalence is 282/100,000 (4), and low tuberculosis case detection rate overall and among PLHIV has been identified as main challenges in the country (5). To increase tuberculosis case detection among PLHIV in resource-constrained areas, great effort has been made to evaluate tuberculosis screening algorithms and diagnostic tests (6–8). In 2011, the World Health Organization (WHO) endorsed a four-symptom-based clinical algorithm as part of their guidelines for tuberculosis screening of PLHIV (9). Moreover, The International Standards for Tuberculosis Care (ISTC) published by WHO and the American Thoracic Society is available to promote high-quality services (10, 11). However, there is little information on the quality of tuberculosis screening practices in the context of routine care. To date, studies have focused on coverage of tuberculosis screening based on retrospective record reviews or routine health information (12–15). These studies reported great variation in coverage with a general increase over time and after education of healthcare providers. Low referral rates for tuberculosis diagnostic test were reported in some of the studies (13, 15). Still, the need remains to assess healthcare providers’ initial tuberculosis suspicion and decision to refer patients for further diagnostic tests as well as identify strategies to improve quality of care under real-life conditions. Clinical audit and feedback is a tool to assess clinical practice and improve the quality of care, and has been found effective to close the gap between routine practice and recommended guidelines (16, 17). Audit has previously been used to assess the quality of diagnosis of smear-negative pulmonary tuberculosis (18) and to improve general tuberculosis diagnostic services in Latin America (19). In collaboration with the National Tuberculosis Control Programme (NTP) in Ghana, we conducted a quality improvement study at 10 large HIV care clinics. The aim was to assess healthcare providers’ adherence to widely recommended standards for HIV-associated tuberculosis screening and evaluate if audit and feedback could improve performance. We used a simple tool for audit and feedback based on prospective self-registration of healthcare providers’ clinical practice focusing on tuberculosis suspicion and referral for sputum smear microscopy. We use the Standards for Quality Improvement Reporting Excellence (SQUIRE 2.0) to report our results (20). Methods Design This is a prospective study with a one-group pre- and post-evaluation of audit and feedback to address and improve the quality of tuberculosis screening practices at HIV care clinics. Study sites and context The study was conducted at 10 large HIV care clinics located in major hospitals from three regions in the southern zone of Ghana. The clinics were selected from the national list of public HIV care clinics based on the following criteria: >3,000 PLHIV enrolled at the clinics; access to X-ray and sputum smear microscopy for acid-fast bacilli at the hospital premises; and availability of a tuberculosis care clinic or a clinic with integrated delivery of HIV and tuberculosis services at the hospital premises. Two of the HIV care clinics were affiliated to a major teaching hospital. Half of the study sites had access to a specific clinic with integrated HIV/tuberculosis services at the hospital. In Ghana, the NTP and the National AIDS Control Programme (NACP) implement HIV and tuberculosis collaborative services and perform routine monitoring and support to health facilities. As in many resource-constrained areas, sputum smear microscopy is the most widely used tuberculosis diagnostic test, while new diagnostic test like the Xpert MTB-RIF is being scaled up (21). National guidelines recommend regular symptom-based tuberculosis screening of PLHIV including questions on cough and duration (≥2 weeks or <2 weeks), cough with blood, weight loss, night sweats, fever, chest pain, and history of any contact to persons known with tuberculosis (22, 23). All individuals with cough ≥2 weeks should be referred for sputum smear microscopy for acid-fast bacilli. The WHO guidelines available for tuberculosis screening in PLHIV recommend that individuals reporting any weight loss, fever, night sweats, and current cough (of any duration) should be referred for further tuberculosis investigations (9). According to the WHO guidelines, isoniazid preventive treatment (IPT) should be offered to those individuals reporting none of these symptoms, but IPT is not yet implemented in Ghana. At the time the study was conducted, 4 of 10 HIV care clinics had access to Xpert MTB/RIF, but none used Xpert MTB/RIF as part of routine diagnostic services for tuberculosis. Participants We invited all healthcare providers involved in tuberculosis screening at the 10 HIV care clinics to participate in the study. Medical doctors and nurses, but also physician assistants and disease control officers, were invited for participation if active in tuberculosis screening. Participants were asked to register consultations with PLHIV during May and October 2014. Data collection – Audit Project Odense We used the Audit Project Odense (APO) method for audit and compilation of feedback. The method is originally developed for quality improvement in general practice but has since been used widely (24–27). The APO method is based on prospective self-registration of clinical practice during patient consultations using a simple paper form. The form used for this study (Supplementary material A) was designed for registrations to take <2 min per consultation. The healthcare providers were asked to register the first 15 consultations with PLHIV daily for 2–4 weeks before and after a feedback intervention – in the following referred to as the first and second audit, respectively. Data registered during consultation included demographic characteristics of patients (age and sex); the consultation type (initial assessment of a newly diagnosed HIV-infected individual vs. follow-up consultations); antiretroviral treatment (ART) status of patients (currently receiving ART, defaulted ART for more than 1 month since last consultation, or never received ART); tuberculosis-related signs and symptoms presented during consultation (cough, weight loss, night sweats, fever); the healthcare providers’ main suspicion or diagnosis (including tuberculosis, pneumonia, upper respiratory tract infection); and prescribed investigations. Healthcare providers were carefully instructed orally and in writing (Supplementary material B) on how to fill in the registration form before the first audit and received a follow-up visit by the research team during registration. They were further asked to complete a form with background information about themselves and the HIV care clinic they represented. Feedback Based on the results of the first audit, performance feedback was developed for each of the healthcare providers. This included a feedback report where the results from the first audit were summarized in words, tables, and diagrams focusing on the healthcare providers’ overall practice for suspecting tuberculosis and referral of tuberculosis suspects for further investigations (Fig. 1). Moreover, a personal feedback sheet was developed for each participant to enable the individual participant to compare their own results with those of their peers. The feedback report and personal feedback sheet were distributed to the healthcare providers prior to a one-day feedback workshop held in August 2015. Here, the feedback report was presented and discussed. The findings were put in context with preliminary results from the national tuberculosis prevalence survey, relevant studies on the area, and the available guidelines for tuberculosis screening. The workshop further comprised group work sessions to identify challenges for adherence to guidelines and possible areas for quality improvements. The principal investigator (SB), a consultant for the APO method (LB), and key persons from the NTP (FB and NNHN) and NACP facilitated the feedback workshop. Fig. 1 Feedback report. Study outcome measures and definitions The primary outcome measure was to determine and characterize healthcare providers’ tuberculosis suspicion rate. We defined a series of secondary outcome measures based on recommended standards for tuberculosis screening of PLHIV and adherence rates to these (Box 1). The performance standards were selected in collaboration with key informants at NTP and NACP and were based on the ISTC and WHO guidelines (9, 10). A positive WHO tuberculosis symptom screen (‘WHO-TB’) is defined as PLHIV presenting with any of the following four symptoms: fever, cough (of any duration, i.e. both the variable cough <2 weeks and cough ≥2 weeks), night sweat, or weight loss (9). Box 1 Outcome measures compared to recommendations Primary outcome measure  Tuberculosis suspicion rate Secondary outcome measures Recommended standards Proportion of individuals with cough ≥2 weeks suspected of tuberculosis ISTC, Standard 1: All persons with otherwise unexplained cough lasting 2–3 weeks or more should be evaluated for tuberculosisa Proportion of individuals with cough ≥2 weeks referred for sputum smear microscopy Proportion of individuals who are suspected of tuberculosis referred for sputum smear microscopy ISTC, Standard 2: All patients (adults, adolescents, and children who are capable of producing sputum) suspected of having pulmonary tuberculosis should have at least two, and preferably three, sputum specimens obtained for microscopic examinationa Proportion of individuals with a positive WHO-TB screen suspected of TB Proportion of individuals with a positive WHO-TB screen referred for sputum smear microscopy WHO: Adults and adolescents living with HIV and screened for TB with a clinical algorithm and who report any one of the symptoms of current cough, fever, weight loss, or night sweats may have active TB and should be evaluated for TB and other diseasesb¶ a International Standards for Tuberculosis Care, 2006 (10). b WHO guidelines for tuberculosis screening among PLHIV, 2011 (9). Statistical analysis Descriptive analysis was used to characterize the participating healthcare providers and the study population using chi-square test for categorical variables and Wilcoxon rank-sum for continuous variables. We used logistic regression models with robust standards errors (SE) to obtain point estimates for the outcome measures with 95% confidence intervals (CI) pre- and post-feedback. As feedback targeted the healthcare providers and not the HIV care clinic, we adjusted for clustering of registrations by the same provider. Difference in outcome measures was analysed for statistical significance using logistic regression with results from the first audit used as reference. Statistical significance was defined as a two-sided p-value less than 0.05, and all analyses were conducted using the statistical software package STATA™ (version 13.1). Ethical clearance The project was approved by the Ghana Health Service Ethical Review Committee (GHS-ERC: 09/03/14), the Ethical and Protocol Review Committee, University of Ghana Medical School (MS-Et./M.4–P 3.3/2012-13) and evaluated by the Developing Country Committee of the Danish National Committee on Health Research Ethics (No. 1302133/1206169). The healthcare providers signed a written informed consent form before participation. We did not register any patient identifiable information apart from age and sex nor expose patients to any intervention, and hence, there were no requirements to collect informed consent from patients contributing data. Results Healthcare providers’ details Healthcare providers (n=22) from 10 HIV care clinics participated in the study including 10 medical doctors, eight nurses, three physician assistants, and one disease control officer. Registrations from one physician assistant were excluded from analysis, since the person failed to register according to instructions. The healthcare providers reported a medium of 5 years (IQR 2–10) of expertise in a HIV care clinic and had a median of 40 (30–70) outpatient consultations per day. All providers received feedback and participated in the feedback workshop; 16 providers participated in both audits. Consultations Data from 2,666 consultations with PLHIV were included in the study based on registration of 1,368 consultations in the first audit and 1,298 consultations in the second audit after performance feedback (Table 1). Overall, consultations comprised of 2,185 (81.9%) follow-up visits and 439 (16.5%) initial assessments of new HIV-positive patients. The majority of patients consulted were females 1,910 (71.6%) and most were receiving ART 1,869 (70.1%). Baseline variables differed between the two audits with regard to consultation type, sex of the individuals, proportion of patients that had defaulted ART, and patients presenting with weight loss or cough <2 weeks. After adjusting for clustering to healthcare provider, only the sex of the patients remained significantly different between the audits. Table 1 Consultation and patient characteristics at first and second audit Overall (N=2,666) First audit (N=1,368) Second audit (N=1,298) n (%) n % n % Consultation  Initial 439 (16.5) 201 14.7 238 18.3  Follow-up 2,185 (81.9) 1,140 83.3 1,045 80.5  Unknown 42 (1.6) 27 2.0 15 1.2 Sex  Male 741 (27.8) 418 30.6 323 24.9  Female 1,910 (71.6) 943 68.9 967 74.5  Unknown 15 (0.6) 7 0.5 8 0.6 Age in years  Median (IQR) 40 (33–48) 41 (33–48) 40 (33–48) HIV Treatment status  Receiving ART 1,869 (70.1) 940 68.7 929 71.6  ART naive 580 (21.8) 291 21.3 289 22.2  Defaulted ART (>1 month) 78 (2.9) 51 3.7 27 2.1  Unknown 139 (5.2) 86 6.3 53 4.1 Signs and symptoms  Weight loss 374 (14.0) 167 12.2 207 16.0  Fever 397 (14.9) 199 14.6 198 15.3  Cough <2 weeks 337 (12.6) 146 10.7 191 14.7  Cough ≥2 weeks 298 (11.2) 144 10.5 154 11.9  Night sweats 172 (6.5) 80 5.9 92 7.1 Positive WHO-TB screena 908 (34.1) 444 32.5 464 35.8 ART, antiretroviral therapy. a Positive WHO symptoms screen if presence of any of the following symptoms: current cough, fever, weight loss, or night sweat (9). Tuberculosis suspicion rate In the first audit, tuberculosis was suspected in 172/1,368 (12.6%) consultations with rates varying from 0.5–35.7% across the healthcare providers. The tuberculosis suspicion rate was higher in consultations with new HIV-positive patients than at follow-up (28.4% vs. 9.9%, odds ratio, OR 3.60; 95% CI: 2.33–5.55). Moreover, patients not receiving ART and patients who had defaulted ART for more than 1 month were more likely to be suspected of tuberculosis than patients receiving ART. Patients aged above 55 years were less likely to be suspected for tuberculosis. The tuberculosis suspicion rate appeared substantially lower for doctor-led consultations than for consultations led by the other staff categories combined, but the difference was borderline significant when accounting for clustering (15.5% vs. 9.5% OR 1.75; 95% CI: 0.99–3.11) (Table 2). Table 2 Tuberculosis suspicion rate overall and by subgroups shown for the first and second audit First audit Second audit Proportion of TB suspects n/N % ORa 95% CI P n/N % ORa 95% CI P Overall 172/1,368 12.6 271/1,298 20.9% By staff category  Medical doctors 63/665 9.5 ref 65/411 15.8% ref  Non-doctors 109/703 15.5 1.75 0.99 3.11 0.054 206/887 23.2% 1.61 0.54 4.79 0.391 By consultation  Follow-up 113/1,140 9.9 ref 140/1,045 13.4% ref  Initial 57/201 28.4 3.60 2.33 5.55 <0.001 128/238 53.8% 7.52 3.94 14.34 <0.001 By patient sex  Male 60/418 14.4 ref 84/323 26.0% ref  Female 111/943 11.8 0.80 0.54 1.18 0.251 186/967 19.2% 0.68 0.52 0.88 0.004 By age (years)  0–17 10/45 22.2 1.51 0.56 4.06 0.416 5/25 20.0% 0.66 0.23 1.87 0.433  18–34.9 57/358 15.9 ref 92/335 27.5% ref  35–54.9 87/794 11.0 0.65 0.39 1.07 0.093 146/768 19.0% 0.62 0.39 0.99 0.044  ≥55 17/170 10.0 0.59 0.38 0.92 0.020 27/167 16.2% 0.51 0.25 1.06 0.069 By ART status  Receiving ART 87/940 9.3 ref 113/929 12.2% ref  ART naive 58/291 19.9 2.44 1.47 4.05 0.001 123/289 42.6% 5.35 3.09 9.27 <0.001  Defaulted ART (>1 month) 14/51 27.5 3.71 1.72 8.02 0.001 15/27 55.6% 9.03 3.14 25.91 <0.001 ART, antiretroviral treatment. a OR: Odds ratio adjusted for clustering of registrations within health provider. Non-doctors includes nurses (n=8), physician assistants (n=2), and disease control officer (n=1). Missing values excluded from analysis; sex (n=15), age (n=4), consultation (n=42), ART status (n=139). P-values in bold indicate values <0.05. In the second audit, tuberculosis was suspected in 271/1,298 (20.9%) consultations. Tuberculosis suspicion rates were higher in consultation with patients of male sex, patients newly diagnosed with HIV, patients not receiving ART, or patients who defaulted ART. Patients in the age groups 35–55 years and above were less likely to be suspected of tuberculosis than patients aged 18–34 years, although borderline significant for the age group above 55 years (Table 2). The increase in tuberculosis suspicion rate from 12.6 to 20.9% after feedback was significant (OR 1.83; 95% CI: 1.09–3.09) (Fig. 2). Fig. 2 Tuberculosis suspicion rate and adherence to standards for tuberculosis screening and referral for sputum smear microscopy. (a) Tuberculosis suspicion rate among healthcare providers at first audit and second audit. (b) Referral rate for sputum smear microscopy among healthcare providers at first audit and second audit. Percentages are given as n/N. OR: Odds ratio shown with results from audit 1 as reference, adjusted for clustering of registrations within the health provider. WHO-TB+: Positive WHO symptoms screen defined as presence of any of the following symptoms; current cough, fever, weight loss or night sweats (9). Missing values excluded from analysis; sex (n=15), age (n=4), consultation (n=42), ART status (n=139). Adherence to standards for tuberculosis screening and referral for sputum smear microscopy The first audit highlighted wide gaps between practices of tuberculosis screening and applicable performance standards. The healthcare providers suspected tuberculosis in 87/144 (60.4%) patients presenting with cough ≥2 weeks and in 166/444 (37.4%) patients with a positive WHO-TB screen. Sputum smear microscopy was requested for 101/172 (58.7%) patients suspected of tuberculosis, for 68/144 (47.2%) patients presenting with cough ≥2 weeks, and for 122/444 (27.5%) patients with a positive WHO-TB screen (Fig. 2). In the second audit, providers’ tuberculosis suspicion rate increased from 60.4 to 73.4% in patients presenting with cough ≥2 weeks and from 37.4 to 56.9% in patients with a positive WHO-TB screen. The increase in tuberculosis suspicion rate was statistically significant for patients with a positive WHO-TB screen (OR 2.21; 95% CI: 1.19–4.09), but not for those with cough ≥2 weeks (OR 1.81; 95% CI: 0.75–4.36). Referral for sputum smear microscopy increased to 74.9% in patients suspected of tuberculosis, to 57.8% in patients presenting with cough ≥2 weeks, and to 50.7% in patients with a positive WHO-TB screen. The increase was significant in referral of patients with a positive WHO-TB screen (OR 2.71; 95% CI: 1.25–5.86), but not for the other groups. Discussion In this study, a simple tool for audit and feedback was integrated into the daily work of healthcare providers at 10 large HIV care clinics in Ghana. The first audit identified low tuberculosis suspicion rates and a marked gap between current practices and widely recommended standards for tuberculosis screening. After feedback to healthcare providers, we observed increased tuberculosis suspicion rates and improved the quality of tuberculosis screening performance. To our knowledge, this is the first study to prospectively evaluate real-life practices for tuberculosis screening at multiple HIV care clinics. We found that providers’ initial tuberculosis suspicion rate was only 12.6% in the first audit. We consider this to be low, especially in the context of the high tuberculosis prevalence in Ghana (4) and the high rates of HIV/tuberculosis co-infections observed in resource-limited settings (28). The prevalence of tuberculosis among PLHIV in Ghana is not systematically described, but we have recently reported a culture-confirmed tuberculosis prevalence of 12.7% among PLHIV eligible for ART from a large teaching hospital in Ghana (29). In this study, tuberculosis suspicion rate was highest in consultation with the younger and newly diagnosed HIV-positive patients. Tuberculosis suspicion rate was very low for patients receiving ART (9%) and at follow-up consultations (10%), although PLHIV remain at risk of tuberculosis throughout the course of HIV disease also after initiating ART (30–32). Our study showed that healthcare providers deviate from recommended guidelines and standards, despite these being evidence-based. Providers did not consistently suspect tuberculosis in patients presenting with cough ≥2 weeks and less so in patients with any of the four symptoms included in the WHO tuberculosis symptom screen. While prolonged cough previously has been identified as the most common symptom screened for in relation to HIV-associated tuberculosis (14), the more broadly defined WHO-TB symptoms seem less well recognised by the healthcare providers as predictive for tuberculosis. Healthcare providers failed to refer a large proportion of patients suspected of tuberculosis for sputum smear microscopy. Other studies have also recognised low referral rates for tuberculosis diagnostic tests as a ‘leaky’ step in the tuberculosis diagnostic cascade (13, 15). Further down the diagnostic cascade, additional leaky steps have been identified including low rates for completion of the tuberculosis diagnostic tests (33, 34) and high rates of individuals that never start treatment despite a positive test (35). Screening guidelines encourage that any individual with a positive WHO-TB screen, including individuals with cough ≥2 weeks, should be referred for further diagnostic tests regardless of whether the health provider suspects tuberculosis (9). We found very low referral rates for patients with a positive WHO-TB screen (27.5%). This could reflect that the WHO-TB screening algorithm is less well integrated at HIV care sites for intensified tuberculosis case finding. Healthcare providers may be reluctant to overburden the often already challenged tuberculosis laboratories. In our study, more than one-third (34.1%) of the patients presented with a positive WHO-TB screen. In studies of HIV-infected individuals eligible to start ART, up to 90% had a positive WHO-TB screen (36, 37) and the consequence of referring all for sputum microscopy could be dire for a fragile health system. The WHO's TB-screening guideline favours a high negative predictive value to identify those individuals, unlikely to have tuberculosis (i.e. none of the four symptoms reported) where IPT could be started to prevent tuberculosis (7, 9). IPT to PLHIV is not programmatically implemented in Ghana, and this may further contribute to low adherence to the WHO-TB-screening guidelines. Major progress has been made in developing accurate and rapid diagnostic tools like the Xpert® MTB/RIF (Cepheid, USA). However, implementation of diagnostic tools like the Xpert MTB/RIF has not yet been able to demonstrate reduced mortality in randomised control trials in Africa, despite the great potentials of the test (38, 39). The trials emphasised that the impact of even the most promising tuberculosis diagnostic test is compromised if not coupled to good standards of clinical care (40). The first steps for tuberculosis screening to be effective, regardless of the test, are that healthcare providers suspect tuberculosis in individuals with relevant signs and symptoms and refer them for further diagnostic tests. In our study, inadequate tuberculosis suspicion and low referral rates for sputum microscopy represent low standards of HIV care and a missed opportunity for tuberculosis case detection. We observed that tuberculosis suspicion rates among healthcare providers increased and standards of care improved with audit and feedback. However, we noted a heterogeneous tuberculosis suspicion rate across healthcare providers, and whereas some providers improved performance significantly after feedback, others did not change practice for tuberculosis screening. Furthermore, there was a trend of higher tuberculosis suspicion among nurses than medical doctors, although it levelled out in the second audit. Moreover, tuberculosis suspicion rate was significantly higher in males in the second audit while not in the first audit, and also in age groups, there was a shift in significant differences. Feedback was, in our study, not tailored to specific staff categories or site-specific barriers for tuberculosis screening. It is possible that audit and feedback, as part of a multi-faceted intervention with targeted initiatives, could improve performance further (16). The tool we used for audit and feedback was simple and operational in the daily work of the healthcare providers and effective to identify quality concerns in provider practices. The same tool could be used to monitor and ensure changes in practices after implementation of other interventions to improve tuberculosis case finding. In Ghana, the NTP has recently modified the algorithm for systematic tuberculosis screening of all outpatient attendees. The audit and feedback tool described here is currently being adapted for programmatic implementation within the health system context and challenges to evaluate the revised systematic tuberculosis screening of outpatients. Our study has limitations. Healthcare providers had to dedicate extra time to fill in the registration form after consultation. In a setting with a high patient flow, this might be difficult and may compromise participation in the audit and quality of entries. Therefore, we instructed the participants to register only the first 15 consultations per day and limited the requirement for details so that entries in the form took <2 minutes. We did not include patient identifiable information and could, therefore, not ascertain if the same patient was entered several times. However, we aimed to characterize the consultation practices rather than the patients. Furthermore, we did not have the capacity to follow up on patients to report the actual test results for patients referred and cannot provide information on the rates for completing a test or starting treatment for tuberculosis given a positive test. Another limitation is that participants may have registered what they perceived as the correct practice rather than what truly happened. This could have been avoided by direct observation of practices that was out of scope for this study. The design of our study prevents us to draw causal inferences, and other factors than the audit and intervention could be responsible for the change seen in performance. Some of the changes observed could be related to concurrent activities to increase tuberculosis case detection or may reflect the increased focus on tuberculosis screening aroused by the national prevalence survey. In future studies, it could be beneficial to include a control group or a qualitative assessment of contextual factors that could have affected the changes in performance observed in our study. Conclusion A simple audit tool for quality development identified a low tuberculosis suspicion rate and substandard performance of healthcare providers’ tuberculosis screening practices at HIV care clinics in Ghana. Performance improved after audit and feedback to the healthcare providers, in particular for adherence to the WHO guidelines for tuberculosis screening of PLHIV. Flaws in healthcare provider practices for tuberculosis screening and referral for sputum smear microscopy at this level represent a lost opportunity for tuberculosis case detection and inadequate HIV care. To harvest the full benefit of new promising diagnostic technologies, the practices of healthcare providers must come more into focus and effort must be placed on identifying and closing gaps in the quality of clinical care. Supplementary Material Tuberculosis screening in patients with HIV: use of audit and feedback to improve quality of care in Ghana Click here for additional data file. Acknowledgements The authors are grateful to the clinicians involved in the study and the heads of Department for providing the necessary support to carry out the study. They further thank the National AIDS Control Programme, Ghana, for their support of the study and their active engagement during feedback. They appreciate funding received from Odense University Hospital (Denmark), University of Southern Denmark, Research Council of Region Southern Denmark, Augustinus Foundation (Denmark). Authors’ contributions SB, LB, AM, ABA, ISJ, and FB conceptualised and designed the study in common. SB, LB, EK, DJ, and AM were responsible for data collection and management. SB, FB, NNHN, and LB were involved in active participation and designed the feedback intervention. SB, LB, and AM contributed to the statistical analysis. All authors contributed to the interpretation of results and writing of the manuscript. Conflict of interest and funding The authors have declared no conflicts of interest related to this study, its findings, or this manuscript. The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the manuscript. Paper context The success of improved tuberculosis diagnostic test relies on healthcare providers’ practices and adherence to evidence-based tuberculosis screening guidelines. We found low tuberculosis suspicion rates among healthcare providers at HIV care clinics in Ghana compounded by low referral rates for sputum smear microscopy. Adherence to guidelines improved with audit and feedback. Our study proposes a simple method of prospective audit and feedback to quantify substandard performance and improve the quality of clinical practices for tuberculosis screening. ==== Refs References 1 Gupta A Nadkarni G Yang WT Chandrasekhar A Gupte N Bisson GP Early mortality in adults initiating antiretroviral therapy (ART) in low- and middle-income countries (LMIC): a systematic review and meta-analysis PLoS One 2011 6 e28691 22220193 2 Gupta RK Lucas SB Fielding KL Lawn SD Prevalence of tuberculosis in post-mortem studies of HIV-infected adults and children in resource-limited settings: a systematic review and meta-analysis AIDS 2015 29 1987 2002 26266773 3 Ghana Aids Commision Country AIDS response progress report, Ghana 2012–2013. Contract No. 15/05/2015 2014 Accra, Ghana Ghana AIDS Commission, UNAIDS, unadis.org 4 World Health Organization Tuberculosis country profile, Ghana 2014 Geneva World Health Organization Available from: https://extranet.who.int/sree/Reports?op=Replet&name=%2FWHO_HQ_Reports%2FG2%2FPROD%2FEXT%2FTBCountryProfile&ISO2=GH&LAN=EN&outtype=html [cited 20 February 2016] 5 National Tuberculosis Control Programme Review Report The national tuberculosis health sector strategic plan for Ghana 2009–2013 2009 Accra, Ghana Ministry of Health 6 Corbett EL MacPherson P Tuberculosis screening in high human immunodeficiency virus prevalence settings: turning promise into reality Int J Tuberc Lung Dis 2013 17 1125 38 23928165 7 Getahun H Kittikraisak W Heilig CM Corbett EL Ayles H Cain KP Development of a standardized screening rule for tuberculosis in people living with HIV in resource-constrained settings: individual participant data meta-analysis of observational studies PLoS Med 2011 8 e1000391 21267059 8 Date A Modi S TB screening among people living with HIV/AIDS in resource-limited settings J Acquir Immune Defic Syndr 2015 68 Suppl 3 S270 3 25768866 9 World Health Organization, Stop TB Dept. DoHA Guidelines for intensified tuberculosis case-finding and isoniazid preventive therapy for people living with HIV in resource-constrained settings 2011 Geneva World Health Organization vii, 39 10 Tuberculosis Coalition for Technical Assistance, World Health Organization International standards for tuberculosis care: diagnosis, treatment, public health 2006 The Hague Tuberculosis Coalition for Technical Assistance 57 11 Hopewell PC Fair EL Uplekar M Updating the International Standards for Tuberculosis Care. Entering the era of molecular diagnostics Ann Am Thorac Soc 2014 11 277 85 24673691 12 Thanprasertsuk S Supawitkul S Lolekha R Ningsanond P Agins BD McConnell MS HIVQUAL-T: monitoring and improving HIV clinical care in Thailand, 2002–08 Int J Qual Health Care 2012 24 338 47 22665387 13 Zaeh S Kempker R Stenehjem E Blumberg HM Temesgen O Ofotokun I Improving tuberculosis screening and isoniazid preventive therapy in an HIV clinic in Addis Ababa, Ethiopia Int J Tuberc Lung Dis 2013 17 1396 401 24125440 14 Auld AF Mbofana F Shiraishi RW Alfredo C Sanchez M Ellerbrock TV Incidence and determinants of tuberculosis among adults initiating antiretroviral therapy–Mozambique, 2004–2008 PLoS One 2013 8 e54665 23349948 15 Van Rie A Clouse K Hanrahan C Selibas K Sanne I Williams S High uptake of systematic HIV counseling and testing and TB symptom screening at a primary care clinic in South Africa PLoS One 2014 9 e105428 25268851 16 Jamtvedt G Young JM Kristoffersen DT O'Brien MA Oxman AD Audit and feedback: effects on professional practice and health care outcomes Cochrane Database Syst Rev 2006 2 CD000259 16625533 17 Siddiqi K Newell J Robinson M Getting evidence into practice: what works in developing countries? Int J Qual Health Care 2005 17 447 54 15872024 18 Harries AD Hargreaves NJ Kwanjana JH Salaniponi FM Clinical diagnosis of smear-negative pulmonary tuberculosis: an audit of diagnostic practice in hospitals in Malawi Int J Tuberc Lung Dis 2001 5 1143 7 11769773 19 Siddiqi K Volz A Armas L Otero L Ugaz R Ochoa E Could clinical audit improve the diagnosis of pulmonary tuberculosis in Cuba, Peru and Bolivia? Trop Med Int Health 2008 13 566 78 18318698 20 Ogrinc G Davies L Goodman D Batalden P Davidoff F Stevens D SQUIRE 2.0 (Standards for QUality Improvement Reporting Excellence): revised publication guidelines from a detailed consensus process BMJ Qual Saf 2015 1 7 21 Ministry of Health Ghana Implementation guideline and scale-up plan for Xpert MTB/RIF Assay in Ghana 2015–202 2015 Accra, Ghana Ministry of Health 22 Ghana Health Services Guidelines for the clinical management of TB and HIV co-infection in Ghana 2007 Accra, Ghana Ghana Health Services 23 Ghana Health Service NTP training manual National Tuberculosis Control Programme (NTP) 2012 Accra, Ghana Ghana Health Service 24 Munck AP Damsgaard JJ Hansen DG Bjerrum L Sondergaard J The APO method–a popular form of quality development in general practice Ugeskr Laeger 2002 164 5390 3 12469386 25 Bjerrum L Munck A Gahrn-Hansen B Hansen MP Jarbol DE Cordoba G Health alliance for prudent antibiotic prescribing in patients with respiratory tract infections (HAPPY AUDIT) -impact of a non-randomised multifaceted intervention programme BMC Fam Pract 2011 12 52 21689406 26 Gahm-Hansen B Aaen-Larsen B Mosgaard L Damsgaard J Munck A Respiratory tract infections in Greenland: results of an audit project Int J Circumpolar Health 2004 63 Suppl 2 209 13 15736654 27 Ingebrigtsen SG Scheel BI Hart B Thorsen T Holtedahl K Frequency of ‘warning signs of cancer’ in Norwegian general practice, with prospective recording of subsequent cancer Fam Pract 2013 30 153 60 23097250 28 Lawn SD Wood R Tuberculosis in antiretroviral treatment services in resource-limited settings: addressing the challenges of screening and diagnosis J Infect Dis 2011 204 Suppl 4 S1159 67 21996698 29 Bjerrum S Oliver-Commey J Kenu E Lartey M Newman MJ Addo KK Tuberculosis and non-tuberculous mycobacteria among HIV-infected individuals in Ghana Trop Med Int Health 2016 doi: http://dx.doi.org/10.1111/tmi.12749 . [Epub ahead of print] 30 Lawn SD Kranzer K Edwards DJ McNally M Bekker LG Wood R Tuberculosis during the first year of antiretroviral therapy in a South African cohort using an intensive pretreatment screening strategy AIDS 2010 24 1323 8 20386425 31 Lawn SD Badri M Wood R Tuberculosis among HIV-infected patients receiving HAART: long term incidence and risk factors in a South African cohort AIDS 2005 19 2109 16 16284460 32 Gupta A Wood R Kaplan R Bekker LG Lawn SD Tuberculosis incidence rates during 8 years of follow-up of an antiretroviral treatment cohort in South Africa: comparison with rates in the community PLoS One 2012 7 e34156 22479548 33 Elden S Lawes T Kudsk-Iversen S Vandelanotte J Nkawanyana S Welfare W Integrating intensified case finding of tuberculosis into HIV care: an evaluation from rural Swaziland BMC Health Serv Res 2011 11 118 21605437 34 Kanara N Cain KP Laserson KF Vannarith C Sameourn K Samnang K Using program evaluation to improve the performance of a TB-HIV project in Banteay Meanchey, Cambodia Int J Tuberc Lung Dis 2008 12 Suppl 1 44 50 18302822 35 Botha E den Boon S Lawrence KA Reuter H Verver S Lombard CJ From suspect to patient: tuberculosis diagnosis and treatment initiation in health facilities in South Africa Int J Tuberc Lung Dis 2008 12 936 41 18647454 36 Bjerrum S Kenu E Lartey M Newman MJ Addo KK Andersen AB Diagnostic accuracy of the rapid urine lipoarabinomannan test for pulmonary tuberculosis among HIV-infected adults in Ghana-findings from the DETECT HIV-TB study BMC Infect Dis 2015 15 407 26427365 37 Balcha TT Skogmar S Sturegard E Schon T Winqvist N Reepalu A A clinical scoring algorithm for determination of the risk of Tuberculosis in HIV-infected adults: a cohort study performed at Ethiopian health centers Open Forum Infect Dis 2014 1 ofu095 25734163 38 Churchyard GJ Stevens WS Mametja LD McCarthy KM Chihota V Nicol MP Xpert MTB/RIF versus sputum microscopy as the initial diagnostic test for tuberculosis: a cluster-randomised trial embedded in South African roll-out of Xpert MTB/RIF Lancet Glob Health 2015 3 e450 7 26187490 39 Theron G Zijenah L Chanda D Clowes P Rachow A Lesosky M Feasibility, accuracy, and clinical effect of point-of-care Xpert MTB/RIF testing for tuberculosis in primary-care settings in Africa: a multicentre, randomised, controlled trial Lancet 2014 383 424 35 24176144 40 Rudolf F Wejse C Tuberculosis case detection revisited: better testing might not improve outcomes Lancet Glob Health 2015 3 e424 5 26187481

PMC005xxxxxx/PMC5002399.txt

==== Front Food Nutr ResFood Nutr ResFNRFood & Nutrition Research1654-661XCo-Action Publishing 3104210.3402/fnr.v60.31042Original ArticleA nine-country study of the protein content and amino acid composition of mature human milk Feng Ping 1Gao Ming 2Burgher Anita 2Zhou Tian Hui 3Pramuk Kathryn 2*1 Wyeth Nutrition, Shanghai, China2 Wyeth Nutrition, King of Prussia, PA, USA3 Bio TX Clinical Research, Pfizer, Inc., Collegeville, PA, USA* Kathryn Pramuk, Nestlé Nutrition R&D, 3000 Horizon Drive, Suite 100, King of Prussia, PA 19406, USA. Email: Kathryn.PramukMari@rd.nestle.com26 8 2016 2016 60 10.3402/fnr.v60.3104218 1 2016 15 7 2016 © 2016 Ping Feng et al.2016This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License, allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material for any purpose, even commercially, provided the original work is properly cited and states its license.Background Numerous studies have evaluated protein and amino acid levels in human milk. However, research in this area has been limited by small sample sizes and study populations with little ethnic or racial diversity. Objective Evaluate the protein and amino acid composition of mature (≥30 days) human milk samples collected from a large, multinational study using highly standardized methods for sample collection, storage, and analysis. Design Using a single, centralized laboratory, human milk samples from 220 women (30–188 days postpartum) from nine countries were analyzed for amino acid composition using Waters AccQ-Tag high-performance liquid chromatography and total nitrogen content using the LECO FP-528 nitrogen analyzer. Total protein was calculated as total nitrogen×6.25. True protein, which includes protein, free amino acids, and peptides, was calculated from the total amino acids. Results Mean total protein from individual countries (standard deviation [SD]) ranged from 1,133 (125.5) to 1,366 (341.4) mg/dL; the mean across all countries (SD) was 1,192 (200.9) mg/dL. Total protein, true protein, and amino acid composition were not significantly different across countries except Chile, which had higher total and true protein. Amino acid profiles (percent of total amino acids) did not differ across countries. Total and true protein concentrations and 16 of 18 amino acid concentrations declined with the stage of lactation. Conclusions Total protein, true protein, and individual amino acid concentrations in human milk steadily decline from 30 to 151 days of lactation, and are significantly higher in the second month of lactation compared with the following 4 months. There is a high level of consistency in the protein content and amino acid composition of human milk across geographic locations. The size and diversity of the study population and highly standardized procedures for the collection, storage, and analysis of human milk support the validity and broad application of these findings. human milkamino acidproteinnitrogenlactation stage ==== Body Human milk is considered the best source of nutrition for term infants. The World Health Organization recommends exclusive breast feeding during the first 6 months of life (1). Within the nutrient-rich matrix of human milk, the quantity and quality of protein are vitally important to provide the infant with a source of peptides, amino acids, and nitrogen for visceral protein synthesis, tissue accretion, and growth. Additionally, human milk provides the amino acids required to synthesize hormones, enzymes, antibodies, and other compounds such as glutathione, nucleotides, and some neurotransmitters (2). Numerous studies have evaluated protein and amino acid levels in human milk. The earliest studies yielded widely divergent findings that were attributed to variability among donors with respect to age, parity, and duration of lactation, as well as differences in the collection and storage of human milk samples, and methods of analysis (3). The introduction of the automated amino acid analyzer (4) represented a clear improvement in methodology that resulted in more consistent data on the protein and amino acid composition of human milk (5–21). Despite such advances in analytical methods, research on the protein and amino acid content of human milk has been limited by small sample sizes and homogeneous study populations. Moreover, studies differ with respect to sample collection, storage, and methods of analysis, all of which can introduce variability to the measurement of protein and amino acid levels in human milk. To our knowledge, this study is the largest, multinational study of protein levels and amino acid composition in human milk. A review of the published literature of total protein and amino acid composition of human milk from various regions and varying collection techniques is summarized in Table 1. In our study, milk samples from 220 women from nine countries across five continents were analyzed for amino acid composition, total nitrogen, and true protein concentration, using a unified protocol and standardized methodology. Therefore, the potential variability inherent from any differences in the sample collection, handling, storage, shipping procedures, and sample analyses was essentially eliminated, thereby increasing our confidence that the variations in our data reflect true biological variations among the samples. Enrolling mothers over a broad range of days, post-partum, permitted the assessment of amino acid and protein levels across several stages of lactation. It has been shown previously that the protein level and amino acid content in human milk decrease over the course of lactation (20), whereas maternal race/ethnicity, age, and maternal dietary protein intake appear to have little effect on the total protein in human milk (22). Table 1 Referenced studies of human milk amino acid and/or total protein composition Authors Location Number of mothers Lactation time Reference Total protein (g/L)a Lonnerdal et al. Sweden 6 2–3 months (5) 10.7+1.34 Raiha et al. Italy 10 Mature (6) 12.0 Picone et al. USA 12 pooled Mature (7) (AA) Atkinson et al. Canada 8 pooled Mature (8) (AA) Renner Germany N/S N/S (9) 8.5 (no SD) USDA N/S N/S N/S (10) 10.3 Svanberg et al. Sweden 8 2–5 months (11) 10.1+0.88 Ethiopia 8 2–5 months (11) 11.4+2.84 Jarvenpaa et al. Sweden Pooled N/S (12) 9.6 Harzer and Bindels Germany N/S 36 days (13) 11.0 Donovan and Lonnerdal USA 5 Mid-lactation (14) 10.4+0.64 Hanning et al. N/S N/S 28–30 days (15) 10.0 Motil et al. USA 24 1–12 months (17) 10.6+1.34 Darragh New Zealand 20 10–14 weeks (18) 11.5+0.19 (SE) Zhao et al. China 91 1–6 months (19) 11.8+0.14 Wu et al. Taiwan, China 105 46–297 days (20) 12.6b (no SD) Feng et al. Nine countries 220 30–188 days Current study 11.9+0.20c N/S, not specified; AA, amino acid composition only. a Total protein=total nitrogen×6.38 for most studies (some values derived from biochemical assay). b Total protein only includes samples with lactation days between 46 and 297 days. c Mean total protein concentration in human milk calculated using 6.25 as the factor=11.9 (range 8.5–22.9 g/L); mean total protein concentration in human milk calculated using 6.38 as the factor=12.2 (range 8.9–23.4 g/L). By using a single commercial entity for shipping, a single commercial laboratory for sample storage, a single research laboratory for sample analysis, and standardized milk collection procedures, our methodology was highly consistent across sites and assures a reliable data set. Materials and methods Study design and subjects The human milk samples were collected as part of a cross-sectional survey of major carotenoids (23) and fatty acids (24) in human milk from healthy, well-nourished lactating women in nine countries: Australia, Canada, Chile, China, Japan, Mexico, the Philippines, the United Kingdom, and the United States. All participants were aged 18–40 years; were mothers of a healthy, full-term singleton infant; and were between 1 and 12 months postpartum at the time of milk collection. Participants signed written, informed consent in their native language prior to enrollment in the study, and the same two individuals conducted on-site training for all study personnel. The study was conducted in accordance with the principles of the Declaration of Helsinki and was approved by the Human Subjects Committee of the University of Arizona and the Human Ethics Committees associated with each participating institution. A minimum of 50 human milk samples were collected from each of the nine countries represented in the study, for a total of 509 samples. Of these, 445 samples were from women who were 30–188 days (1–6 months) postpartum at the time of milk collection. Within this subset of 445 human milk samples, 220 were randomly selected for an analysis of amino acid and nitrogen composition. The random selection of samples was stratified by country to ensure the inclusion of at least 20 samples from each country. Collection and handling of human milk samples Our collection and handling of the human milk samples has been described in detail (23). Briefly, a complete breast expression containing a minimum of 50 mL of human milk was collected between 1:00 p.m. and 5:00 p.m. on the day of the sampling. In all countries except Japan, samples were collected using an electric pump. In Japan, where the use of an electric pump is considered unacceptable to women, samples were collected manually using a hand-held breast pump under the supervision of clinic staff. After collection, milk samples were immediately placed on dry ice or in a freezer at −20°C, then shipped within 10–14 days on dry ice to a central laboratory where the samples were stored at −70°C. Prior to analysis, frozen samples from a single country were thawed overnight in the refrigerator. Under subdued lighting conditions, samples were warmed to 37°C in a water bath and gently stirred, and approximately 10 mL of each sample was transferred to a tube and stored at −70°C for a group analysis. One day before amino acid and nitrogen analyses, samples were thawed overnight in the refrigerator. Thawed samples were warmed to 37°C in a water bath and gently stirred, before subsequent analysis. The analysis of samples was grouped daily by randomly choosing two samples from each country to eliminate day-to-day bias. General amino acid analysis To determine the concentration of 16 amino acids in the human milk samples, 10 mL of 6 M HCl (containing 0.1% phenol) was added to a hydrolysis tube that contained 1 mL of human milk. Following vacuum and nitrogen flush, repeated three times, the tube was sealed under a nitrogen blanket and the sample hydrolyzed by placing in an oven at 110°C for 24 h. The hydrolysate was then quantitatively transferred to a volumetric flask and made up to a volume of 50 mL using distilled water. A total of 15 µL of filtered hydrolysate solution was quantitatively pipetted into a derivatization tube, dried under vacuum, and then combined with alpha-amino butyric acid as an internal standard and analyzed using AccQ-Tag (Waters Corporation, Milford, MA) derivatization and high-performance liquid chromatography (HPLC) (25, 26). Cysteine analysis Distilled water was added to 1 mL of the milk sample in a 50-mL volumetric flask; 15 µL of the solution was quantitatively pipetted into a derivatization tube. The sample was dried under vacuum and oxidized with performic acid for the conversion of cysteine and cystine into cysteic acid; this was followed by vapor phase acidic hydrolysis using a boiling HCl solution at 110°C for 24 h. The sample was combined with alpha-amino butyric acid as an internal standard and then analyzed for cysteic acid using AccQ-Tag derivatization and HPLC (25, 26). Tryptophan analysis A total of 10 mL of 4.2 M NaOH solution was added to a hydrolysis tube that contained 3 mL of human milk. In addition, 800 µL of 5-methyl-tryptophan was added to the hydrolysis tube as an internal standard. Following vacuum and nitrogen flush repeated three times, the tube was sealed under vacuum and placed in an oven at 110°C for 20 h to hydrolyze the sample. Following adjustment to pH 4.2 using 12 M HCl, centrifugation, and filtration, tryptophan was determined using reversed-phase HPLC. Nitrogen analysis Total nitrogen was determined by complete combustion of each human milk sample using the LECO FP-528 nitrogen analyzer (LECO Corporation, St. Joseph, MI). An infant formula standard (NIST infant formula reference material 1846) solution with a nitrogen content of 0.221% (w/w) was used as the calibration standard. Protein and nitrogen calculations Total protein content was calculated from total nitrogen as follows: Total protein=total nitrogen×6.25. True protein, including protein, free amino acids, and peptides, was calculated from total amino acids as follows: True protein=total amino acids×100/116. This calculation corrects the amino acid sum to a corresponding weight of polypeptide. Specifically, 100 g of protein (milk source) generates approximately 116 g of hydrolyzed amino acids due to water molecules added during protein hydrolysis (5, 27). Thus, true protein is a calculation of the amino acid sum, corrected for water added during hydrolysis to individual amino acids. The percentage of protein nitrogen was calculated as true protein divided by total protein. Non-protein nitrogen was calculated as the amount remaining after subtracting the percentage of protein nitrogen from 100. Statistical analysis Descriptive statistics were calculated for all variables, including individual amino acids, total amino acids, and total protein. An ANOVA model was used in the assessment of total protein and amino acid concentrations with the stage of lactation, country, mother's age and parity as covariates. Post-hoc pair-wise comparisons for total protein and total and individual amino acid concentrations in different countries were done by the Fisher's LSD test; as this was an exploratory analysis, no adjustments were made for multiple comparisons. Results Study population Between 20 and 28 human milk samples were analyzed from each of the nine countries included in the study population (Fig. 1). The number of samples varied by the stage of lactation: n=62 for lactation days 30–60; n=66 for lactation days 61–91; n=38 for lactation days 92–121; n=31 for lactation days 122–151; and n=23 for lactation days 152–188. The mean (standard deviation [SD]) age of the mothers who provided the samples was 30 (4.8) years; median (range) parity was 1 (1–4) (Table 2). Fig. 1 Distribution of human milk samples by country and stage of lactation. Table 2 Demographics of study population per country Australia Canada Chile China Japan Mexico Philippines UK USA All Infants' age (Means)  Days (range) 104.7 (68–169) 109.1 (60–163) 65.3 (30–183) 82.1 (33–149) 86.9 (30–188) 99.6 (34–160) 70.6 (30–183) 68.0 (42–114) 103.4 (35–188) 89.4 (30–188) Mother's age (Means)  Years (range) 30.2 (20–36) 32.8 (24–38) 25.7 (18–40) 27.5 (21–34) 31.1 (25–39) 31.1 (25–39) 27.3 (20–36) 32.4 (21–37) 30.6 (21–38) 30.0 (18–40) Parity  Median (range) 2 (1–3) 2 (1–4) 2 (1–4) 1 (1–2) 1 (1–3) 1 (1–3) 2 (1–4) 2 (1–4) 1 (1–3) 1 (1–4) Race 100% Caucasian 86% Caucasian 100% Caucasian 100% Asian 100% Asian 89% Caucasian 100% Asian 100% Caucasian 84% Caucasian – Protein and amino acid concentrations in the overall study population Table 3 summarizes the mean concentrations of amino acids, total protein, and true protein in human milk samples from the overall study population and by stage of lactation. The mean (SD) total protein concentration across all samples was 1,192 (200.9) mg/dL, and the true protein concentration across all samples was 908 (176.2) mg/dL. Overall, the mean concentration of true protein was 76% of the total protein concentration. Mean (SD) total amino acid concentration was 1,053 (204.4) mg/dL. As expected, the true protein concentration and total protein concentration were highly correlated (R2=0.7929) (Fig. 2). Results from multivariable analysis of variance (Table 4) demonstrated that the stage of lactation was significantly correlated with total protein concentration (P<0.0001) and total amino acid concentration (P<0.0001). Correlations with other variables (that is, the country, mother's age, and parity) were not statistically significant for either total protein or total amino acid concentration. Fig. 2 (a) Total protein and true protein over the course of lactation. (b) Correlation between total protein and true protein concentrations. Table 3 Mean concentrations of amino acids, total protein, and true protein in human milk samples from overall study population (n=220) and in samples by stage of lactation All samples Stage of lactation 30–188 days (n=220) 30–60 days (n=62) 61–91 days (n=66) 92–121 days (n=38) 122–151 days (n=31) 152–188 days (n=23) Amino acid: essential, mg/dL  CYS 23 (6.4) 27 (7.6) 22 (6.3) 21 (3.2) 19 (3.1) 21 (4.8)  HIS 26 (5.7) 30 (6.7) 25 (5.4) 25 (3.7) 24 (2.4) 23 (4.4)  ILE 59 (10.4) 67 (12.0)* 58 (9.3) 57 (7.0) 54 (5.1) 53 (8.4)  LEU 107 (20.6) 121 (24.5)* 104 (19.6) 104 (12.5) 99 (9.9) 97 (16.1)  LYS 72 (13.7) 81 (15.2)* 70 (13.2) 69 (9.4) 66 (7.0) 65 (11.0)  MET 17 (3.8) 19 (4.0)* 16 (3.6) 16 (3.3) 15 (2.1) 15 (2.9)  PHE 41 (10.4) 48 (12.7) 40 (10.5) 39 (5.1) 37 (4.1) 37 (7.6)  THR 49 (11.4) 56 (14.1)* 48 (11.5) 46 (5.4) 44 (4.4) 44 (8.9)  TRP 21 (4.8) 24 (5.7)* 21 (4.6) 19 (2.5) 18 (2.4) 18 (3.3)  TYR 48 (11.1) 54 (13.7) 47 (10.7) 45 (5.6) 43 (4.7) 42 (8.7)  VAL 60 (12.2) 69 (14.6)* 59 (11.6) 58 (7.0) 55 (5.8) 54 (9.5) Amino acid: non-essential, mg/dL  ALA 40 (9.4) 46 (11.4) 39 (9.4) 38 (4.9) 37 (4.3) 37 (7.3)  ARG 42 (13.2) 49 (16.4) 41 (14.5) 40 (5.3) 38 (4.9) 39 (10.2)  ASP 90 (18.9) 103 (22.5)* 88 (17.9) 85 (11.6) 83 (9.2) 81 (14.7)  GLU 187 (26.1) 201 (28.9) 184 (25.5) 186 (19.9) 179 (15.8) 173 (25.4)  GLY 25 (8.2) 29 (9.6) 24 (9.5) 24 (3.7) 23 (2.9) 22 (5.9)  PRO 95 (16.6) 107 (19.0)* 92 (15.3) 93 (10.7) 88 (8.8) 87 (13.3)  SER 50 (12.8) 57 (15.9) 48 (13.0) 47 (6.6) 45 (5.1) 44 (9.4)  Total amino acids, mg/dL 1,053 (204.4) 1,188 (241.6)* 1,026 (197.8) 1,013 (117.8) 968 (92.0) 953 (163.0)  Total protein, mg/dL 1,192 (200.9) 1,337 (211.4)* 1,166 (193.9) 1,132 (120.9) 1,103 (142.8) 1,097 (162.1)  True protein, mg/dL 908 (176.2) 1,024 (208.3)* 884 (170.5) 873 (101.6) 834 (79.3) 821 (140.5)  NPN, % 24 (6.6) 24 (0.8) 24 (0.8) 23 (0.9) 24 (1.2) 25 (1.6)  PN, % 76 (6.6) 76 (0.8) 76 (0.8) 77 (0.9) 76 (1.2) 75 (1.6) NPN, non-protein nitrogen; PN, protein nitrogen; SD, standard deviation. Data are mean (SD). Total protein is calculated using total nitrogen×a protein factor of 6.25. True protein is calculated from the total amino acid concentration. * P<0.0001 30–60 days versus following 4 months. Table 4 Analysis of variance results for total protein concentration and total amino acid concentration P Variation (% contribution) Total protein concentration  Stage of lactation <0.0001 22.9  Country 0.1498 5.2  Mother's age 0.3168 0.4  Parity 0.9348 0.2 Total amino acid concentration  Stage of lactation <0.0001 16.9  Country 0.0517 5.9  Mother's age 0.5726 0.1  Parity 0.9194 0.2 Variability in the individual amino acid concentrations, assessed by the coefficient of variation (CV), ranged from 14 to 32% for absolute concentrations (mg/dL) of amino acids in the 220 samples; the mean CV was 23% (Table 5). When normalized according to the percentage of total amino acids, the CVs were much lower, ranging from 3 to 13%, with a mean of 7%. Table 5 Variation in the content of individual amino acids in 220 human milk samples from nine countries Amino acid Mean, mg/dL Coefficient of variation (%) Mean (% of amino acid) Coefficient of variation (%) Essential  CYS 23 28 2.1 13  HIS 26 22 2.5 5  ILE 59 18 5.7 5  LEU 107 19 10.2 3  LYS 72 19 6.8 4  MET 17 23 1.6 11  PHE 41 25 3.9 5  THR 49 23 4.6 4  TRP 21 23 2 11  TYR 48 23 4.5 5  VAL 60 20 5.7 3 Non-essential  ALA 40 23 3.8 6  ARG 42 31 4 12  ASP 90 21 8.5 4  GLU 187 14 17.9 7  GLY 25 32 2.4 11  PRO 95 17 9.1 6  SER 50 26 4.7 6 Total 1,053 19 100 0 Mean 23 7 Protein and amino acid concentrations by stage of lactation The mean concentrations of total protein, true protein, total amino acids, and most individual amino acids in human milk declined steadily from 30 to 188 days of lactation (Table 3). The total protein concentration and total amino acid concentration were both significantly higher (P<0.0001) in the second month of lactation (days 30–60) compared with the following 4 months. In addition, the total amino acid concentration was significantly higher (P=0.029) in the third month of lactation (days 61–91) as compared with the sixth month (days 152–188). The decline in total protein that occurred from the second month of lactation through the sixth month reflected nearly equal declines in the various components of total protein (Fig. 3a). As such, despite steady declines, the proportion of essential amino acids, non-essential amino acids, and non-protein nitrogen components remained relatively unchanged as the duration of lactation increased (Fig. 3b). Similarly, with the exception of cysteine and glutamic acid, the relative contributions of each individual amino acid to total amino acids remained consistent between lactation months 2 and 6 (Table 6). Fig. 3 (a) Decline in total protein (or nitrogen-containing components) and each amino acid group and non-protein nitrogen components by stage of lactation. (b) Relative contributions of individual amino acid groups and non-protein-containing components to total protein content by stage of lactation. BCAA, branched-chain amino acids; Other Ess AA, other essential amino acids; Non Ess AA, non-essential amino acids; NPN Cmpnt, non-protein nitrogen components. Table 6 Amino acid profile by stage of lactation Amino acids as proportion of total amino acids, mean % (SD) Amino acid 30–60 days (n=67) 61–91 days (n=74) 92–121 days (n=40) 122–151 days (n=34) 152–188 days (n=25) P* Essential  CYS 2.2 (0.30) 2.1 (0.28) 2.1 (0.22) 2.0 (0.25) 2.1 (0.22) <0.01  HIS 2.5 (0.17) 2.5 (0.13) 2.5 (0.12) 2.5 (0.08) 2.4 (0.10) NS  ILE 5.6 (0.34) 5.7 (0.30) 5.7 (0.17) 5.6 (0.22) 5.6 (0.22) NS  LEU 10.2 (0.35) 10.2 (0.32) 10.2 (0.28) 10.2 (0.33) 10.2 (0.23) NS  LYS 6.8 (0.30) 6.8 (0.31) 6.8 (0.31) 6.7 (0.28) 6.8 (0.28) NS  MET 1.7 (0.16) 1.6 (0.18) 1.5 (0.24) 1.6 (0.11) 1.6 (0.15) NS  PHE 4.0 (0.22) 3.9 (0.21) 3.9 (0.16) 3.9 (0.12) 3.9 (0.16) NS  THR 4.7 (0.23) 4.7 (0.21) 4.6 (0.12) 4.6 (0.12) 4.6 (0.20) NS  TRP 2.0 (0.24) 2.0 (0.22) 1.9 (0.17) 1.9 (0.15) 1.9 (0.18) NS  TYR 4.6 (0.25) 4.6 (0.22) 4.5 (0.17) 4.5 (0.17) 4.4 (0.25) NS  VAL 5.8 (0.17) 5.7 (0.16) 5.7 (0.14) 5.7 (0.19) 5.7 (0.14) NS Non-essential  ALA 3.9 (0.26) 3.8 (0.25) 3.8 (0.13) 3.8 (0.15) 3.8 (0.17) NS  ARG 4.1 (0.51) 4.0 (0.52) 3.9 (0.28) 4.0 (0.31) 4.1 (0.48) NS  ASP 8.6 (0.45) 8.5 (0.38) 8.4 (0.31) 8.5 (0.32) 8.5 (0.32) NS  GLU 17.0 (1.22) 17.9 (1.17) 18.3 (0.92) 18.4 (0.92) 18.1 (1.08) <0.01  GLY 2.5 (0.29) 2.4 (0.32) 2.4 (0.20) 2.4 (0.14) 2.3 (0.25) NS  PRO 9.1 (0.65) 9.1 (0.66) 9.2 (0.40) 9.1 (0.46) 9.2 (0.46) NS  SER 4.8 (0.40) 4.7 (0.29) 4.7 (0.23) 4.7 (0.16) 4.6 (0.25) NS NS, not statistically significant; SD, standard deviation. * P<0.01, 30–60 days versus following 4 months. Protein and amino acid concentrations by country Mean concentrations of amino acids, total protein, and true protein in human milk samples by country were summarized (Table 7). The mean (SD) total protein concentration in human milk per individual countries ranged from 1,133 (125.5) to 1,366 (341.4) mg/dL. Protein and amino acid concentrations were similar across countries, and the overall effect of the country on the levels of total protein and total amino acids was not statistically significant, with the exception of Chile. Statistical analyses were performed after adjusting for the mother's age and stage of lactation, as the numbers of samples were unevenly distributed across lactation stages. Post-hoc comparisons between Chile and all the other countries were performed to evaluate the significance of the higher protein and amino acid levels in Chile. Even with adjustment, the results of these analyses showed that total protein, total amino acids, and most individual amino acid concentrations were significantly higher in the human milk samples from Chile as compared with the mean concentrations in samples from Australia, China, the Philippines, the United Kingdom, and the United States (P<0.05 for each comparison) (Fig. 4). Fig. 4 Adjusted total amino acid concentration by country. Data were adjusted for the stage of lactation as the number of samples was unevenly distributed across lactation stages due to limited sample availability.The mean total amino acid concentration in human milk samples from Chile was significantly higher than the mean concentration in samples from Australia, China, the Philippines, the United Kingdom, and the United States (P<0.02). Table 7 Mean concentrations of amino acids, total protein, and true protein in human milk samples by country Australia (n=24) Canada (n=28) Chile (n=20) China (n=26) Japan (n=28) Mexico (n=28) Philippines (n=21) UK (n=20) USA (n=25) All (n=220) Amino acids: essential, mg/dL  CYS 21 (4.3) 21 (3.2) 30 (11.6) 21 (5.0) 21 (4.0) 23 (8.8) 23 (4.7) 23 (3.2) 22 (4.7) 23 (6.4)  HIS 24 (3.4) 25 (2.5) 31 (10.0) 26 (4.4) 27 (4.6) 27 (7.0) 27 (4.5) 25 (4.1) 24 (6.0) 26 (5.7)  ILE 56 (7.5) 57 (5.3) 66 (16.6) 59 (10.0) 61 (10.8) 59 (10.2) 60 (9.0) 59 (9.0) 59 (12.2) 59 (10.4)  LEU 100 (13.9) 104 (10.0) 121 (37.8) 107 (17.3) 110 (19.2) 109 (23.5) 108 (15.0) 104 (16.3) 107 (21.2) 107 (20.6)  LYS 67 (10.8) 69 (7.4) 82 (22.9) 69 (10.9) 74 (11.8) 72 (15.8) 71 (9.9) 72 (11.4) 72 (15.3) 72 (13.7)  MET 16 (2.4) 16 (2.1) 19 (6.6) 16 (4.0) 17 (3.6) 17 (4.6) 17 (3.6) 16 (3.0) 17 (3.3) 17 (3.8)  PHE 38 (6.3) 40 (4.5) 51 (20.8) 40 (7.1) 42 (7.8) 42 (14.5) 41 (5.9) 40 (6.3) 40 (8.3) 41 (10.4)  THR 45 (6.6) 47 (4.6) 60 (23.0) 47 (7.7) 49 (7.7) 50 (15.7) 48 (7.0) 48 (7.2) 47 (9.5) 49 (11.4)  TRP 19 (3.2) 20 (2.2) 26 (8.8) 19 (3.4) 20 (3.3) 21 (6.5) 20 (3.0) 23 (3.3) 21 (3.4) 21 (4.8)  TYR 44 (6.5) 46 (4.8) 57 (22.1) 47 (8.2) 48 (9.0) 48 (14.1) 47 (7.1) 46 (7.7) 47 (9.7) 48 (11.1)  VAL 55 (7.7) 58 (5.8) 71 (23.5) 60 (9.1) 62 (10.0) 61 (15.0) 61 (8.7) 59 (8.6) 59 (11.0) 60 (12.2) Amino acids: non-essential, mg/dL  ALA 37 (6.2) 39 (4.6) 50 (17.9) 39 (6.3) 40 (6.5) 42 (12.8) 39 (5.4) 40 (6.2) 40 (7.4) 40 (9.4)  ARG 39 (7.8) 40 (5.4) 56 (26.3) 41 (8.1) 42 (7.5) 44 (20.2) 40 (5.2) 39 (7.6) 41 (10) 42 (13.2)  ASP 83 (13.9) 87 (9.7) 107 (33.9) 87 (14.2) 92 (14.4) 92 (23.4) 86 (13.6) 89 (13.8) 90 (18.8) 90 (18.9)  GLU 178 (23.7) 186 (15.1) 198 (40.5) 183 (23.3) 192 (25.1) 191 (25.1) 180 (22.6) 186 (22.3) 191 (31.6) 187 (26.1)  GLY 23 (4.4) 24 (3.3) 33 (15.9) 24 (4.0) 25 (4.8) 27 (13.7) 25 (4.1) 25 (4.2) 23 (5.2) 25 (8.2)  PRO 88 (11.2) 92 (9.3) 104 (26.8) 96 (16.2) 98 (16.9) 95 (15.5) 100 (15.2) 92 (15.0) 95 (18.1) 95 (16.6)  SER 46 (7.4) 47 (5.2) 62 (25.4) 48 (7.8) 51 (9.4) 51 (17.6) 50 (9.2) 48 (7.7) 47 (10.8) 50 (12.8) Total, mg/dL 978 (139.0) 1,016 (94.3) 1,224 (382.5) 1,029 (159.1) 1,070 (165.9) 1,071 (252.1) 1,045 (141.9) 1,034 (148.8) 1,043 (200.3) 1,053 (204.4) Total protein, mg/dL 1,145 (164.0) 1,133 (125.5) 1,366 (341.4) 1,167 (149.7) 1,205 (160.1) 1,179 (269.2) 1,191 (143.7) 1,215 (139.6) 1,173 (180.9) 1,192 (200.9) True protein, mg/dL 843 (119.8) 876 (81.3) 1,055 (329.7) 887 (137.2) 922 (143.1) 923 (217.3) 901 (122.4) 892 (128.2) 899 (172.7) 908 (176.2) NPN,% 26 (7.3) 22 (5.8) 24 (6.8) 24 (5.9) 23 (6.2) 22 (5.0) 24 (5.7) 27 (6.9) 23 (8.8) 24 (6.6) PN,% 74 (7.3) 78 (5.8) 76 (6.8) 76 (5.9) 77 (6.2) 78 (5.0) 76 (5.7) 73 (6.9) 77 (8.8) 76 (6.6) NPN, non-protein nitrogen; PN, protein nitrogen; SD, standard deviation. Data reflect mean (SD). Total protein is calculated using total nitrogen×a protein factor of 6.25. True protein is calculated from the total amino acid concentration. Discussion Human milk has a substantially lower total protein concentration than the milks of other species. However, human milk provides a richer source of essential amino acids, which allows infants to meet their protein requirements in a lower concentration (28). The unique quantity and quality of proteins in human milk are important because of the elevated requirements for essential amino acids and the conditionally essential nature of certain other amino acids during infancy (29). Human milk changes in both protein content and whey-to-casein composition over the course of lactation. During the first 30 days of lactation, the decline in protein content and the compositional shift in whey-to-casein ratio is clearly apparent. Early milk has a whey-to-casein ratio of approximately 90:10 in early milk, which evolves to approximately 50:50 in late lactation (30, 31); however, beyond the first month, the rate of change in protein content and composition becomes less obvious. It has been estimated that, during infancy, when protein accretion is at its highest, essential amino acids make up one-fifth of protein requirements. By comparison, later in childhood, essential amino acids comprise one-fifth of protein requirements and reflect only one-tenth of the protein requirements in adults (32). Thus, the amino acid composition of human milk has relevance for understanding the nutritional needs of infants. A further measure of nutritional value is true protein, which represents only the polypeptide portion of total protein. (The ‘Methods’ section includes the calculation of true protein). Protein and amino acid analyses in this study included samples from the second to the sixth month of lactation because of the known changes in human milk composition during the first month postpartum. The data from this study show a relatively large variation of amino acids and protein concentration among mothers’ milk samples. However, the variations in amino acid content, protein content, protein nitrogen, and non-protein nitrogen composition observed in this study are consistent with other human milk studies (17, 19). When the absolute concentrations of the individual amino acids were normalized to percentage of total amino acids, the variation in the data decreased considerably, indicating high consistency in the amino acid profile of human milk with very little impact from mother's race/ethnicity or age. Larger CVs in the amino acid profile (the percentage of amino acids) for cysteine and tryptophan may be explained by the use of separate procedures for the analysis of these two amino acids. The higher CVs in both the amino acid amount and the profile for methionine, arginine, and glycine may be due to the susceptibility of methionine and glycine to oxidation under hydrolysis conditions; the oxidized product of methionine may have affected the integration and quantification of arginine in the analysis method. Over the years, the Institute of Medicine of the US National Academy of Sciences has organized scientific expert panels to evaluate the totality of scientific literature on individual nutrients and publish Dietary Recommended Intakes for macronutrients and micronutrients for all age groups. The recommendation for the protein intake for infants during the first 6 months of life is based on the average volume of human milk (0.78 L/day) consumed by infants during this age range at the average protein content of human milk (11.7 g/L), as determined from data from several studies conducted in the United States using various methods of analysis (2). The mean of protein content of human milk from US-based studies is consistent with results described in this paper, which represent a more global analysis. Moreover, the amino acid profile reported in the current study is similar to the calculated mean based on references in the literature, which included amino acid composition of human milk (Fig. 5) (5–19). The mean total protein and amino acid concentrations reported here are also consistent with previously reported data (6, 12, 14, 15, 17, 18, 20). Fig. 5 A comparison of the amino acid profile from this nine-country study with mean values reported in the literature, as detailed in Table 1 (5–21). The human milk samples from Chile contained significantly higher amounts of each amino acid, total amino acids, and protein than most of the other eight countries. These findings are consistent with the results of an analysis of lactoferrin levels in human milk samples from the same study, which found that the mean lactoferrin concentration was significantly higher in samples from Chile compared with the samples from the other eight countries (33), as were levels of zinc (34). Additionally, and also from this same study, the alpha-lactalbumin (a protein fraction) concentration throughout lactation differed between Chile and the other countries (35). The total amino acid concentration and total protein concentration were, however, highly correlated in the samples from Chile just as in the other countries. Thus, we cannot speculate why total protein levels would differ in milks from mothers in Chile versus other countries. Assessment of maternal diet The maternal dietary intake of many important micronutrients has been shown to influence the concentration of those nutrients in human milk including vitamins A and E (36), fatty acids such as DHA (24), and carotenoids such as lutein and beta-carotene (23); however, the protein content of human milk has been shown to be relatively unaffected by maternal diet (36, 37). This lack of influence was one of the reasons we were interested in evaluating protein content across countries and explains why we did not evaluate maternal dietary protein intake in this study. This study provides considerable evidence that the protein content and amino acid composition of human milk are relatively uniform across geographic regions when compared by stage of lactation. Our amino acid analysis revealed little change in amino acid profiles over time. Only cysteine and glutamic acid showed any significant variation with stage of lactation. The shift in these amino acids would be consistent with an increase in the proportion of casein as lactation continues because the concentration of cysteine is lower (and glutamic acid is higher) in the casein fraction of human milk, compared with the whey fraction. In summary, the results from this large-scale, multinational study of 220 human milk samples revealed a high level of uniformity in protein and amino acid composition across a wide range of geographic locations. With the exception of Chile, there were no significant differences between countries in amino acid, true protein, and total protein concentrations. In all nine countries included in the study, protein and amino acid concentrations declined steadily from 30 to 188 days postpartum. Moreover, the proportion of true protein and the amino acid profiles of human milk were generally consistent across lactation stages and countries. Several features of this study, including the size and diversity of the study population, and our use of highly standardized procedures for collection, storage, and analysis of human milk samples strengthen the validity of our findings and enhance their applicability. Acknowledgements The authors thank Mr. John Weaber and Mr. Charles Kuhlman for their leadership and support, and Mr. Timothy Holly and Ms. Judith Nazzario for conducting laboratory analyses, all of whom previously worked for Wyeth Nutrition. Editorial support was provided by MaiLee Wong, PhD of Caudex Medical, funded by Wyeth Nutrition. Conflicts of interest and funding PF and KP are employees of Wyeth Nutrition. MG, THZ, and AB were employees of Wyeth Nutrition at the time of study. This study was funded by Wyeth Nutrition, a Nestle business. ==== Refs References 1 World Health Organization Global strategy for infant and young children feeding 2003 Geneva, Switzerland World Health Organization 2 Food and Nutrition Board, Institute of Medicine, National Academy of Sciences Dietary reference intakes for energy, carbohydrate, fiber, fat, fatty acids, cholesterol, protein, and amino acids (macronutrients) 2005 Washington, DC National Academy Press 620 3 3 Cheung MW Pratt EL Fowler DI Total amino acid composition in mature human milk analysis by the ion exchange resin column chromatographic technic Pediatrics 1953 12 353 7 13099905 4 Spackman DH Stein WH Moore S Automatic recording apparatus for use in the chromatography of amino acids Anal Chem 1958 30 1190 206 5 Lonnerdal B Forsum E Hambraeus L The protein content of human milk Nutr Rep Int 1976 13 125 34 6 Raiha N Minoli I Moro G Bremer HJ Milk protein intake in the term infant. II. Effects on plasma amino acid concentrations Acta Pediatr Scand 1986 75 887 92 7 Picone TA Benson JD Moro G Minoli I Fulconis F Rassin DK Growth, serum biochemistries, and amino acids of term infants fed formulas with amino acid and protein concentrations similar to human milk J Pediatr Gastroenterol Nutr 1989 9 351 60 2614623 8 Atkinson SA Anderson GH Bryan MH Human milk: comparison of the nitrogen composition from mothers of premature and full-term infants Am J Clin Nutr 1980 33 811 15 7361699 9 Renner E Milk and dairy products in human nutrition 1983 Regensburg, Germany Friedrich Pustet 10 United States Department of Agriculture, Agricultural Research Service, Nutrient Data Laboratory USDA National Nutrient Database for Standard Reference, Release 28 Available from: http://www.ars.usda.gov/Services/docs.htm?docid=8964 [cited 3 December 2015] 11 Svanberg U Gebre-Medhin M Ljungqvist B Olsson M Breast milk composition in Ethiopian and Swedish mothers. III. Amino acids and other nitrogenous substances Am J Clin Nutr 1977 30 499 507 557894 12 Jarvenpaa AL Rassin DK Raiha NCR Gaull GE Milk proteins quantity and quality in the term infant. II. Effects on acidic and neutral amino acids Pediatrics 1982 70 221 30 7099788 13 Harzer G Bindels JG M Xanthou Main compositional criteria of human milk and their implications on nutrition in early infancy New aspects of nutrition in pregnancy, infancy and prematurity 1987 Amsterdam Elsevier Science 83 94 14 Donovan SM Lonnerdal B Non-protein nitrogen and true protein in infant formulas Acta Pediatr Scand 1989 78 497 504 15 Hanning RM Paes B Atkinson SA Protein metabolism and growth of term infants in response to a reduced-protein, 40:60 whey: casein formula with added tryptophan Am J Clin Nutr 1992 56 1004 11 1442650 16 Davis TA Nguyen HV Garcia-Bravo R Fiorotto ML Jackson EM Lewis DS Amino acid composition of human milk is not unique J Nutr 1994 124 1126 32 8027865 17 Motil KJ Thotathuchery M Bahar A Montandon CM Marginal dietary protein restriction reduced nonprotein nitrogen, but not protein nitrogen, components of human milk J Am Coll Nutr 1995 14 184 91 7790694 18 Darragh AJ The amino acid composition of human milk corrected for amino acid digestibility Br J Nutr 1998 80 25 34 9797640 19 Zhao X Xu Z Wang Y Sun Y [Studies of the relation between the nutritional status of lactating mothers and milk composition as well as the milk intake and growth of their infants in Beijing. Pt. 4. The protein and amino acid content of breast milk] Ying Yang Xue Bao 1989 11 227 32 [Article in Chinese] 12283956 20 Wu ZC Chijang CC Lau BH Hwang B Sugawara M Idota T Crude protein content and amino acid composition in Taiwanese human milk J Nutr Sci Vitaminol 2000 5 246 51 11234918 21 Lonnerdal B Forsum E Hambraeus L A longitudinal study of protein, nitrogen and lactose contents of human milk from Swedish well-nourished mothers Am J Clin Nutr 1976 29 1127 33 973601 22 Prentice A RG Jensen Regional variations in the composition of human milk Handbook of milk composition 1995 San Diego, CA Academic Press 115 221 23 Canfield LM Clandinin MT Davies DP Fernandez MC Jackson J Hawkes J Multinational study of major breast milk carotenoids of healthy mothers Eur J Nutr 2003 42 133 41 12811470 24 Yuhas B Pramuk K Lien EL Human milk fatty acid composition from nine countries varies most in DHA Lipids 2006 41 851 8 17152922 25 Waters Corporation AccQ-fluor reagent kit care and use manual 2008 Milford, MA Waters Corporation 26 Liu HJ Chang BY Yan HW Yu FH Liu XX Determination of amino acids in food and feed by derivatization with 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate and reversed-phase liquid chromatographic separation J AOAC Int 1995 78 736 44 27 Soupart P Moore S Bigwood EJ Amino acid composition of human milk J Biol Chem 1954 206 699 704 13143030 28 Heine WE Klein PD Reeds PJ The importance of alpha-lactalbumin in infant nutrition J Nutr 1991 121 277 83 2002399 29 Heird WC BA Bowman RM Russell Infant nutrition Present knowledge in nutrition 2006 2 9th ed Washington, DC International Life Sciences Institute 556 30 Kunz C Lonnerdal B Re-evaluation of the whey protein/casein ratio of human milk Acta Paediatr 1992 81 107 12 1515752 31 Lonnerdal B Nutritional and physiologic significance of human milk proteins Am J Clin Nutr 2003 77 Suppl 1537S 43S 12812151 32 Young VR Adult amino acid requirements: the case for major revision in current recommendations J Nutr 1994 124 8 Suppl 1517S 23S 8064412 33 Lien E Jackson J Kuhlman C Pramuk K Lönnerdal B Janszen D Variations in concentrations of lactoferrin in human milk: a nine country survey Adv Exp Med Biol 2004 554 423 6 15384615 34 Radzanowski GM Jackson JK Pramuk K Kaup SM Comparison of trace elements and macronutrients in breast milk of women from eight different geographical locations 1999 Presented at Trace Elements in Man and Animals, TEMA-10, Evian May 2–7 France 35 Jackson JG Janszen DB Lonnerdal B Lien EL Pramuk KP Kuhlman CF A multinational study of alpha-lactalbumin concentrations in human milk J Nutr Biochem 2004 15 517 21 15350982 36 Lonnerdal B Effect of maternal dietary intake on human milk composition J Nutr 1986 116 499 513 3514820 37 Nommsen LA Lovelady CA Heinig MJ Lonnerdal B Dewey KG Determinants of energy, protein, lipid, and lactose concentrations in human milk during the first 12 months of lactation: the DARLING Study Am J Clin Nutr 1991 53 457 65 1989413

PMC005xxxxxx/PMC5002400.txt

==== Front Front PsycholFront PsycholFront. Psychol.Frontiers in Psychology1664-1078Frontiers Media S.A. 10.3389/fpsyg.2016.01272PsychologyReviewWhat Is the Sense of Agency and Why Does it Matter? Moore James W. *Department of Psychology, Goldsmiths, University of LondonLondon, UKEdited by: Morten Overgaard, Aarhus University, Denmark Reviewed by: Felipe De Brigard, Duke University, USA; Roy Salomon, École Polytechnique Fédérale de Lausanne, Switzerland *Correspondence: James W. Moore, j.moore@gold.ac.ukThis article was submitted to Consciousness Research, a section of the journal Frontiers in Psychology 29 8 2016 2016 7 127211 5 2016 10 8 2016 Copyright © 2016 Moore.2016MooreThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.Sense of agency refers to the feeling of control over actions and their consequences. In this article I summarize what we currently know about sense of agency; looking at how it is measured and what theories there are to explain it. I then explore some of the potential applications of this research, something that the sense of agency research field has been slow to identify and implement. This is a pressing concern given the increasing importance of ‘research impact.’ consciousnessfree willresponsibilityhuman-computer-interactionlegalagingschizophreniaOCD ==== Body Introduction This article aims to serve two purposes. First, I wish to provide a general overview of research on sense of agency. This is by no means exhaustive, and is instead intended to give the reader a broad introduction to the topic. Second I wish to explore some areas in which this research may have some kind of impact. Impact is becoming an increasingly important issue for research scientists. According to the UK’s 2014 Research Excellence Framework, impact can be defined as research having an “an effect on, change or benefit to the economy, society, culture, public policy or services, health, the environment or quality of life, beyond academia.” Although many scientists are dubious of this increased focus on impact, it is almost certainly here to stay, at least for the foreseeable future. We all, therefore, would be well-served by being mindful of it in our own research. Some research topics lend themselves to impact more than others. Sense of agency falls more into the latter group; those working on this topic may struggle to articulate the relevance and potential impact of what they do. This article is, in part, an attempt to address this issue. What Is the Sense of Agency? Background When we make voluntary actions we tend not to feel as though they simply happen to us, instead we feel as though we are in charge. The sense of agency refers to this feeling of being in the driving seat when it comes to our actions. Synofzik et al. (2008) draw an important distinction between the Feeling of agency (FOA) and the Judgment of agency (JOA). FOA is a lower level non-conceptual feeling of being an agent; it is the background buzz of control we feel for our voluntary actions when not explicitly thinking about them. JOA, on the other hand is a higher-level conceptual judgment of agency, and arises in situations where we make explicit attributions of agency to the self or other. The FOA is linked to low-level sensorimotor processes, whilst the JOA to higher-level cognitive processes such as background beliefs and contextual knowledge relating to the action. These two levels of agency processing, although related, can be dissociated from one another. For example, an unexpected action outcome, which would signal non-agency at the FOA level, can nevertheless be attributed to the self at the JOA level if beliefs and contextual factors imply self-causation (for example, when an unexpected action outcome happens when acting alone in a room). This distinction between FOA and JOA also highlights an important distinction between agency and causality. Whilst the low-level FOA automatically registers agency or non-agency by tracking sensorimotor contingencies, the higher-level JOA deploys more general-purpose causal attribution processes that are not specifically linked to the sensorimotor system (see Gallagher, 2000, for a detailed discussion of this). As with other aspects of conscious experience, the sense of agency is not an infallible reproduction of objective reality. As a consequence of this our experiences of agency can go awry. This is quite common in gambling, where players often feel an exaggerated sense of agency. An example of just such an illusion of control was noted by Henslin in the 1960s (Henslin, 1967). Henslin was a sociologist and spent a number of weeks observing cab drivers in St. Louis in the USA. A popular past-time among the cabbies was craps, a dice rolling gambling game. Henslin joined in with these games and made an intriguing observation. When these cab drivers came to roll the dice they altered their behavior depending on the number they needed, throwing harder for higher numbers and more gently for lower numbers. What is striking about this kind of behavior is that the outcome of dice rolling is objectively uncontrollable. Nevertheless these cabbies clearly felt otherwise. This is an example of where the sense of agency can be quite divorced from objective reality. You might think that you are immune to such cognitive foibles, but you would almost certainly be mistaken. I would bet that most of us have fallen foul, at some point, of so-called ‘placebo buttons.’ These are buttons that we encounter every day that we think do things, but actually do nothing (McRaney, 2013). Buttons at pedestrian crossings are a common example of placebo-buttons. Most of these buttons are ineffective and instead the changing of the traffic lights are linked to timers. This was shown by a recent survey of pedestrian crossings in New York (McRaney, 2013). Intriguingly, most of us fail to notice the causal inefficacy of our button presses. Other examples of placebo buttons include ‘close door’ buttons in lifts and even thermostats in offices (many of which, apparently, do not work). There are two reasons for flagging up the occasional lapses in our sense of agency. The first is to show that the accuracy of this experience is not a given. Instead, the brain appears to actively construct the sense of agency, and because of this, our experiences of agency can be quite divorced from the facts of agency. The second reason is that these lapses reveal something quite remarkable about our sense of agency: its impressive flexibility. Beyond the examples I have given here, we see over and over again that people come to experience control over outcomes in many weird and wonderful situations. The voodoo doll is another example; people taking part in this practice genuinely believe that sticking a pin in an effigy of someone causes actual physical harm in that person. At first blush this inference seems irrational. However, examples like the voodoo doll actually hint at the adaptability and flexibility of the agency processing system. It is worth reminding ourselves that causal mechanisms are quite opaque in a lot of modern technology (consider the simple act of tapping on a keyboard and seeing a letter appear on the screen in front of you – there are a lot of steps in this causal chain that are hidden from you). Despite this causal opacity, we feel in control of these interactions. So the flexibility that might make us vulnerable to agency errors in things like placebo buttons and voodoo dolls, can also allow our experience of agency to extend into new domains and track the rapidly changing agentic structure of our environment. Rather than our agency processing system breaking down with the development of tools, which have changed and extended our agentic capabilities, it has been flexible and adaptable, allowing us to accommodate these changes. Measures The number of scientific investigations of sense of agency has increased considerably over the past 20 years or so. This increase is despite the fact that experiments on sense of agency face certain methodological problems. A major one is that the sense of agency is phenomenologically thin (Haggard, 2005). That is, when we make actions we are typically only minimally aware of our agentic experiences. This is quite unlike conscious experience in other modalities, especially vision, where our experiences are typically phenomenologically strong and stable. What this means is that sense of agency can be difficult to measure. As a result of this, experimenters have had to be quite inventive in order to develop paradigms that capture this rather elusive experience. You can generally group these paradigms into implicit or explicit measures. Implicit measures assess a correlate of voluntary action and infer something about the agentic experience on the basis of this. In these paradigms no one is ever asked, directly, about their agentic experience. Probably the most widely used implicit measure of sense of agency is intentional binding (for a review see Moore and Obhi, 2012). This was developed by Haggard et al. (2002) and is based on time perception. Haggard et al. (2002) found that when we make a voluntary action, the perceived times of the action and its effect are shifted toward each other. This change in time perception is taken to be an implicit marker of sense of agency. Other implicit measures of sense of agency include sensory attenuation paradigms. It has been shown that the perceived intensity of the sensory consequences of voluntary action is lower than for passive movements (Blakemore et al., 1998, 1999). This can explain why we are unable to tickle ourselves (Blakemore et al., 1998). In these sensory attenuation paradigms, researchers use changes in perceived intensity of sensory feedback to infer something about the participant’s sense of agency. Explicit measures, on the other hand, directly ask the participant to report something about their agentic experience. These measures are more intuitive but they can be vulnerable to problems like demand effects. A number of these paradigms require participants to make action recognition judgments. Typically the participant makes an action, but does not directly see that action. Instead they are shown some kind of feedback on a screen. This feedback may depict the participant’s action or it might depict the action of someone or something else (perhaps an experimenter or a computer), and the participant is asked whose movement it is. Importantly, the experimenter ensures that there is some uncertainty over the agent of the action being displayed. An example of this kind of task was used by Farrer et al. (2008). They had participants perform regular finger tapping movements while wearing a glove. They could not directly see these movements, and instead they were shown video feedback of the movement on the screen. A delay was inserted between the movement and the feedback presented to the participant. The participants were not aware that the movement was always their own, and instead were led to believe that the movement was either their own or an experimenter performing the same movement, and that this could switch at any time. The participant simply had to indicate when they thought they were seeing their own movement and when they thought they were seeing the experimenter’s movement. Farrer et al. (2008) found that participants experienced a bi-stable impression of agency in this situation, with judgments of agency spontaneously flipping between self and experimenter. Other explicit measures also use visual feedback about movements, but will not create this kind of self/other confusion. Instead, the participant is required to make a judgment about the feedback itself. An example of this kind of action monitoring task can be seen in an experiment carried out by Synofzik et al. (2010). In this experiment participants made pointing movements under a screen, meaning that they could not directly see the movement. On the screen participants were shown a visual marker (white disk) that tracked the pointing movement. This marker was rotated by varying degrees relative to the actual movement. The participants had to indicate the direction in which the visual feedback was rotated relative to the actual movement. This gave the experimenters a measure of action awareness and, more specifically, sensitivity to distortions in action-relevant feedback. A final kind of explicit measure requires participants to report on their feeling of agency for certain action outcomes that their movements might have caused. A simple example of this would involve a key press that causes an outcome after a variable delay. Participants would then judge how much they felt their action caused the outcome. A common finding is that such causal judgments are stronger for shorter delays (e.g., Shanks et al., 1989; Chambon et al., 2015). Interestingly, this kind of explicit measure taps into a slightly different aspect of the agentic experience compared with the other two kinds of explicit measure described in this section. Action recognition/monitoring tasks focus more on the action element, whereas, causal judgment tasks focus on the outcome component. Although, both of these are central to the agentic experience, this difference is often overlooked and not very well-understood. Theories The two most influential theories of sense of agency have been the ‘Comparator Model’ developed by Frith et al. (2000) and Frith (2005), the ‘Theory of apparent mental causation’ developed by Wegner and Wheatley (1999) and Wegner (2002). The comparator model takes as its starting point the motor control system. We now know a great deal about the computational processes underpinning the control of voluntary movement (see Wolpert and Miall, 1996, for a review). According to the comparator model, some of these processes also inform the sense of agency. On this view, our actions start with intentions or goals, which enables a representation to be formed of the desired state of the motor system. Controllers within the motor control system then use this information about the desired states to generate a motor command. This motor command produces a movement, which changes the state of the motor system, and generates sensory feedback. On the basis of this information the new state of the system can be estimated. This estimate is compared with the desired state at a comparator. If there is a mismatch then an updated motor command is issued. This process can continue until the desired state is achieved (indicated by the absence of a mismatch at the comparator). The issue with a motor system operating only in this way is that it is slow to respond to error. Because of this, the organism is vulnerable. The solution, it would appear, is to have an additional predictive component within the motor system, and it is this that is thought to be particularly relevant to sense of agency. This predictive component uses a copy of the motor command that is issued (a so-called ‘efference copy’) to predict the future state of the system. This includes predictions about changes to the motor system as well as the sensory consequences resulting from those changes. On the basis of these predictions, a representation of the predicted state of the system can be formed, and this representation can be compared both with the desired state of the system and with the actual state of the system. The former comparison is important for motor control, as it allows the organism to rapidly adjust motor commands in advance of incorrect actions being performed. The latter comparison is thought to be important for sense of agency. According to the comparator model, the output of the comparison between predicted and actual states determines whether or not we feel a sense of agency. If there is a match, then we feel a sense of agency; if there is mismatch then we do not. A number of studies support this idea that sense of agency is closely tied to sensorimotor processes. For example, as predicted by the comparator model, it has been shown that the perceived intensity of self-produced tactile sensations is attenuated relative to externally produced ones (Blakemore et al., 1998, 1999). It has also been shown that sensorimotor prediction contributes to intentional binding (Moore and Haggard, 2008). Finally, mismatches between expected and actual sensory feedback influence action recognition judgments (e.g., Daprati et al., 1997; Franck et al., 2001). The ‘Theory of apparent mental causation’ approaches sense of agency from a quite different angle. Whereas the comparator model places a heavy emphasis on the contribution of the motor system to sense of agency, the theory of apparent mental causation explicitly downplays this contribution. Indeed, according to this theory, it is because we do not have conscious access to the motor control system that our sense of agency can, at times, be so misleading, as seen in phenomena like voodoo dolls and placebo buttons. According the theory of apparent mental causation when we make a voluntary action there is an unconscious causal pathway that is responsible for the action. This pathway corresponds to the workings of the motor control system. There is also an unconscious causal pathway that is responsible for the associated thoughts about actions (i.e., intentions). In addition to these unconscious causal pathways, there are certain events that we are conscious of, namely the intention to act and the act itself. According to Wegner it is the relationship between the thought and the action that determines the sense of agency (or in Wegner’s term, the ‘experience of conscious will’). If our intention to act happens before we act, is consistent with the action, and is the only plausible cause of the action, then we feel as though we have caused the action. A fundamentally important feature of Wegner’s theory is the additional claim that this feeling is illusory – the inference that our intentions have caused our actions is erroneous (the unconscious pathways are the real causes of our actions. A number of studies support Wegner’s theory of apparent mental causation. For example, it has been shown that priming thoughts about an upcoming action fosters an illusory sense of agency for that action (Wegner and Wheatley, 1999; Wegner et al., 2004). It has also been shown that manipulating high-level contextual information about an action (in the form of causal beliefs) alters sense of agency, as measured by intentional binding (Desantis et al., 2011). These two theories, the comparator model and the theory of apparent mental causation, offer competing accounts of sense of agency. They differ in terms of the sources of information thought to be most important for producing sense of agency. For the comparator model, sensorimotor processes are key. For the theory of apparent mental causation, the emphasis, at least in the experimental work carried out to test it, has been on information that is external to the motor system, such as environmental and social cues (for a more extensive discussion, see Wegner and Sparrow, 2004). The traditional assumption has been, therefore, that these two views are mutually exclusive. However, this assumption has challenged by a number of studies. For example, using the intentional binding measure Moore and Haggard (2008) showed that both internal sensorimotor prediction and external action outcomes contributed to the sense of agency. It was found that binding of the action to the tone outcome was present when the probability of that outcome was high, even when it did not occur. This suggests that if sensorimotor prediction is sufficiently strong binding will occur. On the other hand, it was found that when sensorimotor prediction was weak, binding would occur but only when the key press actually caused the tone outcome. This would suggest that the presence of an external tone outcome retrospectively triggered the binding effect. Findings such as these led us to develop an alternative ‘cue integration’ theory of sense of agency (Moore et al., 2009; Moore and Fletcher, 2012). This helped us move beyond the debate over whether sense of agency was based on sensorimotor information (comparator model) or information external to the motor system (theory of apparent mental causation). Instead, according to the cue integration theory both views have merit, and in fact the sense of agency is based on various different sources of information (or agency cues). We have also suggested that the relative influence of the different sources of information may be linked to their reliability, with the more reliable source of information dominating the agentic experience. We can see evidence of this in Moore and Haggard’s (2008) study, described above, where the influence of external action outcomes on intentional binding increased when the reliability of sensorimotor prediction decreased. We can also see evidence of this in patients with schizophrenia. Using an agency attribution paradigm, Synofzik et al. (2010) showed that agency judgments in people with schizophrenia relied more strongly on visual feedback about an action rather than on internal sensorimotor cues. This reliance on external visual feedback is consistent with the cue integration theory, as it has been shown that sensorimotor prediction is unreliable in people with schizophrenia (a similar finding was also obtained by Voss et al., 2010). Although, a more thorough examination of this theory is needed, it does promise to help us understand the processes underpinning sense of agency in health and disease. Why Does Sense of Agency Matter? The previous section provides an overview of sense of agency research and theory. However, from this overview it would not be entirely clear why any of this matters, particularly from an impact point of view. In the following section I want to address this. I will look at the possible impact of sense of agency research in the context of health and well-being, human-computer-interaction, and the broader issues of free will and responsibility. Health and Well-being Schizophrenia and Other Disorders Schizophrenia is the classic disorder of sense of agency and has been the subject of more agency research than any other disorder. The symptoms of schizophrenia are grouped into two categories: ‘positive symptoms’ and ‘negative symptoms.’ Negative symptoms are defined by the absence of a normal function (for example, ‘alogia’ or reduced speech). Positive symptoms, on the other hand, are defined by the abnormal presence of perceptions (hallucinations) or beliefs (delusions). Abnormal experiences of agency fall within the positive symptom category. Although these abnormal experiences can take many forms, the most common are passivity symptoms (or delusions of control). A patient with passivity symptoms will feel as though his or her actions are not under their control. You can see this in the following patient reported by Mellor (1970, p. 18): ‘It is my hand and arm which move, and my fingers pick up the pen, but I don’t control them. What they do is nothing to do with me.’ Research on patients with schizophrenia has confirmed that these individuals have agency processing problems. In one relatively early study by Daprati et al. (1997), healthy controls and patients with schizophrenia made simple hand movements. They did not directly see their own movements. Instead they saw visual feedback of the movement on screen via a video link. These movements were either (a) their own actual movements, (b) the same movements made by an experimenter in another room, or (c) the movement of that experimenter performing a different movement. The participants and the experimenter were wearing gloves to prevent any visual identity clues. After each trial the participant simply had to say whether the movement on the screen was their own movement or the experimenter’s. Compared with controls, patients – especially those experiencing passivity symptoms – made more errors in attributing the action to its correct source when the experimenter made the same movements as them. In this situation of agentic uncertainty, patients struggled to recognize their own movements. These action recognition problems have since been confirmed in a number of other studies. For example, Franck et al. (2001) tested patients and healthy controls on an action recognition task. In this experiment they made movements and again only saw video feedback of the movement. In one condition different levels of spatial distortion were introduced. In another condition different time delays were introduced. After each trial participants had to say whether the hand movements on the screen matched their own. Healthy participants tended to say no earlier in both conditions than patients who took much longer to detect these mismatches. Again this suggests abnormal action awareness in patients. Where agency research on patients has been particularly useful is in uncovering the information processing abnormalities underpinning these disordered experiences of agency. Patients with schizophrenia seem to have specific problems with sensorimotor prediction, which, as we saw in the previous section, is crucial for the sense of agency. One line of evidence comes from studies on sensory attenuation. The neural response to sensory feedback generated by a voluntary action is attenuated – the brain cares less about the things it can predict (Blakemore et al., 1999). This can explain our inability to tickle ourselves: self-tickling is less effective because we can predict the sensory consequences of our actions, resulting in the sensory percept being attenuated. Interestingly, patients with schizophrenia can tickle themselves (Blakemore et al., 2000). This finding strongly suggests that patients struggle predicting the sensory consequences of their actions. Another line of evidence comes from research on intentional binding. As we saw in the previous section, voluntary actions are associated with a compression of the perceived interval between the action and its effect: the perceived time of action is shifted toward the effect, and the perceived time of the effect is shift back toward the action. In a study on healthy adults, Moore and Haggard (2008) showed that the action component of the binding effect (the shift in perceived time of action toward the outcome) is partly linked to prediction. When the outcome was probable, but not always present, there was a shift in perceived time of action even on those trials where the outcome did not occur. This was not the case when the outcome was unpredictable. In follow-up work by Voss et al. (2010), it was found that this predictive effect was absent in patients with schizophrenia. These problems with prediction can help us explain the behavioral findings from individuals with schizophrenia that I described above. For example, Franck et al. (2001) found that patients struggled to detect temporal and spatial discrepancies between their movements and the feedback of those movements. If the patient struggles to predict where their hand should be during movement, they will struggle to detect spatial distortions. Furthermore, if the patient struggles to predict when their hand should move, they will struggle to detect temporal delays. More recently it has been suggested that problems with prediction represent a core deficit in the disorder (Fletcher and Frith, 2009). On this view predictive deficits can explain positive symptoms more generally, not just passivity symptoms. Clearly we need to find out more about the nature and origins of this predictive deficit in individuals with schizophrenia, but it at least offers us a starting point in the quest to understand and ultimately treat the disorder. Developing interventions to remedy these agency processing problems is one possible avenue for impact. Aberrant experiences of agency are not just confined to patients with schizophrenia. Indeed, aberrant experiences of agency can be seen in various disorders. Anosognosia for hemiplegia is one such disorder, and is attracting growing interest in the field. ‘Anosognosia’ comes from Greek words nosos (meaning “disease”), and gnosis (meaning “knowledge”), so patients with anosognosia are unaware of their disease or impairment. There are many kinds of anosognias, but the most relevant for us is anosognosia for hemiplegia. These are patients who are paralyzed, usually following stroke, but who are unaware of this impairment. The following description from Berti et al. (2007) is of a patient with anosognosia for hemiplegia: “CR presented severe and persistent anosognosia for her left hemiplegia... She never spontaneously reported her motor problems. When questioned about her left arm, she always claimed that it could move without any problem. When asked to actually perform movements, she attempted to perform the action, and after a few seconds she appeared to be satisfied with her performance” (p. 172). From an agency point of view this disorder is intriguing. It suggests that an individual can experience a sense of agency for movements that they cannot make, and for which there is compelling sensory evidence to confirm their paralysis. Research carried out by Fotopoulou et al. (2008) shows that patients do in fact discount sensory evidence in their agency assessments. When instructed to make a movement, they will claim to have moved despite contradictory visual feedback. What this implies is that the experience of agency in these individuals is strongly governed by pre-motor agency cues, such as intentions and sensorimotor predictions. As with the schizophrenia patients, we clearly we need to find out more about the exact nature of this deficit, but it again gives us a useful starting point for the development of therapeutic interventions. For example, it might be useful to try to find ways of increasing the weighting that anosognosia for hemiplegia patients give to sensory feedback, either through cognitive/behavioral interventions or through neural interventions (e.g., pharmacological). Beyond anosognosia for hemiplegia and schizophrenia there are a number of other disorders that are beginning to attract interest from agency researchers. In Obsessive Compulsive Disorder, for example, it has been shown that patients have deficits in sensorimotor prediction resulting in a reduction of sensory suppression (Gentsch et al., 2012). This finding echoes those from patients with schizophrenia described above. It has also been shown that individuals with high obsessive-compulsive tendencies tend to omit agency from spoken language, perhaps indicating a reduced sense of agency in these individuals (Oren et al., 2016). From this brief (and far from complete) survey of the clinical research on sense of agency, it should be apparent that aberrant experiences of agency are strikingly common in a range of different disorders. In the field of schizophrenia research, some have claimed that such disturbances in self-awareness are, in fact, a core feature of the disorder (e.g., Sass, 2014). It is a question for future research to find out whether we should also start thinking about certain other disorders in the same way. Whether or not this comes to pass, it is now incumbent on agency researchers to use these findings from patients with disorders of sense of agency to begin developing interventions aimed at remedying them. Healthy Aging With an aging population there has been a strong push to understand cognitive changes across the lifespan. This is with a view to mitigating some of the negative effects of these changes in older adults. Changes in sense of agency seem to be a feature of old age and therefore warrant further investigation. Understanding these changes, and developing interventions aimed at remedying them could serve to improve well-being in older adulthood. The majority of work on sense of agency in old age has focussed on examining the link between general changes in sense of agency (based on self-reports) and old age, and how these relate to various indices of health and well-being. Based on this work it is clear that old age is associated with a reduction in the sense of agency. For example, in a large-scale survey of Americans, Lachman and Firth (2004) found that 62% of older adults disagreed with the statement “What happens in my life is beyond my control” whereas almost 80% of young adults (25–39 years) disagreed with it. This reduction begins at around the age of 50 years and continues into older adulthood, with the most rapid decline occurring between 60 and 80 years (Mirowsky, 1995). Importantly, this reduction in sense of agency is associated with poor health and a reduction in quality of life (Langer and Rodin, 1976; Rodin and Langer, 1977), which itself highlights the pressing need for rigorous experimental research. A key factor in this reduced feeling of control is likely to be a reduction in the basic capacity for agency due physical impairment (Mirowsky, 1995). However, there might also be neurocognitive factors underpinning this reduction in the sense of agency. To-date, few studies have directly examined this from an experimental psychology or cognitive neuroscience perspective. One of the few that have is a study by Metcalfe et al. (2010). They found that the experience of control in older adults less sensitive to three external performance manipulations (such as the insertion of a temporal delay between their movement and a cursor moving on the screen) than a control group of younger adults. The pattern of results is intriguing, suggesting that older adults have a reduced sensitivity to external sensory cues to agency. Future research should explore this in more detail. By uncovering the agency processing abnormalities in older adults it will then be possible to start developing interventions aimed at remedying them. Applications Beyond Health and Well-being The potential impact of agency research extends beyond health and well-being. In this section I consider two of the areas where agency is (or should be) having an impact. Human-Computer-Interaction (HCI) Most of us spend a lot of time interacting with computers, both for work purposes and for social and leisure purposes. Perhaps because of the ubiquity of our interactions with computers we tend not to spend much time thinking about them (unless things go wrong or we buy a new computer/adopt a new operating system). Another reason why these interactions often go unnoticed might be that a lot of thought goes into designing the interface that sits between the computer and the user. User experience is at the heart of interface design, and this is where sense of agency comes in. It has long been recognized that the user’s sense of agency is an important consideration when designing new interfaces. Indeed, the seventh of Shneiderman’s Eight Golden Rules of Interface Design states that designers should create interfaces that “support an internal locus of control” (Shneiderman, 1992). This is based on the idea that users “strongly desire the sense that they are in charge of the system and that the system responds to their actions” (Shneiderman, 1992). In light of this, interface design will benefit greatly from scientific research on sense of agency – both in terms of measures that have been developed and in the understanding of what neurocognitive processes shape sense of agency. Some researchers are already trying to bridge the gap between HCI and sense of agency. In previous work we have looked at how different user interfaces impact on the user’s sense of agency. In one study (Coyle et al., 2012) we compared intentional binding for traditional keyboard input with intentional binding for a novel input modality called ‘skinput’ (whereby a user controls the computer by tapping on their own skin). We found that intentional binding was stronger for skinput, suggesting that this input modality increases the user’s sense of agency. In another study (Limerick et al., 2014) we measured intentional binding for a speech interface. We found that intentional binding was significantly reduced for the speech interface compared with the keyboard interface. Findings like these are potentially useful for those working in interface design. We have shown the utility of intentional binding as a measure of sense of agency – this may offer a more rigorous measure of the user’s sense of agency than currently used measures. We have also shown that different modalities are associated with differences in sense of agency. Given the importance of sense of agency for interface design, this is information is potentially useful – for example, when it comes to speech, our findings may help explain the difficulty and unpopularity of speech interfaces (Aylett et al., 2014). Interfaces such as skinput and speech require the user to make an over behavioral response. In brain-machine-interfaces (BMIs), there is no such requirement. As such, BMIs are a particularly exciting development in terms of our interaction with computers. By exploiting relatively recent advances in the acquisition and analysis of neuroimaging data, these technologies allow users to interact with computers and other devices without the need for overt behavioral responses. Of particular interest here is the user’s experience of agency given the absence of overt behavior. This was investigated by Evans et al. (2015). They found that agency judgments were modulated in a way that was similar to what one would expect for bodily movements, suggesting that BMIs can generate experiences of agency in users. Despite these similarities, they also observed that the effect of manipulated visual feedback in these tasks had a more pronounced effect on the user’s sense of agency than one would expect with bodily movements. This finding is consistent with the cue integration theory (e.g., Moore and Fletcher, 2012) as it shows that in the absence of strong internal motor cues, external visual feedback dominates. A more pressing practical concern in the context of HCI is the user’s sense of agency when interacting with automated technologies. Indeed, an increasing number of our interactions with computers and technology are being automated, with the system taking over a lot of the control that would have been in the hands of the user. Examples of automation include things like auto-correct in word processors through to driverless cars. Automation raises a number of issues when it comes to sense of agency; the most of obvious is the potential loss of sense of agency in the user. This is important if we remind ourselves of the prominence that the user’s sense of agency has in the design of interfaces. One area where automation is well-established is aircraft control – much of the pilot’s work is now carried out by a computer. Berberian et al. (2012) examined the sense of agency under different levels of automation in a flight simulator. They found that increasing automation reduced sense of agency. This finding is important, especially if sense of agency is linked to performance (something which is not yet established but would be interesting to look at). It is also important in situations where the automated system goes wrong – the attribution of responsibility in these situations, which is incredibly important socially and legally, might be guided by findings such as these. Freedom and Responsibility The attribution of responsibility, something that I touched on at the end of the previous section, is one of the key social functions of sense of agency (Frith, 2014). Humans seem to place a premium on responsibility – most, if not all, societies require that their members are held responsible for what they do. Haggard and Tsakiris (2009) have argued persuasively that sense of agency plays a key role in guiding attributions of responsibility. For Frith (2014) this bearing of responsibility for one’s own actions plays an important social function. It means that people can be held account for what they do which in turn allows behavior to be legitimately managed through punishment or reward. This behavioral management can be implemented so as to benefit the social group and promote social cohesion. In many societies the legal system is a tool which facilitates this kind of behavioral management. Given the importance of sense of agency for establishing responsibility, research in this area is therefore likely to have implications for the legal system. Responsibility is closely related the concept of free will. That is, for most people it only makes sense to hold someone responsible for their actions if they are freely in control of them. Indeed, the The Oxford English Dictionary defines free will as ’the power of an individual to make free choices, not determined by divine predestination, the laws of physical causality, fate, etc. Also: the doctrine that human beings possess this power and are hence able to direct and bear responsibility for their own actions’ (emphasis added). Free will is the elephant in the room when it comes to sense of agency research. Researchers tend to side step the issue of free will and instead focus solely on uncovering things like the neurocognitive basis of agentic experience. That is, whether or not we have free will, we unquestionably do have the experience of agency when we make actions and scientific research has tended to focus on understanding this experience. This evasion of the free will debate is understandable; philosophical debates on free will are often quite complex and confusing, especially for scientists with no background in philosophy. However, I think those of us working on this topic should try to engage more with this debate. In terms of impact, the social and legal consequences of this debate are immense, and our findings should be helping to inform this debate. Nichols (2011) has highlighted an interesting point of contact between sense of agency research and the free will debate. The free will problem arises because on the one hand we feel like conscious, rational free agents, and yet we recognize that this is incompatible with determinism. The relevance of sense of agency to this issue is that it is these experiences of agency surrounding our voluntary actions that give rise to the general feeling that we are conscious, rational free agents. According to Nichols, understanding the neurocognitive origins of free will beliefs will not tell us if they are true or not, but will help us evaluate whether or not those beliefs are justified. Although, this is just one of many possible links between free will and sense of agency, it does offer a potentially useful starting point for bringing the two fields together. There are of course some scientists who are willing to engage with the free will debate, and they are to be commended (see, for example, Wegner, 2002; Gazzaniga, 2011). However, these contributions can often take the form of what many philosophers regard as overly strong claims about our lack of free will. Mele (2014), for example, argues that many scientists who make such claims have a conception of free will that most philosophers, and indeed a lot of the general public, would not subscribe to (see also Dennett, 1984, 2004, for extensive discussions of different conceptions of free will). Other philosophers have contested this scientific work on different grounds. For example, Gallagher (2008) suggests that the scientific challenges to free will are misguided because they confuse the issue of mental causation (free will), with the issue of motor control. Because of this, these experiments can tell us something about motor control, but not free will. These examples from the philosophical literature show that if we scientists who work on sense of agency are to contribute meaningfully to the free will debate, there is a need for us to be more engaged with the philosophical work on the topic. Concluding Remarks In line with the stated aims of this article, I hope that I have provided the reader with a useful overview of the literature on sense of agency. I also hope that I have given the reader a sense of the potential importance and impact of this research. Much of the applied work is still in its infancy and I have only scratched the surface in terms of the potential applications of agency research. It is now up to us to drive this forward and to make a concerted effort to translate the findings of our more basic research on sense of agency into useful and effective applications. Author Contribution The author confirms being the sole contributor of this work and approved it for publication. Conflict of Interest Statement The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. ==== Refs References Aylett M. P. Kristensson P. O. Whittaker S. Vazquez Alvarez Y. (2014 ). “None of a CHInd: relationship counselling for HCI and speech technology ,” in Proceedings of the Extended Abstracts of ACM CHI 2014 (Toronto, Canada) (New York, NY : ACM ) 749 –760 . Berberian B. Sarrazin J.-C. Le Blaye P. Haggard P. (2012 ). Automation technology and sense of control: a window on human agency. PLoS ONE 7 :e34075 10.1371/journal.pone.0034075 Berti A. Spinazzola L. Pia L. Rabuffetti M. Haggard P. Rossetti Y. (2007 ). Motor awareness and motor intention in anosognosia for hemiplegia. Sensorimotor Found. High. Cogn. 163 –181 . Blakemore S. J. Smith J. Steel R. Johnstone C. E. Frith C. D. (2000 ). The perception of self-produced sensory stimuli in patients with auditory hallucinations and passivity experiences: evidence for a breakdown in self-monitoring. Psychol. Med. 30 1131 –1139 . 10.1017/S0033291799002676 12027049 Blakemore S. J. Wolpert D. M. Frith C. D. (1998 ). Central cancellation of self-produced tickle sensation. Nat. Neurosci. 1 635 –640 . 10.1038/2870 10196573 Blakemore S. J. Wolpert D. M. Frith C. D. (1999 ). The cerebellum contributes to somatosensory cortical activity during self-produced tactile stimulation. NeuroImage 10 448 –459 . 10.1006/nimg.1999.0478 10493902 Chambon V. Moore J. W. Haggard P. (2015 ). TMS stimulation over the inferior parietal cortex disrupts prospective sense of agency. Brain Struct. Funct. 220 3627 –3639 . 10.1007/s00429-014-0878-6 25134684 Coyle D. Moore J. Kristensson P. O. Fletcher P. Blackwell A. (2012 ). “I did that!: measuring users’ experience of agency in their own actions ,” in Proceedings of the 2012 ACM Annual Conference on Human Factors in Computing Systems (New York, NY : ACM ) 2025 –2034 . Daprati E. Franck N. Georgieff N. Proust J. Pacherie E. Dalery J. (1997 ). Looking for the agent: an investigation into consciousness of action and self-consciousness in schizophrenic patients. Cognition 65 71 –86 . 10.1016/S0010-0277(97)00039-5 9455171 Dennett D. C. (1984 ). Elbow Room: The Varieties of Free Will Worth Wanting. Cambridge, MA : MIT Press . Dennett D. C. (2004 ). Freedom Evolves. London : Penguin . Desantis A. Roussel C. Waszak F. (2011 ). On the influence of causal beliefs on the feeling of agency. Conscious. Cogn. 20 1211 –1220 . 10.1016/j.concog.2011.02.012 21396831 Evans N. Gale S. Schurger A. Blanke O. (2015 ). Visual Feedback dominates the sense of agency for brain-machine actions. PLoS ONE 10 :e0130019 10.1371/journal.pone.0130019 Farrer C. Frey S. H. Van Horn J. D. Tunik E. Turk D. Inati S. (2008 ). The angular gyrus computes action awareness representations. Cereb. Cortex 18 254 –261 . 10.1093/cercor/bhm050 17490989 Fletcher P. C. Frith C. D. (2009 ). Perceiving is believing: a bayesian approach to explaining the positive symptoms of schizophrenia. Nat. Rev. Neurosci. 10 48 –58 . 10.1038/nrn2536 19050712 Fotopoulou A. Tsakiris M. Haggard P. Vagopoulou A. Rudd A. Kopelman M. (2008 ). The role of motor intention in motor awareness: an experimental study on anosognosia for hemiplegia. Brain 131(Pt 12) 3432 –3442 . 10.1093/brain/awn225 18812442 Franck N. Farrer C. Georgieff N. Marie-Cardine M. Daléry J. d’Amato T. (2001 ). Defective recognition of one’s own actions in patients with schizophrenia. Am. J. Psychiatry 158 454 –459 . 10.1176/appi.ajp.158.3.454 11229988 Frith C. (2005 ). The self in action: lessons from delusions of control. Conscious. Cogn. 14 752 –770 . 10.1016/j.concog.2005.04.002 16098765 Frith C. D. (2014 ). Action, agency and responsibility. Neuropsychologia 55 137 –142 . 10.1016/j.neuropsychologia.2013.09.007 24036357 Frith C. D. Blakemore S. J. Wolpert D. M. (2000 ). Abnormalities in the awareness and control of action. Philos. Trans. R. Soc. Lond. B Biol. Sci. 355 1771 –1788 . 10.1098/rstb.2000.0734 11205340 Gallagher S. (2000 ). Philosophical conceptions of the self: implications for cognitive science. Trends Cogn. Sci. 4 14 –21 . 10.1016/S1364-6613(99)01417-5 10637618 Gallagher S. (2008 ). “Self-agency and mental causality ,” in Philosophical Issues in Psychiatry: Explanation, Phenomenology, and Nosology eds Kendler K. S. Parnas J. (Baltimore, MD : The Johns Hopkins University Press ). Gazzaniga M. S. (2011 ). Who’s in Charge?: Free Will and the Science of the Brain. New York, NY : HarperCollins . Gentsch A. Schütz-Bosbach S. Endrass T. Kathmann N. (2012 ). Dysfunctional forward model mechanisms and aberrant sense of agency in obsessive-compulsive disorder. Biol. Psychiatry 71 652 –659 . 10.1016/j.biopsych.2011.12.022 22305109 Haggard P. (2005 ). Conscious intention and motor cognition. Trends Cogn. Sci. 9 290 –295 . 10.1016/j.tics.2005.04.012 15925808 Haggard P. Clark S. Kalogeras J. (2002 ). Voluntary action and conscious awareness. Nat. Neurosci. 5 382 –385 . 10.1038/nn827 11896397 Haggard P. Tsakiris M. (2009 ). The Experience of agency feelings, judgments, and responsibility. Curr. Dir. Psychol. Sci. 18 242 –246 . 10.1111/j.1467-8721.2009.01644.x Henslin J. M. (1967 ). Craps and Magic. Am. J. Sociol. 73 316 –330 . 10.1086/224479 Lachman M. E. Firth K. M. (2004 ). “The adaptive value of feeling in control during midlife ,” in How Healthy Are We?: A National Study of Well-Being at Midlife eds Brim O. G. Ryff C. D. Kessler R. (Chicago : University of Chicago Press ) 320 –349 . Langer E. J. Rodin J. (1976 ). The effects of choice and enhanced personal responsibility for the aged: a field experiment in an institutional setting. J. Pers. Soc. Psychol. 34 191 –198 . 10.1037/0022-3514.34.2.191 1011073 Limerick H. Coyle D. Moore J. W. (2014 ). The experience of agency in human-computer interactions: a review. Front. Hum. Neurosci. 8 :643 10.3389/fnhum.2014.00643 McRaney D. (2013 ). You Can Beat Your Brain: How to Turn Your Enemies Into Friends, How to Make Better Decisions, and Other Ways to Be Less Dumb. London : Oneworld Publications . Mele A. R. (2014 ). Free: Why Science Hasn’t Disproved Free Will. Oxford : Oxford University Press . Mellor C. S. (1970 ). First rank symptoms of schizophrenia: I. The frequency in schizophrenics on admission to hospital. II. Differences between individual first rank symptoms. Br. J. Psychiatry 117 15 –23 . 10.1192/bjp.117.536.15 5479324 Metcalfe J. Eich T. S. Castel A. D. (2010 ). Metacognition of agency across the lifespan. Cognition 116 267 –282 . 10.1016/j.cognition.2010.05.009 20570251 Mirowsky J. (1995 ). Age and the Sense of Control. Soc. Psychol. Q. 58 31 –43 . 10.2307/2787141 Moore J. Haggard P. (2008 ). Awareness of action: inference and prediction. Conscious. Cogn. 17 136 –144 . 10.1016/j.concog.2006.12.004 17306565 Moore J. W. Fletcher P. C. (2012 ). Sense of agency in health and disease: a review of cue integration approaches. Conscious. Cogn. 21 59 –68 . 10.1016/j.concog.2011.08.010 21920777 Moore J. W. Obhi S. S. (2012 ). Intentional binding and the sense of agency: a review. Conscious. Cogn. 21 546 –561 . 10.1016/j.concog.2011.12.002 22240158 Moore J. W. Wegner D. M. Haggard P. (2009 ). Modulating the sense of agency with external cues. Conscious. Cogn. 18 1056 –1064 . 10.1016/j.concog.2009.05.004 19515577 Nichols S. (2011 ). Experimental philosophy and the problem of free will. Science 331 1401 –1403 . 10.1126/science.1192931 21415346 Oren E. Friedmann N. Dar R. (2016 ). Things happen: individuals with high obsessive-compulsive tendencies omit agency in their spoken language. Conscious. Cogn. 42 125 –134 . 10.1016/j.concog.2016.03.012 27003263 Rodin J. Langer E. J. (1977 ). Long-term effects of a control-relevant intervention with the institutionalized aged. J. Pers. Soc. Psychol. 35 897 –902 . 10.1037/0022-3514.35.12.897 592095 Sass L. A. (2014 ). Self-disturbance and schizophrenia: structure, specificity, pathogenesis (current issues, new directions). Schizophr. Res. 152 5 –11 . 10.1016/j.schres.2013.05.017 23773296 Shanks D. R. Pearson S. M. Dickinson A. (1989 ). Temporal contiguity and the judgement of causality by human subjects. Q. J. Exp. Psychol. B 41 139 –159 . 10.1080/14640748908401189 Shneiderman B. (1992 ). Designing the User Interface: Strategies for Effective Human-Computer Interaction (Vol. 2). Reading, MA : Addison-Wesley . Synofzik M. Thier P. Leube D. T. Schlotterbeck P. Lindner A. (2010 ). Misattributions of agency in schizophrenia are based on imprecise predictions about the sensory consequences of one’s actions. Brain 133 262 –271 . 10.1093/brain/awp291 19995870 Synofzik M. Vosgerau G. Newen A. (2008 ). Beyond the comparator model: a multifactorial two-step account of agency. Conscious. Cogn. 17 219 –239 . 10.1016/j.concog.2007.03.010 17482480 Voss M. Moore J. Hauser M. Gallinat J. Heinz A. Haggard P. (2010 ). Altered awareness of action in schizophrenia: a specific deficit in predicting action consequences. Brain 133 3104 –3112 . 10.1093/brain/awq152 20685805 Wegner D. M. (2002 ). The Illusion of Conscious Will. Cambridge, MA : MIT Press . Wegner D. M. Sparrow B. (2004 ). Authorship Processing. Available at: http://doi.apa.org/psycinfo/2005-01373-079 Wegner D. M. Sparrow B. Winerman L. (2004 ). Vicarious agency: experiencing control over the movements of others. J. Pers. Soc. Psychol. 86 838 10.1037/0022-3514.86.6.838 Wegner D. M. Wheatley T. (1999 ). Apparent mental causation. Sources of the experience of will. Am. Psychol. 54 480 –492 .10424155 Wolpert D. M. Miall R. C. (1996 ). Forward models for physiological motor control. Neural Netw. 9 1265 –1279 . 10.1016/S0893-6080(96)00035-4 12662535

PMC005xxxxxx/PMC5002401.txt

==== Front Front Plant SciFront Plant SciFront. Plant Sci.Frontiers in Plant Science1664-462XFrontiers Media S.A. 10.3389/fpls.2016.01270Plant ScienceOriginal ResearchGenome-Wide Association Studies Reveal that Diverse Heading Date Genes Respond to Short and Long Day Lengths between Indica and Japonica Rice Han Zhongmin 1Zhang Bo 1Zhao Hu 1Ayaad Mohammed 1Xing Yongzhong 1*1National Key Laboratory of Crop Genetic Improvement and National Center of Plant Gene Research, Huazhong Agricultural UniversityWuhan, China2Hubei Collaborative Innovation Center for Grain IndustryJingzhou, ChinaEdited by: Henry T. Nguyen, University of Missouri, USA Reviewed by: Mingsheng Chen, Chinese Academy of Sciences, China; Rupesh Kailasrao Deshmukh, Laval University, Canada *Correspondence: Yongzhong Xing, yzxing@mail.hzau.edu.cnThis article was submitted to Plant Genetics and Genomics, a section of the journal Frontiers in Plant Science 29 8 2016 2016 7 127022 1 2016 10 8 2016 Copyright © 2016 Han, Zhang, Zhao, Ayaad and Xing.2016Han, Zhang, Zhao, Ayaad and XingThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.Rice is a short-day plant. Short-day length promotes heading, and long-day length suppresses heading. Many studies have evaluated rice heading in field conditions in which some individuals in the population were exposed to various day lengths, including short and long days, prior to a growth phase transition. In this study, we investigated heading date under natural short-day conditions (SD) and long-day conditions (LD) for 100s of accessions and separately conducted genome-wide association studies within indica and japonica subpopulations. Under LD, three and four quantitative trait loci (QTLs) were identified in indica and japonica subpopulations, respectively, two of which were less than 80 kb from the known genes Hd17 and Ghd7. But no common QTLs were detected in both subpopulations. Under SD, six QTLs were detected in indica, three of which were less than 80 kb from the known heading date genes Ghd7, Ehd1, and RCN1. But no QTLs were detected in japonica subpopulation. qHd3 under SD and qHd4 under LD were two novel major QTLs, which deserve isolation in the future. Eleven known heading date genes were used to test the power of association mapping at the haplotype level. Hd17, Ghd7, Ehd1, and RCN1 were again detected at more significant level and three additional genes, Hd3a, OsMADS56, and Ghd7.1, were detected. However, of the detected seven genes, only one gene, Hd17, was commonly detected in both subpopulations and two genes, Ghd7 and Ghd7.1, were commonly detected in indica subpopulation under both conditions. Moreover, haplotype analysis identified favorable haplotypes of Ghd7 and OsMADS56 for breeding design. In conclusion, diverse heading date genes/QTLs between indica and japonica subpopulations responded to SD and LD, and haplotype-level association mapping was more powerful than SNP-level association in rice. heading datelong and short-day conditionsgenome-wide association studieshaplotype-level association ==== Body Introduction Rice as a short-day plant is cultivated from latitudes of 55°N in China to 36°S in Chile (Khush, 1997). Rice heading date is an important trait to adapt to various regional environments and is usually related to grain yield. Many isolated QTLs showed varying photoperiod sensitivities between SD and LD. Under SD, heading date is promoted by two independent pathways mediated by Heading date 1 (Hd1) and Early heading date 1 (Ehd1). Hd1 is a homolog of CONSTANS in Arabidopsis. Ehd1 encoding a B-type response regulator has no ortholog in Arabidopsis. Hd1 and Ehd1 up-regulated the expression of the FT-like (florigen) Hd3a genes (Yano et al., 2000; Doi et al., 2004). Ehd2, Ehd3, and Ehd4 promote heading by up-regulating Ehd1 under SD conditions (Matsubara et al., 2008, 2011; Gao et al., 2013). Most day-length-sensitive heading date QTLs, such as Hd6, Ghd7, OsMADS56, OsMADS50, Hd17, DTH2, Hd16, and Ghd7.1, function specifically under LD (Takahashi et al., 2001; Lee et al., 2004; Xue et al., 2008; Ryu et al., 2009; Matsubara et al., 2012; Hori et al., 2013; Wu et al., 2013; Zhang et al., 2013). Among these QTLs, OsMADS50, Hd17, and DTH2 promote heading but the others delay heading under LD. Interestingly, both Hd1 and DTH8/Ghd8 delay heading under LD and promote heading under SD (Yano et al., 2000; Wei et al., 2010). A genome-wide association study (GWAS) combining high-density markers with a diverse germplasm collection provides higher mapping resolution than conventional QTL mapping based on bi-parent derived segregating populations and enables the prediction or identification of causal genes. GWAS has been widely used to identify loci related to various important agronomical traits in rice (Huang et al., 2010, 2012; Zhao et al., 2011). For heading date, Huang et al. (2012) associated three known heading date genes of OsGI, Hd3a, and RCN1 with worldwide rice varieties in natural conditions. They were all detected in indica subpopulation but OsGI was also detected in japonica subpopulation. However, the major gene Ghd7 was not detected. Zhao et al. (2011) conducted experiments in three environments: LD (12–14 h), 12–13 h day-length condition in the field and one very long to very short day-length condition (~18–6 h) in a greenhouse. They identified 10 candidate heading date genes. Hd1 was the only one repeatedly detected in multiple environments. In both studies, the power of genome-wide association mapping for heading date QTLs is low because few known genes have been identified. The low power of association mapping at the single-nucleotide polymorphism (SNP) level is primarily caused by low resolution in grouping samples into two genotypes. Notably, for adaptive genes in rice such as Ghd7 and Ghd7.1, several alleles/haplotypes exist in nature. Each allele has a different genetic effect and results in a specific geographic distribution (Xue et al., 2008; Zhang et al., 2013). Therefore, association mapping at the haplotype level is expected to be more powerful because the germplasm collection can be classified into several genotypes (as opposed to two genotypes via SNP). Additionally, the heading date data collected from the population grown under mixed LD and SD conditions for GWAS likely leads to deviation because some heading date genes such as Hd1 have opposing effects on heading between SD and LD, which eliminate their genetic effects and result in failed detection. Indica and japonica rice varieties have distinct growing zones and cropping seasons. It is naturally questioned whether different genes control heading date between indica and japonica subpopulations under LD and SD. To answer the questions, heading date was recorded for a global rice collection under SD and LD in this study, and GWAS for heading date was independently performed in indica and japonica subpopulations under SD and LD. To check the power of haplotype-level GWAS, 11 known heading date genes were tested. Our results revealed a diverse genetic basis for heading date between indica and japonica subpopulations under two conditions and a greatly improved power of haplotype-level GWAS as compared to SNP-level GWAS. Materials and Methods Plant Material, Field Experiments, and Heading Date Record A diverse worldwide collection of 529 Oryza sativa accessions were planted on April 17 in a bird net-equipped field in the experimental farm of Huazhong Agricultural University, Wuhan, China (114° 21′ E, 30° 28′ N) during the summer seasons of 2013 and 2014 and in the winter season, 2012–2013 (on December 5) in Hainan Island, China (110° 01′ E, 18° 30′ N). Field trials were carried out following a randomized complete block design with two replications within each year. The seedlings of each accession were transplanted after 25 days into one row in the field, with a distance of 16.5 cm × 26.4 cm within and between the rows. The heading dates of five plants in the middle of each row were individually recorded. Every 2 days, the heading date was recorded as the days from sowing to the first panicle appearance. The average trait value for each accession across two replicates within each year was used for separate GWAS. SNP Database In total, 529 O. sativa landrace and elite accessions were genotyped via sequencing (Chen et al., 2014). The SNP information is available on RiceVarMap1, a comprehensive database of rice genomic variations, while the SNP physical locations were cited using the genome of TIGR Rice Loci 6. Genome-Wide Association Study A total of 2,767,159 and 1,857,845 SNPs (minor allele frequency ≥0.05; the number of accessions with minor alleles ≥6) were used for GWAS in indica and japonica subpopulations, respectively (Chen et al., 2014). Linear mixed models (LMM) were used to make associations by running the Fast-LMM program (Lippert et al., 2011). Population structure was controlled using a kinship matrix constructed with all SNPs. Using a method described by (Li et al., 2012), the effective numbers of independent SNPs were calculated as 571,843 and 245,348 for both indica and japonica subpopulations, respectively. The suggested P-values were specified as 1.8 × 10-6 in indica and 1.3 × 10-6 in japonica (Chen et al., 2014). The thresholds were then set at P = 8.0 × 10-7 in Wuhan and P = 3.8 × 10-7 in Hainan to identify significant association signals via LMM. To obtain independent association signals, multiple SNPs exceeding the threshold in a 5-Mb sliding window were clustered by r2 of linkage disequilibrium ≥0.25, and SNPs with the minimum P-value in a cluster were considered lead SNPs. The detailed method was previously described (Chen et al., 2014; Wang et al., 2015). Haplotype and Statistical Analyses SNPs within genes used to define haplotypes for 11 known heading date genes including Hd17, Ghd7, Ehd1, RCN1, Hd1, Ghd8, Ghd7.1, DTH2, Hd3a, OsMADS50, and OsMADS56 were downloaded2; the haplotypes were then reconstructed using PHASE software (Stephens et al., 2001) to do imputation for the missing and heterozygous SNPs. The haplotypes with an allele frequency ≥0.01 (5 accessions) were included for association analysis via ANOVA; then, a Duncan test (Post-hoc analysis) was conducted to show the difference in heading date between all possible haplotype pairs. For each condition, the lead SNPs of the associations detected via GWAS were fit into a linear model using an lm function in R software3, and the phenotype variations explained by each lead SNP and by all lead SNPs together were estimated. Results Day Length of the Rice-Growing Season in Hainan and Wuhan The day length of 13.5 h was set as the boundary between SD and LD (Itoh et al., 2010). Throughout the winter growing season (December–April) in Hainan, the day length ranged from 11.0 to 12.5 h, which is a typical SD. Therefore, all 529 accessions completed their life cycles under SD in Hainan. However, in Wuhan, the summer season day length increased from 13.0 to 14.2 h (April 17–June 22) and then decreased from 14.2 to 12.2 h (June 22–September 21). Therefore, some accessions completed heading under LD; however, some accessions first encountered LD and then experienced SD prior to heading. Approximately 25 days were required for rice to complete panicle development from the growth phase transition to heading (Vergara et al., 1966; Chang et al., 1969; Liu et al., 2013). In Wuhan, the day length decreased from 13.5 h after August 6, thus the accessions that completed heading prior to August 30 initiated the transition from a vegetative to a reproductive phase under LD. To precisely evaluate the effects of LD conditions on heading date, we removed 26 accessions that experienced mixed LD and SD prior to the phase transition and kept 503 accessions that completed the phase transition under LD (headed prior to August 30) in Wuhan. Variation of Heading Date in 503 Diverse Accessions A structural analysis indicated that the entire collection was divided into several distinct subpopulations such as indica, japonica, aus and mixture (Chen et al., 2014). Heading date exhibited large variation in the 503 accessions, the mean values of heading date in the indica subpopulation were equivalent to those in japonica rice either in Wuhan or Hainan, with the exception of heading date in 2012 in Hainan (Figure 1). The correlation coefficient of heading date between the 2 years was 0.94 in Wuhan and 0.86 in Hainan. The correlation coefficients between different locations were significant but substantially lower than that between years in the same locations (Table 1). Two-way ANOVA showed significant genotype and environment effects on heading date (Table 2). The genetic factor accounted for the most phenotypic variance (59.5%) and the environmental effect contributed 4.6% of the phenotypic variance, indicating that many accessions showed differing heading dates between the four environments. FIGURE 1 Performance of heading date in the indica (A,B) and japonica (C,D) subpopulations. The circles indicated mean values of heading date. Table 1 Correlations of heading dates of 503 accessions in different environments. Environments HN2013 WH2013 WH2014 HN2012 0.86 0.37 0.31 HN2013 0.44 0.42 WH2013 0.94 Table 2 Genotypes and environments analysis of variance for heading date using 503 accessions. SV Heading date SSG/SST (%) F P G 59.5 4.8 <0.0001 E 4.6 63.5 <0.0001 SV, source of variance; G, genotype; E, environment.QTL Detection Using GWAS in Wuhan (LD) The 278 indica and 147 japonica accessions (after removed 78 of the aus or admixed subpopulations) were utilized for GWAS to avoid structure noise. GWAS was separately performed in indica and japonica subpopulations. A total of seven associations with heading date were detected in Wuhan, three and four QTLs were detected in indica and japonica subpopulations, respectively (Figure 2; Table 3). Two QTLs (qHd6.1 and qHd7.1) on chromosomes 6 and 7 were repeatedly detected in indica subpopulation and individually explained 18.7–19.7% and 7.1–7.3% of the phenotypic variance, respectively. The lead SNP Sf602312322 for qHd6.1 was 78.2 kb to Hd17, and the lead SNP Sf709177919 for qHd7.1 located in 25.5 kb from Ghd7. In japonica subpopulation, qHd6.2 was repeatedly associated with the heading date in 2 years, however, there are no reported heading date genes nearby qHd6.2. qHd4 was detected only in 2013 but exhibited the largest contribution (26.1%) to the variation in heading date. No common association was detected in either subpopulation. The QTLs detected in indica subpopulation cumulatively explained 27.1 and 35.3% of the phenotypic variations in 2013 and 2014, respectively. The QTLs in japonica subpopulations explained more variation in the heading date (39.4 and 40.8% in the 2 years, respectively). FIGURE 2 Genome-wide P-values (A,C,E,G) and QQ plots (B,D,F,H) from the LMM model for heading date in indica and japonica in Wuhan are demonstrated in the four panels. The horizontal dashed line (A,C,E,G) indicates the genome-wide significance threshold (P = 8.0 × 107). Table 3 Genome-wide significant association signals for heading date in indica and japonica subpopulations using the LMM method in Wuhan (long day conditions). QTL Lead SNP Chr. Linkage disequilibrium interval (bp) Distance to gene Indica 2013 Indica 2014 Japonica 2013 Japonica 2014 P P.V (%) P P.V (%) P P.V (%) P P.V (%) qHd4 Sf0403512804 4 3145576–3733894 1.8E-7 26.1 qHd6.1 Sf0602312322 6 2311322–2931955 78.2 kb to Hd17 1.6E-6 19.7 4.4E-7 18.7 qHd6.2 Sf0602663020 6 2396433–2733565 8.0E-7 13.2 4.4E-6 16.8 qHd7.1 Sf0709177919 7 9137593–9196135 -25.5 kb to Ghd7 5.7E-7 7.3 1.1E-6 7.1 qHd7.2 Sf0711412802 7 11316108–11898697 3.1E-7 16.6 qHd8 Sf0804089477 8 3942483–4245642 3.5E-8 9.5 qHd10.1 Sf1001634949 10 1596335–1634949 2.2E-7 7.4 Total estimation 7.2E-19 27.1 4.0E-25 35.3 3.8E-16 39.4 3.2E-15 40.8 Linkage disequilibrium interval means the linkage disequilibrium (r2 > 0.2) extended genome region surrounding the lead SNP. P is P-value estimated in LMM, P.V (%) is phenotype variance explained by the QTL in percent.QTL Detection via GWAS in Hainan (SD) Six heading date QTLs in indica subpopulation and no QTLs in japonica subpopulation were detected under SD (Figure 3; Table 4). Two QTLs (qHd6.3 and qHd7.1) on chromosomes 6 and 7 were repeatedly detected in 2 years indica subpopulation. qHd6.3 and qHd7.1 primarily explained 26.5% and 10.0% of the phenotypic variations, respectively. In total, the QTLs explained 57.7% and 41.3% of the phenotypic variation, respectively, in 2 years in indica subpopulation. Three QTLs were located near known heading date genes: the lead SNP Sf709158944 for qHd7.1 was 6.5 kb to Ghd7, the lead SNP Sf1017100637 for qHd10.1 was 24.5 kb to Ehd1 and the lead SNP Sf1102480009 for qHd11 was 27.5 kb to RCN1. FIGURE 3 Genome-wide P-values (A,C,E,G) and QQ plots (B,D,F,H) from the LMM model for heading date in indica and japonica in Hainan are demonstrated in the four panels. The horizontal dashed line (A,C,E,G) indicates the genome-wide significance threshold (P = 3.8 × 107). Table 4 Genome-wide significant association signals for heading date in the indica subpopulations using the LMM method in Hainan (short day conditions). QTL Lead SNP Chr. Linkage disequilibrium interval (bp) Distance to gene Indica 2012 Indica 2013 P P.V (%) P P.V (%) qHd3 Sf0322143338 3 22079360–22143338 1.4E-8 23.3 qHd6.3 Sf0607922178 6 7917158–8424734 1.3E-9 22.0 1.1E-8 26.5 qHd7.1 Sf0709158944 7 9137593–9196135 -6.5 kb to Ghd7 1.2E-7 7.7 2.0E-6 10.0 qHd7.3 Sf0712842846 7 12774834–12894276 1.4E-7 1.1 qHd10.2 Sf1017100637 10 16691541–17385511 -24.5 kb to Ehd1 2.9E-7 4.7 qHd11 Sf1102480009 11 2357177–2488304 -27.5 kb to RCN1 3.8E-07 3.5 Total estimation 2.1E-46 57.7 1.0E-30 41.3 Linkage disequilibrium interval means the linkage disequilibrium (r2> 0.2) extended genome region surrounding the lead SNP. P is P-value estimated in LMM, P.V (%) is phenotype variance explained by the QTL in percent.Haplotype-Level Association Analysis of 11 Known Heading Date Genes To date, dozens of heading date genes have been cloned in rice. However, only four QTLs in our study were located near heading date genes, which were less than 80 kb from the corresponding lead SNPs. The average linkage disequilibrium decay in these indica and japonica groups extended 93 and 171 kb, respectively4, which was similar to the estimates of 75–123 kb in indica and 150–167 kb in japonica in other studies (Zhao et al., 2011; Huang et al., 2012). In our study, lead SNPs pointed to the four QTLs (qHd6.1, qHd7.1, qHd10.1, and qHd11) were in linkage disequilibrium with known heading date genes of Hd17, Ghd7, Ehd1, and RCN1, respectively (Tables 3 and 4). Thus, these heading date genes most likely underlined these four QTLs. We questioned whether known heading gene had a greater chance to be associated by haplotype, a combination of a set of SNPs within gene. We then extracted SNPs from the initiation codon to the termination codon of 11 genes: Hd17, Ghd7, Ehd1, RCN1, Hd1, Ghd8, Ghd7.1, DTH2, Hd3a, OsMADS50, and OsMADS56; haplotypes were then constructed. Most of these genes had more haplotypes in indica subpopulation than japonica subpopulation, with the exception for Ghd7.1 and OsMADS56 (Table 5). The associations were re-estimated at the haplotype level. The four heading date genes detected using GWAS at the SNP level were identified via haplotype association. Accordingly, Ghd7 was identified in the indica subpopulation under both SD and LD. However, RCN1 was detected in indica under SD, and Hd17 was detected in both subpopulations under LD. Ehd1 was detected in japonica under SD. Additionally, Ghd7.1, Hd3a, and OsMADS56 were detected via haplotype association (Supplementary Figure S1; Table 5). Ghd7.1 was identified in japonica subpopulation under both SD and LD. OsMADS56 was specifically identified under SD in japonica subpopulation. However, Hd3a was specifically identified in indica under LD. Table 5 Association analyses for known heading date genes in the indica and japonica subpopulations based on haplotypes using 425 accessions. Gene N (Ind, Jap)∗ Subpop Wuhan Hainan P (2013) P (2014) P (2012) P (2013) Hd17 6 (6, 2) Ind 9.4E-06 2.6E-08 Jap 1.3E-08 1.5E-08 Hd3a 6 (6, 2) Ind 1.8E-07 3.8E-09 RCN1 3 (3, 1) Ind 2.6E-11 2.3E-08 Ehd1 5 (3, 3) Jap 5.3E-16 2.0E-12 OsMADS56 7 (3, 4) Jap 1.8E-09 6.3E-09 Ghd7 6 (4, 2) Ind 1.5E-19 8.2E-17 3.1E-18 9.4E-10 Ghd7.1 14 (5, 10) Jap 1.5E-11 5.9E-11 3.5E-21 1.9E-23 ∗N was the number of haplotypes in 425 accessions. The numbers inside the parentheses are the numbers of haplotypes in the indica and japonica subpopulations, respectively. P is P-value in ANOVA.Haplotype Analysis for Ghd7 and OsMADS56 Ghd7 is a key regulator of heading date and has five haplotypes with strong, weak or null-functions in cultivars (Xue et al., 2008; Lu et al., 2012). To obtain insight into why Ghd7 was not detected in japonica, we compared the haplotypes of Ghd7 in indica and japonica subpopulations (Table 6). In addition to the previously reported five haplotypes, one additional haplotype (Ghd7-4) was identified. In indica subpopulation, strong alleles of Ghd7-1 and Ghd7-3, weak allele of Ghd7-4 and non-functional allele of Ghd7-0 were observed, which caused a large variation in heading date and resulted in a successful detection. However, in japonica subpopulation, we only detected weak functional alleles of Ghd7-2 and non-functional alleles of Ghd7-0a, resulted a small variation and failed detection. The similar result was observed for OsMADS56 (Table 7). That is, indica rice carried haplotypes with similar genetic effects on heading date and japonica rice carried haplotypes with distinct genetic effects, these data explain why OsMADS56 was detected in only japonica subpopulation. Table 6 Comparison of heading date (days) among haplotypes of Ghd7 in indica and japonica subpopulations. Pop Haplotypes Number Hainan 2012 Hainan 2013 Wuhan 2013 Wuhan 2014 Indica Ghd7-1 161 98.7 ± 11.4C 98.4 ± 7.7C 101.4 ± 11.5C 103.7 ± 11.7C Ghd7-3 70 96.0 ± 10.1C 96.6 ± 7.0C 96.6 ± 6.8B 97.4 ± 6.0B Ghd7-4 18 72.4 ± 12.4B 86.4 ± 7.9B 76.8 ± 10.9A 82.5 ± 12.4A Ghd7-0 6 79.8 ± 16.6B 88.1 ± 10.1B 78.7 ± 11.2A 82.6 ± 9.1A Japonica Ghd7-2 136 87.2 ± 22.3b 94.4 ± 12.2b 97.3 ± 17.2b 100.0 ± 17.4b Ghd7-0a 5 59.0 ± 5.1a 79.9 ± 4.2a 61.0 ± 5.7a 68.2 ± 4.5a The phenotype values are represented as the means ± SD, letters are ranked via Duncan test at P < 0.01.Table 7 Comparison of heading date among haplotypes of OsMADS56 in indica and japonica subpopulations in Hainan. Pop Haplotypes Number Hainan 2012 Hainan 2013 Indica OsMADS56-1 59 97.5 ± 10.9A 98.0 ± 7.8A OsMADS56-2 145 93.3 ± 15.2A 95.7 ± 9.0A OsMADS56-3 28 99.8 ± 11.1A 99 ± 7.9A Japonica OsMADS56-4 64 73.7 ± 18.5a 87.9 ± 7.6a OsMADS56-5 13 101.5 ± 20c 101.5 ± 10.5c OsMADS56-6 32 92.7 ± 22.8b 95.9 ± 13.5b OsMADS56-7 9 111.6 ± 8.1d 108.4 ± 7.0d The phenotype values are represented as the means ± SD, letters are ranked via Duncan test at P < 0.01.Discussion Diverse Genetic Basis of Heading Date between Indica and Japonica Subpopulations The large range in heading dates in the indica subpopulation was similar to that in the japonica subpopulation in each environment (Figure 1), which indicated a complicated genetic basis heading date in rice. However, no common QTLs were detected at the SNP level in two subpopulations (Tables 3 and 4). Under LD, Hd17 and Ghd7 were the primary heading genes in indica subpopulation; whereas qHd4, qHd6.2, and qHd7.2 were the primary players in japonica subpopulation. Under SD conditions, there were no genomic regions associated with heading date in japonica subpopulation, which indicated that the heading date genes in japonica had a photoperiod sensitivity that was too weak to efficiently function under SD conditions and led to a failed detection (Table 4). On contrary, a few QTLs were identified in indica subpopulation, which indicated that indica rice carried strong photoperiod-sensitive genes, which affected heading date even under SD. These results show that indica and japonica subpopulations have different gene/allele responses to day length. The power of association mapping was greatly improved at the haplotype level; however, only one heading gene (Hd17) was commonly identified in both japonica and indica subpopulations under LD (Table 5). Moreover, the 7 associated known heading genes (with the exception of OsMADS56 and Ghd7.1) had more haplotypes in indica than japonica, which indicated that indica rice in regulating heading date was more genetically diverse than japonica. This result is in agreement with the previous finding that the genetic diversity in indica (0.0016) was much higher than that (0.0006) in japonica (Huang et al., 2010). There were 10 alleles of Ghd7.1 in japonica and five alleles in indica; however, only one allele was common across the two subpopulations. This finding was consistent with the reported assumption that indica and japonica alleles of Ghd7.1 evolved independently (Zhang et al., 2013). In most cases, japonica rice carries weak or non-functional alleles of photoperiod-sensitive genes, which enable it to be grown at high latitudes with LD during the rice-growing season. Indica rice carries diverse alleles, including strong, weak and non-functional alleles, which enable its growth across various climatic regions. Different Heading Date Gene Responses to Long and Short Day Lengths In this study, we detected QTLs for heading date using GWAS under SD (Hainan) and LD (Wuhan). At the SNP level, no common QTL was identified between SD and LD in japonica (Tables 3 and 4). In indica, only the known major gene Ghd7 was commonly detected under both conditions. For the 11 known heading date genes, seven were identified at the haplotype level. However, only Ghd7 in indica and Ghd7.1 in japonica were commonly detected under both conditions (Table 5). This indicated that Ghd7 and Ghd7.1 simultaneously contributed to the variation in heading date under both conditions, which is in agreement with the findings that Ghd7 and Ghd7.1 primarily function under LD, and they have weak effects under natural short day conditions (Xue et al., 2008; Zhang et al., 2013). As we expected, Ehd1, a major heading date gene that primarily functions under SD conditions (Doi et al., 2004), was detected under SD conditions. Accordingly, the gene Hd17, which primarily functions under LD conditions (Matsubara et al., 2012), was detected under LD conditions. Hd3a, a rice florigen gene with expression that is repressed under LD and promoted under SD (Tamaki et al., 2007; Tsuji et al., 2011), was detected under LD in this study. This finding indicated that probably the upstream regulators act with Hd3a more actively under LD than SD. However, OsMADS56 and RCN1, two major heading date genes primarily working under LD (Nakagawa et al., 2002; Ryu et al., 2009), were detected under SD conditions. This inconsistency indicated that these genes likely have a new function, or association mapping is not sufficiently powerful so that some key genes are not detected. These unknown QTL detected under SD or LD may primarily regulate heading date dependent on short or long day length, respectively. Different QTL detected under SD and LD suggested that heading date genes have various photoperiod sensitivities. A parallel experimental design similar to this study is suggested for identifying and characterizing the genes with responses to environmental factors such as temperature or fertilizers. No heading date genes have been reported around the major QTL qHd3 in indica under SD and qHd4 in japonica under LD. They are likely new heading date genes. Populations designed for mapping both QTL are encouraged to confirm their identities. Haplotype-Level Association Analysis Improved the Power of Mapping In rice, several papers reported GWAS for important agronomic or biological traits at the SNP level. However, at least some or even a large portion of known functional genes was not identified (Zhao et al., 2011; Huang et al., 2012). The most likely reason for this is that in the most cases, SNP classification divided the collection into only two classes, whereas a gene has multiple alleles exhibiting different genetic effects in a natural population. Therefore, an SNP-derived low-resolution classification confused functional and non-functional alleles that were distinguished in multiple SNP combinations and compromised the real differences between alleles, thereby producing a failed detection. A haplotype is the genetic constitution of one locus/gene, which is specified by a minimum number of SNPs. At the single gene level, a haplotype is equal to an allele. In consequence, a haplotype-level GWAS is expected to mine more genes hidden in the whole genome because a haplotype-level GWAS shows the difference between haplotypes rather than conflating these differences. Notably, most genes excluding housekeeping genes include several haplotypes/alleles, such as Ghd7 and OsMADS56, in the germplasm collection (Tables 6 and 7). Only four known heading genes (Ehd1, Hd17, RCN1, and Ghd7) were identified at the SNP level. However, for the 11 known genes, seven (including the four detected via SNP-level GWAS (SGWAS)) were identified via haplotype-level association analysis (Supplementary Figure S1; Table 5). Moreover, the order of magnitudes of the P-values for these four genes was greatly decreased comparing with SGWAS (Tables 3–5). Obviously, haplotype-level association analysis immensely increased the mapping power. In this study, we choose only 11 known genes as candidates to test the haplotype-level association analysis. Moreover, haplotype analysis ranked the genetic effects of haplotypes, which provided favorable haplotypes for breeding design. However, a haplotype-level genome-wide association study (HGWAS) faces substantial challenges because several points should be considered when constructing genome-wide haplotypes. (i) The haplotype specification should be made per gene unit or coding sequence of one gene that potentially matches a phenotypic variation, (ii) The species has a high-quality reference genome sequence for SNP screening and high-quality full-length cDNA sequences to delimit the gene region, (iii) The minimum deep sequencing and high-quality sequencing data are required for the identification of haplotypes by decreasing imputations of the missing and heterozygous sites and alleviating the calculation burden. The association mapping methods should also be further developed for HGWAS. Conclusion GWAS revealed that diverse genes/QTLs between indica and japonica subpopulations regulated heading date under SD and LD. Test of haplotype-level association mapping for 11 known heading date genes confirmed that the power of association mapping was greatly improved as compared to SNP-level association mapping and HGWAS approach is encouraged to develop in rice. Author Contributions ZH performed the experiments, analyzed the data and wrote the paper. BZ performed the experiments, HZ analyzed data, and MA wrote the paper. YX conceived the project, designed the research, and wrote the paper. Conflict of Interest Statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Funding. This work was supported by grants from the National Special Program for Research of Transgenic Plant of China (2011ZX08009-001-002), the National Natural Science Foundation of China (91335201). We thank the field technician Mr. Jianbo Wang for his excellent work in the experimental field. 1 http://ricevarmap.ncpgr.cn 2 http://ricevarmap.ncpgr.cn 3 http://www.r-project.org/ 4 http://ricevarmap.ncpgr.cn Supplementary Material The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2016.01270 FIGURE S1 Physic map of associated QTLs and candidate genes via SNP-level GWAS (SGWAS) and haplotype-level GWAS (HGWAS). The black QTLs are detected by SGWAS. The red heading date genes are detected by HGWAS. The blue heading date genes are detected by both SGWAS and HGWAS. The purples on the left side of each chromosome are heading date genes are failed in detection by GWAS. Click here for additional data file. Click here for additional data file. ==== Refs References Chang T. T. Li C. C. Vergara B. S. (1969 ). Component analysis of duration from seeding to heading in rice by the basic vegetative phase and the photoperiod-sensitive phase. Euphytica 18 79 –91 . 10.1007/BF00021985 Chen W. Gao Y. Xie W. Gong L. Lu K. Wang W. (2014 ). Genome-wide association analyses provide genetic and biochemical insights into natural variation in rice metabolism. Nat. Genet. 46 714 –721 . 10.1038/ng.3007 24908251 Doi K. Izawa T. Fuse T. Yamanouchi U. Kubo T. Shimatani Z. (2004 ). Ehd1, a B-type response regulator in rice, confers short-day promotion of flowering and controls FT-like gene expression independently of Hd1. Genes Dev. 18 926 –936 . 10.1101/gad.1189604 15078816 Gao H. Zheng X. M. Fei G. L. Chen J. Jin M. N. Ren Y. L. (2013 ). Ehd4 encodes a novel and Oryza-genus-specific regulator of photoperiodic flowering in rice. PLoS Genet. 9 :e1003281 10.1371/journal.pgen.1003281 Hori K. Ogiso-Tanaka E. Matsubara K. Yamanouchi U. Ebana K. Yano M. (2013 ). Hd16, a gene for casein kinase i, is involved in the control of rice flowering time by modulating the day-length response. Plant J. 76 36 –46 . 10.1111/tpj.12268 23789941 Huang X. Wei X. Sang T. Zhao Q. Feng Q. Zhao Y. (2010 ). Genome-wide association studies of 14 agronomic traits in rice landraces. Nat. Genet. 42 961 –967 . 10.1038/ng.695 20972439 Huang X. Zhao Y. Wei X. Li C. Wang A. Zhao Q. (2012 ). Genome-wide association study of flowering time and grain yield traits in a worldwide collection of rice germplasm. Nat. Genet. 44 32 –39 . 10.1038/ng.1018 22138690 Itoh H. Nonoue Y. Yano M. Izawa T. (2010 ). A pair of floral regulators sets critical day length for Hd3a florigen expression in rice. Nat. Genet. 42 635 –638 . 10.1038/ng.606 20543848 Khush G. S. (1997 ). “Origin, dispersal, cultivation and variation of rice ,” in Oryza: From Molecule to Plant , eds Sasaki T. Moore G. (Berlin : Springer ), 25 –34 . Lee S. Kim J. Han J. J. Han M. J. An G. (2004 ). Functional analyses of the flowering time gene OsMADS50, the putative suppressor of overexpression of co 1/AGAMOUS-like 20 (SOC1/AGL20) ortholog in rice. Plant J. 38 754 –764 . 10.1111/j.1365-313X.2004.02082.x 15144377 Li M. X. Yeung J. M. Cherny S. S. Sham P. C. (2012 ). Evaluating the effective numbers of independent tests and significant p-value thresholds in commercial genotyping arrays and public imputation reference datasets. Hum. Genet. 131 747 –756 . 10.1007/s00439-011-1118-2 22143225 Lippert C. Listgarten J. Liu Y. Kadie C. M. Davidson R. I. Heckerman D. (2011 ). FaST linear mixed models for genome-wide association studies. Nat. Methods 8 833 –835 . 10.1038/nmeth.1681 21892150 Liu T. Liu H. Zhang H. Xing Y. (2013 ). Validation and characterization of Ghd7. 1, a major quantitative trait locus with pleiotropic effects on spikelets per panicle, plant height, and heading date in rice (Oryza sativa L.). J. Integr. Plant Biol. 55 917 –927 . 10.1111/jipb.12070 23692054 Lu L. Yan W. Xue W. Shao D. Xing Y. (2012 ). Evolution and association analysis of Ghd7 in rice. PLoS ONE 7 :e34021 10.1371/journal.pone.0034021 Matsubara K. Ogiso-Tanaka E. Hori K. Ebana K. Ando T. Yano M. (2012 ). Natural variation in Hd17 a homolog of Arabidopsis ELF3 that is involved in rice photoperiodic flowering. Plant Cell Physiol. 53 709 –716 . 10.1093/pcp/pcs028 22399582 Matsubara K. Yamanouchi U. Nonoue Y. Sugimoto K. Wang Z. X. Minobe Y. (2011 ). Ehd3, encoding a plant homeodomain finger-containing protein, is a critical promoter of rice flowering. Plant J. 66 603 –612 . 10.1111/j.1365-313X.2011.04517.x 21284756 Matsubara K. Yamanouchi U. Wang Z. X. Minobe Y. Izawa T. Yano M. (2008 ). Ehd2, a rice ortholog of the maize INDETERMINATE1 gene, promotes flowering by up-regulating Ehd1. Plant Physiol. 148 1425 –1435 . 10.1104/pp.108.125542 18790997 Nakagawa M. Shimamoto K. Kyozuka J. (2002 ). Overexpression of RCN1 and RCN2, rice TERMINAL FLOWER 1/CENTRORADIALIS homologs, confers delay of phase transition and altered panicle morphology in rice. Plant J. 29 743 –750 . 10.1046/j.1365-313X.2002.01255.x 12148532 Ryu C. H. Lee S. CHO L. H. Kim S. L. LEE Y. S. Choi S. C. (2009 ). OsMADS50 and OsMADS56 function antagonistically in regulating long day (LD)-dependent flowering in rice. Plant Cell Environ. 32 1412 –1427 . 10.1111/j.1365-3040.2009.02008.x 19558411 Stephens M. Smith N. J. Donnelly P. (2001 ). A new statistical method for haplotype reconstruction from population data. Am. J. Hum. Genet. 68 978 –989 . 10.1086/319501 11254454 Takahashi Y. Shomura A. Sasaki T. Yano M. (2001 ). Hd6, a rice quantitative trait locus involved in photoperiod sensitivity, encodes the α subunit of protein kinase CK2. Proc. Natl. Acad. Sci. U.S.A. 98 7922 –7927 . 10.1073/pnas.111136798 11416158 Tamaki S. Matsuo S. Wong H. L. Yokoi S. Shimamoto K. (2007 ). Hd3a protein is a mobile flowering signal in rice. Science 316 1033 –1036 . 10.1126/science.1141753 17446351 Tsuji H. Taoka K.-I. Shimamoto K. (2011 ). Regulation of flowering in rice: two florigen genes, a complex gene network, and natural variation. Curr. Opin. Plant Biol. 14 45 –52 . 10.1016/j.pbi.2010.08.016 20864385 Vergara B. Tanaka A. Lilis R. Puranabhavung S. (1966 ). Relationship between growth duration and grain yield of rice plants. Soil Sci. Plant Nutr. 12 31 –39 . 10.1080/00380768.1966.10431180 Wang Q. Xie W. Xing H. Yan J. Meng X. Li X. (2015 ). Genetic architecture of natural variation in rice chlorophyll content revealed by a genome-wide association study. Mol. Plant 8 946 –957 . 10.1016/j.molp.2015.02.014 25747843 Wei X. Xu J. Guo H. Jiang L. Chen S. Yu C. (2010 ). DTH8 suppresses flowering in rice, influencing plant height and yield potential simultaneously. Plant Physiol. 153 1747 –1758 . 10.1104/pp.110.156943 20566706 Wu W. Zheng X. M. Lu G. Zhong Z. Gao H. Chen L. (2013 ). Association of functional nucleotide polymorphisms at DTH2 with the northward expansion of rice cultivation in Asia. Proc. Natl. Acad. Sci. U.S.A. 110 2775 –2780 . 10.1073/pnas.1213962110 23388640 Xue W. Xing Y. Weng X. Zhao Y. Tang W. Wang L. (2008 ). Natural variation in Ghd7 is an important regulator of heading date and yield potential in rice. Nat. Genet. 40 761 –767 . 10.1038/ng.143 18454147 Yano M. Katayose Y. Ashikari M. Yamanouchi U. Monna L. Fuse T. (2000 ). Hd1, a major photoperiod sensitivity quantitative trait locus in rice, is closely related to the Arabidopsis flowering time gene CONSTANS. Plant Cell 12 2473 –2483 . 10.2307/3871242 11148291 Zhang Y. -C. Yu Y. Wang C. -Y. Li Z. -Y. Liu Q. Xu J. (2013 ). Overexpression of microRNA OsmiR397 improves rice yield by increasing grain size and promoting panicle branching. Nat. Biotechnol. 31 848 –852 . 10.1038/nbt.2646 23873084 Zhao K. Tung C. -W. Eizenga G. C. Wright M. H. Ali M. L. Price A. H. (2011 ). Genome-wide association mapping reveals a rich genetic architecture of complex traits in Oryza sativa. Nat. commun. 2 467 10.1038/ncomms1467

PMC005xxxxxx/PMC5002402.txt

==== Front Front Public HealthFront Public HealthFront. Public HealthFrontiers in Public Health2296-2565Frontiers Media S.A. 10.3389/fpubh.2016.00174Public HealthOriginal ResearchA Feasibility Study for an Integrated Approach to Fall Prevention in Community Care: Stay Up and Active in Orange County Lindgren Spencer W. 1Kwaschyn Katie 2Roberts Ellen 2Busby-Whitehead Jan 3Evarts Lori A. 4Shubert Tiffany 5*1Orange County Emergency Services, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA2University of North Carolina at Chapel Hill, Chapel Hill, NC, USA3Division of Geriatric Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA4Public Health Leadership Program, Gillings School of Global Public Health, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA5Center for Aging and Health, University of North Carolina at Chapel Hill, Chapel Hill, NC, USAEdited by: Matthew Lee Smith, University of Georgia, USA Reviewed by: William Augustine Toscano, University of Minnesota, USA; Janice Elisabeth Frates, California State University Long Beach, USA *Correspondence: Tiffany Shubert, tiffany_shubert@med.unc.eduSpecialty section: This article was submitted to Public Health Education and Promotion, a section of the journal Frontiers in Public Health 29 8 2016 2016 4 17417 6 2016 08 8 2016 Copyright © 2016 Lindgren, Kwaschyn, Roberts, Busby-Whitehead, Evarts and Shubert.2016Lindgren, Kwaschyn, Roberts, Busby-Whitehead, Evarts and ShubertThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.Introduction Falls among persons over 60 present significant risks for serious injury or debility. Falls place burdens on Emergency Medical Services (EMS), hospitals, and the adults themselves. Recognizing a need to provide interventions to minimize risk, Orange County Emergency Services (OCES), the Orange County Department on Aging (OCDoA), and the University of North Carolina at Chapel Hill (UNC) partnered to create the Stay Up and Active Program (SUAA). The purpose of this study was to determine if SUAA was a feasible program to implement in the community. Methods A streamlined workflow algorithm between the OCES and OCDoA was created and employed to provide falls risk assessment and necessary services. Qualitative techniques were used to assess the need for such a program and its potential impact. A subset of individuals was interviewed 3 months after the intervention to assess the impact of the intervention on their fall risk. Formal stakeholder interviews were not conducted, but anecdotal information from EMS providers was obtained and reported. Results In the first 7 months, 478 instances of individuals who called OCES screened positive for falls risk. Of the 478 positive screenings, 55 individuals were identified as having received more than one positive fall screen due to multiple calls. The maximum number of positive screenings by one individual was 14. More women (61.3%) than men screened positive for fall risk. Individuals 88 years of age (6.9%) represented the highest number of individuals with positive screens. Nineteen (4.0%) people who called OCES and received the intervention completed a 3-month follow-up survey. Of the 19, 86% (n = 16) reported no recurrent fall. Conclusion The number of individuals who screened positive supports the need for early identification and intervention through SUAA. This program identified several challenges connecting older adults with services already available to keep them independent, which provided insight to all stakeholders regarding factors that inhibit the program’s success. The program evaluation should continue to provide suggestions for improvement and ensure sustainability. first respondersfall prevention programaging and longetivityemergency medical servicesSTEADI toolkitHealth Resources and Services Administration10.13039/100000102UB4HP19053 ==== Body Introduction Falls in older adults comprise a significant portion of health-care expenditures and resource use in the United States. One of every three older adults falls annually resulting in a total of 12 million falls (1). In 2013, these falls represented approximately $34 billion in direct medical costs and led to 21,700 deaths among older adults (2). North Carolina is ranked fifth in the United States for the greatest number of older adults. It is projected that by 2030, there will be a 32% increase in the state’s population aged 65 years and older (3). Given these demographics, the state is particularly concerned by this public health issue, keeping in mind that the costs of falls and the burden on the health-care system are already substantial. In 2012, there were nearly 195,000 Emergency Department (ED) visits as a result of unintentional falls (4). Of these ED visits, 900 resulted in deaths, constituting a 74.5% increase in deaths from falls between 1999 and 2012 (4). Orange County is one of the 100 counties in North Carolina. The county measures approximately 398 square miles and is home to 141,354 citizens (5). The county is home to several towns including Chapel Hill, Carrboro, and Hillsborough. The University of North Carolina at Chapel Hill (UNC), along with the UNC Hospitals System, are located in the southern part of the county in a more urban setting, while the northern part of the county is mostly rural with a significantly lower population density. The area has gained popularity with retirees and is the home of five large retirement communities as well as several assisted living and skilled nursing facilities. Persons age 65 years or older make up 11.2% of the population and it is projected that by 2030 18% of the population will be 65 years or older (3, 5). Females comprise 52.2% of the population, Caucasians account for 76.8% of the population, African-American for 12.2%, Hispanic for 8.4%, Asian for 7.7%, and Native American for 0.6% (5). The Emergency Medical Services Division (EMS) of the Orange County Emergency Services (OCES) is the sole provider of Advanced Life Support (ALS) services in the county. EMS consists of 75 full-time and 20 part-time employees, and staffs 5–9 ambulances any given day. OCES began tracking EMS calls classified as falls-related in 2010. Between 2010 and 2013, the EMS Division averaged 10,384 calls per year (6). Of these calls, 0.9% were lift assist EMS calls (where a person needs help transferring from a bed to a chair or similar situation and has not actually fallen), and 10.7% were falls-related calls, for a combined average of 11.6% of all calls being falls-related over the 4-year period; these data are summarized in Table 1 (6). Table 1 Baseline falls data for Orange County emergency medical services (EMS), 2010–2013 (6). Year Total EMS calls Lifting assistance Falls Combined lifting and falls Percent lifting and falls 2010 9,585 159 984 1,143 11.9 2011 10,333 101 1,117 1,218 11.8 2012 10,636 64 1,165 1,229 11.6 2013 10,983 63 1,182 1,245 11.3 During the period of 2010–2013, there was no tracking to identify repeat fall victims that utilized an EMS ambulance for transport or non-transport purposes. The Orange County Department on Aging (OCDoA) encounters over 190,000 participants per year at its hosted events (7). The OCDoA has expertise and resources to help older adults manage their fall risk and achieve the goals of aging in community. The Aging Transitions Unit, a group within OCDoA, employs five full-time and several part-time employees to provide in-home assessments, caregiver referral, low-cost support services, and other age-related services to citizens (7). The Aging Transitions Unit spends an average of 150 h per month providing information and case assistance to citizens (7). Emergency Medical Services has frequent contact with older adults who would benefit from OCDoA services to minimize their risk of falling. The opportunity to leverage the “first responder” relationship and connect older, at-risk adults with the resources in the community was the inspiration for the Stay Up and Active Program (SUAA). SUAA was designed to be a fall risk identification and management program implemented by EMS to connect at-risk older adults with the services they need. The intent of connecting these adults with services is to reduce the number of falls by older adults in their homes. This program represents the first time EMS formally collaborated with the Department on Aging to meet a need in the community. Initial discussions between the OCDoA and EMS supported that an EMS-centric model would be optimal for a community falls prevention program. As EMS providers are frequently the first caregivers in any fall, OCES was well positioned to link the at-risk population with the services and resources provided by the OCDoA. EMS providers have the opportunity to obtain accurate and complete histories from the older adults and possible bystanders on scene, and can assess the older adult’s safety in their environment. EMS and OCDoA agreed to initiate SUAA with EMS as the first point of contact with potential at-risk individuals. EMS would then schedule an in-home visit to further assess the older adult and communicate with OCDoA to connect the older adults with community resources. The purpose of this study was to determine if SUAA was a feasible program to implement in the community. Specifically, it was necessary to know if the perceived need for the program was accurate, if the workflow developed to implement SUAA for EMS staff was efficient and effective, if older adults who called EMS for a falls-related issue would be receptive to a second home visit, and if the system designed to facilitate communication between OCES and OCDoA achieved the goals of the project. The information from this study will help inform future steps to this collaborative project to address the problem of falls in the community. Materials and Methods Development of Workflow The OCDoA provides services to all county residents aged 60 years and older. Therefore, the SUAA program included any adult who is 60 years of age or older in Orange County who received EMS support resulting from a call for service. An algorithm and workflow were developed by both organizations to identify the level of risk and appropriate intervention (Figure 1). Figure 1 WebEOC flow chart. All adults over the age of 60 years, who called EMS, were screened for fall risk. During the course of an EMS intervention, providers would ask the following screening questions from the American Geriatrics Society (AGS) Clinical Practice Guidelines questions: • Are you worried that you are going to fall? • Have you fallen in the past year? • Are you unsteady when walking or standing? (8) A positive screen was based on a “Yes” response to any of the questions. Those that screened positive were asked if they would like to receive a follow-up phone call and additional home safety services from EMS. Those who agreed signed a form allowing EMS to access their name and phone number for further follow-up. Any EMS provider, regardless of their certification level, was able to conduct a fall risk screen. This screening was designed to supplement the standard patient assessment, and was incorporated to be as streamlined as possible for field EMS staff. The older adults who agreed to follow-up were entered into the EMS WebEOC (online emergency incident management technology) database to track their status. The purpose of WebEOC was to notify the SUAA team that an older adult screened positive for the program, to track their status and to communicate outcomes between agencies. Seventy-two hours following the initial EMS service call, a follow-up telephone call by EMS was initiated to schedule a home visit. If no contact was made after three telephone call attempts, the patient record was closed. If contact was made, a home visit from EMS personnel would be scheduled. The home visit consisted of a translation of the Centers for Disease Control’s STEADI (Stopping Elderly Accidents, Deaths, and Injuries) toolkit, which is an evidence-based fall risk management algorithm for clinicians (9). The STEADI algorithm includes assessment of the following risk factors: falls history, fear of falling, polypharmacy, leg weakness, balance impairments, low vision, cognitive impairment, depression as well as environmental factors. At the scheduled EMS home visit, the patient would be asked background information including current medications, medication history, and the current status of their health. Additionally, they would be screened with validated tools for cognitive impairment (Mini-Cog Assessment), depression (PHQ-2), elder abuse, and vision impairment (9). The older adult completed a Timed Up and Go Test, the 30-S Chair Stand, and 4-Stage Balance Test Full Tandem Stance (9). Finally, an assessment of their current living conditions and any observed safety concerns or risk factors were discussed with the patient. The results of the EMS home visit were then entered into WebEOC. Subsequent to a home visit from EMS, and with approval from the older adult, a notification was sent to OCDoA from the WebEOC database. The goal was for OCDoA to make an assessment and connect the older adult with the appropriate resources in the community. At the OCDoA follow-up, appropriate referrals for occupational therapy, physical therapy, counseling, caregiver support group, in-home health-care services, and others were made. Communication In an effort to streamline communication and share findings, the assessments and referrals were recorded in WebEOC by both EMS and by OCDoA and used for participant monitoring. Following the completion of a WebEOC ticket, the initial EMS crews were notified of the outcome of OCES and OCDoA follow-up with the older adults. Prior to implementation, SUAA was reviewed by the UNC Office of Human Research Ethics Institutional Review Board as Study 13-2942 and was granted exempt status from further review as the submission was considered a quality improvement program and did not constitute human subjects research under 45 CFR 46.102 (d or f) and 21 CFR 56.102(c)(e)(l). Additionally, neither special funding was allocated nor was grant funding obtained to implement this program. All resources were obtained from preexisting sources within county offices. Any materials given to adult participants were free of charge and donated by relevant organizations. Evaluation The SUAA program was evaluated by a team based at the UNC Chapel Hill Center for Aging and Health to determine the feasibility of implementing the project. The first part of the evaluation determined the SUAA program met a need in the community. Implementation based on the established workflow and related IRB status enabled the data collection and analysis effort. The second part of the SUAA program evaluation consisted of interviews with a subset of individuals who received the home visit to assess their response to the program. The evaluation work was funded by a grant received by the University of North Carolina. Formal interviews of EMS providers were not conducted; however, anecdotal information obtained by providers was obtained when reviewing cases with these providers. The primary method of obtaining information about patients and their condition was by reviewing the patient care reports submitted by EMS providers after the initial 9-1-1 call. These reports provided valuable information about the patient’s health and current social environment. Results Between September 1, 2013 and March 31, 2014, there were a total of 704 EMS calls for a fall and 37 EMS calls for “Lift Assistance” (6). There were a total of 478 instances of a positive screen using the Falls Risk Assessment criteria. Of these, there were a total of 55 unique individuals who had at least one repeat positive screen as a result of a 9-1-1 call. The range of repeat screenings by an individual was between two and fourteen. There were 32 individuals who experienced a total of two positive screenings and 303 individuals with only one positive screening. The available data are presented in Table 2. The age range of patients was 60–99 years, with the largest group (6.9%) being adults aged 88 years. Females made up 61.3% of the positive screenings. Positive screenings are plotted by age of the patient in Figure 2. The age demographics of Orange County EMS patients decreases significantly after age 90 years, accounting for the drop in positive fall screenings. Of the patients who screened positive for fall risk, 316 instances (66.1%) were transported to an ED, and the remaining 162 instances (33.9%) were non-transport either by Refusal Against Medical Advice or Referral to a Physician within 4 or 24 h. Table 2 Positive screening rates for September 1, 2013 through March 31, 2014 (6). Total number of positive screenings Number of individuals 1 303 2 32 3 10 4 5 5 2 6 1 7 2 8 1 9 1 10 0 11 0 12 0 13 0 14 1 Figure 2 Positive fall screenings by patient age, Orange County, NC, USA, September 1, 2013–March 31, 2014. Of patients who screened positive, accurate phone numbers were only recorded in 31% (148 instances) of the patient care reports. Of the 148 instances of positive screening for falls risk and accurate phone number collected, 54 participants agreed to a home visit by EMS. Of those patients who received a home visit by EMS, 20 participants agreed to a follow-up visit from the OCDoA. Nineteen of the participants who received follow-up from the OCDoA agreed to further follow-up from UNC in the form of a 3-month follow-up survey conducted by phone interview. Of the 19 participants who completed the 3-month follow-up survey, 86% did not report a recurrent fall at 3-month follow-up. A total of 74% were very satisfied, and 26% were satisfied with the home visit from EMS. When asked about the value of the program, 5% found it not helpful, 16% found it somewhat helpful, 21% found it helpful, and 42% very helpful; 10 respondents answered other. As a result of the home visit, 16% felt somewhat confident, 16% felt confident, and 32% felt very confident that he or she could take actions to reduce risk of falling. Ten participants remarked other and commented, “They have already done everything they could to prevent falls” and, “The visit helped to heighten their awareness that they had to do something to prevent falls.” In the survey, 95% would recommend the EMS home visit program to a friend who may need help to stay independent in their home and 5% responded maybe. Discussion Older adult falls, and older adults who fall more than once, are a public health problem in Orange County, NC, USA. With more than 11% of all EMS calls being fall-related, coupled with a rapidly expanding aging population within the community, there was a definite need for a local structured falls prevention program. An efficient workflow incorporating evidence-based assessments was constructed and adopted by OCES and OCDoA. The WebEOC system allowed for transparent and timely exchanges of information between providers. The program recruited only 54 older adults in a 7-month pilot period, a lower than expected number. Of the 19 participants who completed the 3-month follow-up survey, 86% did not report a recurrent fall and the overall satisfaction rate was positive. During the 7-month trial period, there were a total of 704 EMS calls for a fall, and 37 EMS calls for “Lift Assistance.” Based on this information alone, the EMS unit hour utilization and ED bed time use expected as a result of these calls causes a significant burden to health-care resources. Further studies are warranted to investigate if SUAA has any impact on decreasing the number of annual falls related EMS calls. First responders should continue to be utilized as they offer a unique and innovative way to access older, at-risk adults who would otherwise be left underserved by their community resources. There were a total of 741 falls related EMS calls during the study period, but only 478 instances of positive screenings. As the falls risk assessment could be performed on any patient aged 60 years or greater no matter what the nature of the call (Fall, Lift Assist, Chest Pain, Dyspnea, etc.), it was expected that at least as many positive screenings would be recorded. Since this is not the case, further investigation is needed to determine and quantify whether or not all fall victims were screened or if they screened negative. If they are simply were not being screened, then further training and emphasis will need to be placed on the necessity for asking the three falls risk assessment questions with field EMS staff. If the patients are screening negative, then evaluation of EMS recording and other potential areas of outreach need to be explored with this program. In examining the demographics of the patients who screened positive, more women screened positive than men, consistent with the national data that shows women over 60 falls more frequently than men (1). The Orange County data do show, however, that there is no correlation between age and a positive falls screening. The most common age in Orange County for falls risk was 88 years, but otherwise, there was no ability to predict a person’s risk based on age. EMS providers in Orange County anecdotally believe that more falls calls occur at assisted living facilities than at private residences. Based on the screenings performed by EMS, these data revealed that there were, in fact, more people at risk for falling in private homes than in assisted living facilities. There were several barriers and limitations discovered during implementation of this program. Barriers fell into two categories: system-based change and older adults. Initially, information sharing to track participants who agreed to follow-up was difficult. In response, the WebEOC boards were reviewed, modified, and republished to ensure ease of access and use for all agencies. A second barrier encountered was the poor phone number collection by field EMS staff. Without an accurate phone number, patients could not be contacted for follow-up which was reflected by low participant rates. There was significant difficulty getting participants’ agreement to a home visit by EMS based on the EMS telephone contact effort. Several factors that contributed to this were failure to self-identify as at-risk, currently receiving care at the time of phone call (including hospitalization), unable to contact, and no interest in speaking with a representative from EMS. It was also difficult to find one time frame (e.g., 1-week post initial EMS call) that could be applied to all patients to call to schedule a follow-up. To address these barriers, patients are now to be asked at the time of the field assessment if they would like a follow-up and contact information will be obtained for both the patient and their primary caregiver (if possible). This process amendment will hopefully reduce the difficulty in trying to explain the program over the phone and make it easier to schedule a follow-up visit. The 3-month follow-up survey revealed areas of success and room for improvement. One participant remarked that the, “EMS made suggestions to get the wheelchair through the doorway” while a family member of another participant commented, “The older adult won’t comply with the recommendations.” Some of the additional comments included: (1) “Daughter was very frustrated – she has been spending considerable time caring for her mother and needs help. The daughter has minimal transportation and hasn’t been able to go to work due to caring for her mother. Her mother had been taken to the hospital in the morning with a mini-stroke. The daughter repeatedly said that her mother needs a wheelchair,” (2) “Talked with son-in-law of patient. He was present at the EMS home visit and very enthusiastically supports it.” The idea of utilizing EMS to provide population health services is not novel; programs have been established for the EMS staff to augment immunization and fall prevention services provided in the rural areas of upstate New York (10). SUAA successfully expands this model of care beyond a rural setting. In addition, SUAA partners with the Department of Aging in order to bolster program recruitment and to offer evaluation and meaningful interventions in the care of falls risk patients. The follow-up rate for the study in upstate New York was 61% with the survey completed 14 days after interview; follow-up calls were attempted for up to 4 weeks to contact individuals (10). The 3-month wait time for SUAA allowed adequate time for all planned interventions to be performed prior to assessing the patient outcomes. Still, the SUAA respondent rates are lower than NY study, and further studies looking into barriers to communication may be warranted. The study shows SUAA addresses a need within the community, but adjustments are needed to improve processes to ensure sustainability. Conclusion Elderly, frail patients with multi-morbidity require greater time and resources to maintain independent living. In an effort to intercept the unique health-care concerns of a rapidly expanding aging population, the SUAA offers a potential solution by targeting at-risk individuals and providing assessment and resources. The goal is to not only decrease the number of EMS calls for falls but also the overall community morbidity as a result of preventable falls in older adults. This program represents a tremendous effort put forth by UNC, OCDoA, and OCES. The historical data and results from the pilot phase of Stay Up and Active demonstrate the need in Orange County for more than simple emergency response to injury and illness. Orange County EMS is in a prime position to provide the falls assessment questions as an integrated part of their services, and must continue implementation of this program as well as address the barriers identified in this report. Furthermore, SUAA represents a national trend for EMS systems to address community needs of their patients and begin to shift resources toward population health as a means to alleviate the burdens they face. However, with a large aging population, both local and national attention should be given to help individuals safely age in place as a way to help offset future health-care costs. Author Contributions SL is the primary author for this work. LE served as content and conceptual review for the work throughout the process. TS was the program implementation lead. KK made iterative changes and periodic updates to the work throughout the process. ER and JB-W provided subject-matter expertise to the research and evaluation oversight. Conflict of Interest Statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Funding This work was supported by the Bureau of Health Professions (BHPr), Health Resources and Services Administration (HRSA), Department of Health and Human Services (DHHS) under grant #UB4HP19053, Carolina Geriatric Education Center. This information, content, and conclusions are those of the author and should not be construed as the official position or policy of, nor should any endorsements be inferred by the BHPr, HRSA, DHHS, or the U.S. Government. Abbreviations AGS, The American Geriatrics Society; ALS, advanced life support; CDC, Centers for Disease Control and Prevention; C-Spine, cervical spine; ED, emergency department; EMS, emergency medical services; EMT, emergency medical technician; OCDoA, Orange County Department on Aging; OCES, Orange County Emergency Services; STEADI, stopping elderly accidents, deaths, and injuries; SUAA, stay up and active; TBI, traumatic brain injury; UNC, University of North Carolina at Chapel Hill. ==== Refs References 1 National Council on Aging . Debunking the Myths of Older Adult Falls . (2016 ). Available from: http://www.ncoa.org/improve-health/falls-prevention/debunking-the-myths-of-older.html 2 Centers for Disease Control and Prevention . Falls among Older Adults: An Overview . (2016 ). Available from: http://www.cdc.gov/homeandrecreationalsafety/falls/adultfalls.html 3 UNC Carolina Population Center Carolina Demography . Population Growth & Population Aging in North Carolina Counties . (2016 ). Available from: http://demography.cpc.unc.edu/2013/10/14/population-growth-population-aging-in-north-carolina-counties/ 4 Austin A Unintentional fall injury data . Data Presented at: North Carolina Falls Prevention Summit . Raleigh, NC (2015 ). 5 United States Census Bureau . Quick Facts – Orange County, North Carolina . (2016 ). Available from: http://www.census.gov/quickfacts/table/PST045215/37135,00 6 Orange County 9-1-1 Center . Data from: Orange County 9-1-1 Data, 2010–2013 . (Data File and Code Book). Hillsborough, NC (2016 ). 7 Orange County Department on Aging . FY 2015–16 Annual Operating Budget and Capital Investment Plan . Hillsborough, NC (2016 ). Available from: http://www.orangecountync.gov/2015_16_Commissioner_Approved_Operating_Budget_and_Capital_Investment_Plan.pdf 8 American Geriatrics Society . Prevention of Falls in Older Persons . (2016 ). Available from: http://www.medcats.com/FALLS/frameset.htm 9 Centers for Disease Control and Prevention . STEADI Tool Kit for Health Care Providers . (2016 ). Available from: http://www.cdc.gov/steadi/materials.html 10 Shah M Clarkson L Lerner E Fairbanks R Mccann R Schneider S . An emergency medical services program to promote the health of older adults . J Am Geriatr Soc (2006 ) 54 (6 ):956 –62 .10.1111/j.1532-5415.2006.00736.x 16776792

PMC005xxxxxx/PMC5002403.txt

==== Front Front Hum NeurosciFront Hum NeurosciFront. Hum. Neurosci.Frontiers in Human Neuroscience1662-5161Frontiers Media S.A. 10.3389/fnhum.2016.00436NeuroscienceOriginal ResearchDevelopment of Proprioceptive Acuity in Typically Developing Children: Normative Data on Forearm Position Sense Holst-Wolf Jessica M. *Yeh I-Ling Konczak Jürgen Human Sensorimotor Control Laboratory, School of Kinesiology, University of MinnesotaMinneapolis, MN, USAEdited by: Rachael D. Seidler, University of Michigan, USA Reviewed by: Douglas Owen Cheyne, Hospital for Sick Children, Canada; Alessandro Farne, French Institute of Health and Medical Research, France *Correspondence: Jessica M. Holst-Wolf hols0078@umn.edu29 8 2016 2016 10 43604 5 2016 15 8 2016 Copyright © 2016 Holst-Wolf, Yeh and Konczak.2016Holst-Wolf, Yeh and KonczakThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution and reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.This study mapped the development of proprioception in healthy, typically developing children by objectively measuring forearm position sense acuity. We assessed position sense acuity in a cross-sectional sample of 308 children (5–17 years old; M/F = 127/181) and a reference group of 26 healthy adults (18–25 years old; M/F = 12/14) using a body-scalable bimanual manipulandum that allowed forearm flexion/extension in the horizontal plane. The non-dominant forearm was passively displaced to one of three target positions. Then participants actively matched the target limb position with their dominant forearm. Each of three positions was matched five times. Position error (PE), calculated as the mean difference between the angular positions of the matching and reference arms, measured position sense bias or systematic error. The respective standard deviation of the differences between the match and reference arm angular positions (SDPdiff) indicated position sense precision or random error. The main results are as follows: First, systematic error, measured by PE, did not change significantly from early childhood to late adolescence (Median PE at 90° target: −2.85° in early childhood; −2.28° in adolescence; and 1.30° in adults). Second, response variability as measured by SDPdiff significantly decreased with age (Median SDPdiff at 90° target: 9.66° in early childhood; 5.30° in late adolescence; and 3.97° in adults). The data of this large cross-sectional sample of children document that proprioceptive development in typically developing children is characterized as an age-related improvement in precision, not as a development or change in bias. In other words, it is the reliability of the perceptual response that improves between early childhood and adulthood. This study provides normative data against which position sense acuity in pediatric patient populations can be compared. The underlying neurophysiological processes that could explain the observed proprioceptive development include changes in the tuning of muscle spindles at the spinal level, the maturation of supraspinal somatosensory pathways and the development of interhemispheric callosal connections responsible for the transfer of somatosensory information. sensorimotorproprioceptionposition sensehumandevelopment ==== Body Introduction Proprioception refers to the sense of relative position and movement of the limbs and body (Konczak et al., 2009). Proprioceptive information is provided through mechanoreceptors embedded in the joints, muscles, tendons and skin. It has long been recognized that intact proprioception is essential for the control of muscle tone and voluntary movement. It is further known that pediatric conditions such as cerebral palsy, autism or developmental coordination disorder are associated with somatosensory or proprioceptive deficits that negatively affect movement control (Coleman et al., 2001; Kaufman and Schilling, 2007; Goble et al., 2009; Wang et al., 2009; Zwicker et al., 2013; Li et al., 2015). In contrast to other senses such as vision or audition, the development of proprioception in typically developing children has never been fully mapped. Previous research has established that kinesthetic sensitivity and acuity improves in middle childhood and continues to approach adult levels during late adolescence (Laszlo and Bairstow, 1980; Bairstow and Laszlo, 1981; Elliott et al., 1988; Visser and Geuze, 2000; Goble et al., 2005; Pickett and Konczak, 2009). However, no consistent assessment protocol has been established. The available empirical results were either based on small samples, had methodological shortcomings, or were restricted to limited age groups that did not provide a comprehensive understanding of proprioceptive development across childhood and adolescence. Moreover, most reported results are not comparable between studies because of differences in protocol and test equipment. One reason for the current lack of reliable normative data against which the proprioceptive status of pediatric populations can be compared, lies in the fact that proprioceptive function is not easily examined. Clinical rating scales are coarse and lack sufficient sensitivity, which allow only for detection of severe deficits in proprioception. Objective testing of proprioceptive function may require specialized equipment that is expensive and typically not available in clinical settings. In addition, testing takes too long or demands high levels of attention, which becomes problematic especially for younger children. Moreover, available kinesthetic acuity tests have failed to consistently differentiate pediatric populations from healthy children (Lord and Hulme, 1987; Hoare and Larkin, 1991; Piek and Coleman-Carman, 1995). Early work to characterize the development of proprioception by Laszlo and Bairstow (1980) produced the Kinesthetic Sensitivity Test. However, in this test, approximately 25% of the 5- and 6-year old participants never responded correctly for more than half of the time (Bairstow and Laszlo, 1981). This is equivalent to random guessing and illustrates the difficulty in designing a reliable and valid test of proprioceptive function useful for clinical assessment. Other available tests such as the Movement Assessment Battery for Children 2nd edition (Kirby and Sugden, 2007; Li et al., 2015) focus primarily on motor but not on proprioceptive deficits. When designing a test of proprioceptive assessment for children, one needs to consider the following: first, which proprioceptive modality shall be examined: position or motion sense? Position sense testing is technically simpler, because it does not require an apparatus to control the velocity of limb motion. Second, will sensitivity or acuity be tested? (We here refer to sensitivity as the ability to detect a proprioceptive stimulus and acuity as the ability to discriminate between two detectable stimuli). Third, is the test duration short and simple enough to assure that even young children can complete it without being mentally and cognitively challenged? Fourth, does the test equipment conform to the large differences in anthropometrics that are observed during childhood? Here, we attempted to address the above challenges. Based on our previous work (Elangovan et al., 2014) and the work of others (Goble et al., 2005; Goble, 2010), we opted to employ a contralateral arm matching task to measure position sense acuity in children. We designed a bimanual manipulandum that can be body-scaled and is able to conform to anthropometrics of children, adolescents, and adults, while providing highly precise objective measures of forearm position (and velocity). We selected a bimanual, contralateral matching task, because it removes any confounds related to memorizing a position or having to generate an internal representation of the limb position (Goble et al., 2005; Goble, 2010; Elangovan et al., 2014). The method generates objective measures of acuity in a relatively short time span and is thus ideal for assessing proprioceptive function in children. This study mapped the development of proprioceptive acuity in a large, cross-sectional sample of typically developing children with the aim to obtain a normative data set against which the proprioceptive status of pediatric populations can be compared. When envisioning the developmental time course of proprioceptive acuity, one needs to consider that acuity has two aspects: bias and precision (Figure 1A). Here bias indicates how close a sensed limb position corresponds to the true physical position of the forearm (I.S.O., 1994). For a true response, there is no systematic error or bias. Precision represents the random error or the agreement between independent repeated responses (I.S.O., 1994). This implies that during development either an initially high systematic error is reduced, or precision increases with age, or proprioceptive development is characterized by a combination of both. Figure 1 (A) Bias and precision are the two components of acuity. Shown is a hypothetical population or individual probability density function for limb position matching responses the difference between the match and reference arm angular positions (degrees; Pdiff). The reference position is the true position. Accuracy requires minimal or no bias (low position error (PE)) and high precision (low SDPdiff). (B) Experimental setup with a participant seated in front of the bimanual manipulandum. Here the experimenter passively moved the non-dominant, right arm and the participant actively matched the position with the dominant left arm. The white lines depict the reference, or start position of 30° for the left arm as well as the three target positions. (C) Interchangeable lever arms accommodated a wide range of forearm lengths from children to adults fitting forearms lengths between 15 and 35 cm. (D) The left lever arm was adjustable to accommodate differences in shoulder width. Materials and Methods Participants Twenty-six healthy adults (M age: 20 years 11 months (SD: 1 year 8 months); 12 ♂, 14 ♀) and 308 children participated (age range: 5 years 4 months to 17 years 10 months; 127 ♂, 181 ♀). Recruitment and data collection occurred at a local primary school, on the University of Minnesota campus, and at the Minnesota State Fair. Appropriate parental consent and child assent was obtained before data collection. The study was approved by the University of Minnesota Internal Review Board. All subjects completed a modified Edinburgh handedness inventory to determine the dominant upper limb. Exclusion criteria for participation were the documented presence of central or peripheral nervous system disease, current injury to the upper limbs and/or implanted medical devices in the arms which may impair sensorimotor function. Apparatus A bimanual manipulandum with one degree of freedom in the horizontal plane was used to perform the bilateral arm position matching task performed in this study (Figure 1). Two US Digital H6 optical encoders (2500 quadrature count/revolution; spatial resolution: 0.14°), housed at the rotating point of the manipulandum lever arms, recorded the angular position of each arm at a sampling frequency of 43 Hz. Procedure Participants were seated, and they placed each arm on the bimanual levers. Chair height, lever arm length and handle placement were adjusted to the anthropometrics of each participant such that the approximate joint axis of the elbows aligned directly with encoder shaft axis. Distance between the two levers was adjusted such that the participants’ elbows were a comfortable distance apart (45°–85° of shoulder abduction). At the start of each trial and between target positions participants rested each arm at a start position set at 30° from the frontal plane (see Figure 1B). Participants wore vision occluding glasses during all trials and could not see their arms. The non-dominant arm was passively moved in the horizontal plane away from the body at a consistent speed of 20°–30°/s to one of three target positions, 40°, 60° or 90° from the frontal plane of the participant (see Figure 1B). We tested three different positions because displacing a limb to a higher amplitude is associated with a greater error (Goble et al., 2005) and we here sought to account for possible developmental differences of position sensing across the forearm workspace. Once the target position was reached, the participant moved the dominant arm to match the position of the other arm. Participants had as much time as desired to match the position. Participants verbally indicated when the matching position was achieved and held this position. Positions of each arm were then recorded for 1.6 s (collecting 70 samples of current limb position). Then the researcher moved the non-dominant hand back to the start position and gave a verbal cue for the participant to actively move the dominant arm back to the start position. Target positions were presented in pseudo random order such that each position was repeated 5 times for a total of 15 trials. One or two practice trials preceded data collection to familiarize the participants to the task and ensure that the instructions of the task were understood. If the researcher or child visibly moved during the 1.6 s recording, for example, if an arm was brought back to the start position early, the trial was repeated. Total testing time was approximately 10 min. Measurements For each trial the physical position of each arm was determined by computing the average position over the recorded 70 samples (i.e., while holding the position). Subsequently, the average position of the matching arm was subtracted from the average position of the reference arm (Pdiff). Finally, for each participant, the position error (PE) was computed as the mean of Pdiff across the five trials for each target position (40°, 60° and 90°). Similarly, the corresponding standard deviation was computed as the standard deviation (SDPdiff) of Pdiff across the five trials for each participant and target position. PE indicates a measure of bias or systematic error, while SDPdiff indicates the response precision across trials. Each participant’s chronological age at the date of data collection was calculated in years and months. Months were converted to base 10 (i.e., a year was subdivided into 10 sections). Results To characterize the breadth of response behavior across the development, Figure 2 shows the individual responses (Pdiff) of all 334 participants for each target position. Across the cross-sectional age sample, Pdiff was distributed about zero degrees for each target position, meaning that there was no systematic age-related trend to undershoot (more flexed than the reference position) or overshoot (more extended than the reference position). Figure 2 Differences between arm positions (Pdiff) for each trial across age for each target positions. A perfect forearm position match would have a Pdiff value of zero, a positive value indicates overshooting and a negative value indicates undershooting. Note that the data distribute about zero across age for each target. We further characterized the development of position sense acuity by investigating both bias and precision. To determine systematic changes in bias during development, a repeated measures analysis of variance (ANOVA) of PE was performed (chronological age × target position × gender × handedness). No significant main effects of PE for age, gender and handedness were found (p’s > 0.05). A significant interaction between age and target position was found (p = 0.048). However, linear regression procedures on PE by chronological age for the 40° and 60° target position did not reach significance (p’s > 0.05). The regression fit for the 90° target position reached significance (p = 0.04). However, the adjusted coefficient of determination for this linear model was low indicating that only age explained 1% of the variability in PE for the 90° target position (R2adjusted = 0.01). To determine developmental change in position sense precision, the corresponding SDPdiff data were analyzed. The distribution of the SDPdiff data was significantly different from normal which was corrected with a log10 transformation. A full model repeated measures ANOVA of the log transformed SDPdiff data (chronological age × target position × gender × handedness) was then performed. No significant main effects and interactions of gender and handedness were found (p’s > 0.05). The reduced model (chronological age × target position) revealed significant main effects of both amplitude (p < 0.001) and chronological age (p < 0.001), but no significant interaction. That is, the precision component of acuity changed with age and target position. Non-linear quantile regression procedures were conducted to estimate the SDPdiff percentiles across age for each target position. Figure 3 maps this age-related change in SDPdiff separately, for each target amplitude, illustrating that for each target SDPdiff decreased with increasing age. Figure 3 Developmental changes in precision for each target position. Each data point represents a participant’s SDPdiff. The quantile regression estimates for the 5th, 25th, 50th, 75th and 95th percentile are shown. Note that SDPdiff decreased with age, indicating an age-related improvement in precision. The median fit for each target position is: SDPdiff,40 = 8.43 *age−0.61, SDPdiff,60 = 19.83 *age−0.60, SDPdiff,90 = 32.18 *age−0.70. Grouping participants into developmentally appropriate classes (Payne and Isaacs, 2008) underlines the assessment that there are minimal age-related differences in PE (see Figure 4). For example, median PE for the 90° target ranged between −4.8° and 1.3° with no age-related, systematic change in PE discernible. To appreciate the developmental trend toward increased precision, consider that at 5–6 years of age (early childhood) the 5% to 95% range of SDPdiff for the 90° target was 2.4°–23.0°, compared to 1.2°–10.6° in adolescence and dropped to 1.9°–7.7° in adulthood. Figure 4 Development of bias and precision across developmental periods. Shown are boxplots for PE (left) and SDPdiff (right). The box lines indicate the 25th, 50th and 75th percentiles and the whiskers extend to the 5th and 95th percentile, outliers have been omitted for clarity. There is no trend of the median PE across development indicating no change in bias across age. However, the median SDPdiff for each age period decreases with increasing age indicating an improvement in precision during development. Early childhood (5–6 years old, n = 25), Middle childhood (7–8 years old, n = 62), Late Childhood (9–11 years old, n = 97), Adolescence (12–17 years old, n = 124), Adults (18–25 years old, n = 26). To investigate, if the age-related decrease in SDPdiff is a result of an age-related decrease in PE, we performed linear regression procedures for SDPdiff with PE as the predictor for each of the three target positions. All regression procedures yielded significant correlations (R2adjusted: 0.12 for 40°, 0.10 for 60°, and 0.05 for 90°; p’s < 0.05). However, no more than 12% of the age-related change in SDPdiff was explained by changes in PE, indicating that PE was not a strong predictor of SDPdiff. Discussion This is the first study to map proprioceptive development from early childhood to young adulthood in a large cross-sectional sample with an objective, quantitative measure. We employed a simple, child-appropriate method that conforms to a child’s anthropometrics and provides objective measures of forearm position sense acuity. Specifically, we systematically examined how bias and precision components of proprioceptive acuity changed during development. The main findings are that proprioceptive development between the ages of 5 and 18 years is not characterized as a development of decreasing bias, or inversely an increase in the accuracy of sensing. Instead, it is understood as a development where the precision improves with age. That is, the sensory response variability decreases with age. In summary, position sense does not become more accurate, but it does become more reliable during development. Neurophysiological Factors Underlying Proprioceptive Development What are the possible neural correlates of proprioceptive development in our sample of children? First, the observed age-related change in precision is not likely due to maturation of peripheral proprioceptors. We examined children as young as 5 years of age, but muscle spindles, which contribute significantly to position sense, are known to be mature in children as young as 3 years of age (Österlund et al., 2011). Moreover, spinal level somatosensory circuitry is functional at a young age. This is demonstrated by findings that threshold amplitudes for eliciting stretch and Hoffman reflex responses reach adult levels by 6–7 years of age (O’Sullivan et al., 1991; Grosset et al., 2007). Second, changes in muscle spindle sensitivity across age may in part explain the age related differences in position sense response variability. Research demonstrating that developmental changes in stretch response amplitude resulting from muscle spindle tuning by altered γ motoneuron activation (Grosset et al., 2007) underline this view. Given that muscle spindles and the relevant spinal circuitry are largely functional by middle childhood, developmental changes in spindle sensitivity would likely be influenced by changes in the descending signals projecting from supraspinal centers that target γ motorneurons which, in turn, innervate the intrafusal fibers of the spindles. Third, developmental changes in proprioceptive precision could be influenced by the development of supraspinal structures involved in the inter-hemispheric transfer of somatosensory information such as the corpus callosum. This is plausible, because our limb position matching task required the transfer of position information of the reference limb to the opposite hemisphere to correctly position the matching arm (Goble et al., 2005; Goble, 2010). Thus, the development of callosal projections that enable the communication between brain hemispheres could also have influenced the perceptual outcome measure. Axon growth and axon cytoskeletal changes of the corpus callosum are known to occur throughout childhood progressing into adulthood (Keshavan et al., 2002; Lebel et al., 2010). Finally, maturation of cortical structures responsible for somatosensory processing occurs at least until middle to late childhood. In typically developing children, the morphology of somatosensory evoked potentials (SEP) by median nerve stimulation is already mature by 3 years of age (Laget et al., 1976), but latencies are not mature until around 6–8 years of age (Boor et al., 1998; Boor and Goebel, 2000). This likely reflects developmental changes in the thalamo-cortical and lemniscal segments of somatosensory pathways. Contralateral somatosensory evoked fields (SEF) produced by index finger tactile stimulation become adult-like around the age of 2 years, while their morphology and latency continues to change through at least until 6 years of age (Pihko et al., 2009). Moreover, cortical changes such as maturation of the neurotransmitter system and axon and dendrite growth occur through adolescence (Nevalainen et al., 2014). Recent brain imaging examined how proprioceptive cortical networks develop in adolescence (Cignetti et al., 2016). Employing tendon vibration to stimulate muscle spindles which then induce illusory movement, the study demonstrated that the proprioceptive brain network is already firmly established prior to early adolescence but undergoes continued refinement as evidenced by a shift from diffuse to focal functional connectivity that was especially pronounced in fronto-striatal connections. In summary, an array of neural changes influence signal generation at the receptor level and the subsequent processing of these signals in supraspinal networks. Given that muscle spindles are unique sensors, because their response sensitivity to muscle tension can be altered by descending input to γ motorneurons, the fine tuning of these γ motorneurons pools is plausibly a major determinant of proprioceptive development. Alternative Explanations and Limitations This study employed a method in which participants had to match a given joint position by actively moving the forearm. This implies that an efference copy of the motor commands underlying such active movements was available to the nervous system to predict movement outcome. Thus, the sensing of joint position could have been influenced by at least three neural processes: the generation of internal predicted sensory feedback the processing of afferent sensory feedback and the integration of these two types of feedback to obtain a stable percept of limb position (Von Holst, 1954; Blakemore et al., 2001; Konczak et al., 2012). One therefore needs to recognize that this method does not provide a “pure” sensory measure of proprioceptive acuity (i.e., the sole processing of proprioceptive afferents). However, it is unlikely that the active movement to match a target position caused an unsystematic bias, across ages, that confounds the results. From a clinical perspective it may even be argued that the active movement testing represents a more functional scenario. Finally, one needs to consider if developmental differences in motor control could account for the differences in the proprioceptive precision. While one cannot fully exclude this possibility, we had attempted to address this potential confound between motor ability and proprioceptive acuity by: (a) measuring forearm position at rest; and (b) by allowing participants as much time as needed to achieve their desired limb position. Thus, possible differences in movement kinematics (speed, smoothness) had little to no influence on our perceptual outcome measure. In summary, the active joint position matching method provides a functional measure of a child’s proprioceptive status, but its results do not represent a measure of “pure” sensory proprioceptive acuity. It constitutes a trade-off between simplicity of testing and a higher measurement accuracy attainable through the use of specialized equipment such as a passive motion apparatus. Conclusion Here we presented a simple, time- and cost-effective method to objectively measure proprioceptive acuity in children using a bimanual manipulandum that can be body-scaled to the anthropometric dimensions of the child. Assessing limb position sense acuity of the forearm in a large cross-sectional sample of children between the ages of 5 and 18 years revealed that the development of proprioceptive acuity in typically developing children is not characterized by a change in systematic bias but rather is associated with an increase in precision with increasing age. Several neurodevelopmental processes contribute to the observed improvement of proprioceptive precision with age including changes in the regulation of muscle spindle sensitivity, the maturation of the corpus callosum, and the development of central somatosensory pathways. Further work to isolate structures/processes which are involved may prove beneficial for the design and implementation of effective training interventions for children with proprioceptive and associated sensorimotor deficits. Author Contributions JMH-W, I-LY and JK each made substantial contributions to the study design and analysis, as well as aided in drafting, revising and approving the final version of this written work. All authors listed meet the authorship criteria and no one qualified for authorship has been omitted. The first author, JMH-W, collected data, aided in the design of data analysis procedures, performed data analysis generated the first draft of this manuscript and contributed to edits and updates to the document. The second author, I-LY, collected data, aided in the design of data analysis procedures, performed data analysis and contributed to edits and updates of the manuscript. The third author, JK, provided guidance for refining the data collection method, aided in the design of data analysis procedures and contributed to edits and updates of the manuscript. Conflict of Interest Statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Data collection for this study was supported by numerous members of the Human Sensorimotor Control Laboratory. We here would like to thank Sara Norgren, Joseph Wentzel, Yu-Ting Tseng, Sanaz Khosravani, Naveen Elangovan, Anna Vera Cuppone, Allison Giddings and Dr. Joshua Aman for their help. We are grateful to Dr. Jin Hoon Park for introducing this project. Additionally, we want to extend our gratitude to the Twin Cities German Immersion School and Driven to Discover Project 2015 for welcoming our research into their communities. Finally, a special thanks goes to all the children, adults and families who volunteered for this study. Abbreviations Pdiffthe difference between the match and reference arm angular positions (degrees) PEmean of the differences between the match and reference arm angular positions SDPdiffstandard deviation of the differences between the match and reference arm angular positions. ==== Refs References Bairstow P. J. Laszlo J. I. (1981 ). Kinaesthetic sensitivity to passive movements and its relationship to motor development and motor control . Dev. Med. Child Neurol. 23 , 606 –616 . 10.1111/j.1469-8749.1981.tb02042.x 7286454 Blakemore S.-J. Frith C. D. Wolpert D. M. (2001 ). The cerebellum is involved in predicting the sensory consequences of action . Neuroreport 12 , 1879 –1884 . 10.1097/00001756-200107030-00023 11435916 Boor R. Goebel B. (2000 ). Maturation of near-field and far-field somatosensory evoked potentials after median nerve stimulation in children under 4 years of age . Clin. Neurophysiol. 111 , 1070 –1081 . 10.1016/s1388-2457(00)00262-5 10825715 Boor R. Goebel B. Taylor M. J. (1998 ). Subcortical somatosensory evoked potentials after median nerve stimulation in children . Eur. J. Paediatr. Neurol. 2 , 137 –143 . 10.1016/s1090-3798(98)80029-9 10726835 Cignetti F. Fontan A. Menant J. Nazarian B. Anton J. L. Vaugoyeau M. . (2016 ). Protracted development of the proprioceptive brain network during and beyond adolescence . Cereb. Cortex [Epub ahead of print]. 10.1093/cercor/bhv323 26733535 Coleman R. Piek J. P. Livesey D. J. (2001 ). A longitudinal study of motor ability and kinaesthetic acuity in young children at risk of developmental coordination disorder . Hum. Mov. Sci. 20 , 95 –110 . 10.1016/s0167-9457(01)00030-6 11471400 Elangovan N. Herrmann A. Konczak J. (2014 ). Assessing proprioceptive function: evaluating joint position matching methods against psychophysical thresholds . Phys. Ther. 94 , 553 –561 . 10.2522/ptj.20130103 24262599 Elliott J. M. Connolly K. J. Doyle A. J. (1988 ). Development of kinaesthetic sensitivity and motor performance in children . Dev. Med. Child Neurol. 30 , 80 –92 . 10.1111/j.1469-8749.1988.tb04728.x 3371573 Goble D. J. (2010 ). Proprioceptive acuity assessment via joint position matching: from basic science to general practice . Phys. Ther. 90 , 1176 –1184 . 10.2522/ptj.20090399 20522675 Goble D. Hurvitz E. Brown S. (2009 ). Deficits in the ability to use proprioceptive feedback in children with hemiplegic cerebral palsy . Int. J. Rehabil. Res. 32 , 267 –269 . 10.1097/MRR.0b013e32832a62d5 19318973 Goble D. J. Lewis C. A. Hurvitz E. A. Brown S. H. (2005 ). Development of upper limb proprioceptive accuracy in children and adolescents . Hum. Mov. Sci. 24 , 155 –170 . 10.1016/j.humov.2005.05.004 16043248 Grosset J. F. Mora I. Lambertz D. Pérot C. (2007 ). Changes in stretch reflexes and muscle stiffness with age in prepubescent children . J. Appl. Physiol. (1985) 102 , 2352 –2360 . 10.1152/japplphysiol.01045.2006 17347384 Hoare D. Larkin D. (1991 ). Kinesthetic abilities of clumsy children . Dev. Med. Child Neurol. 33 , 671 –678 . 10.1111/j.1469-8749.1991.tb14944.x 1916023 I.S.O. (1994 ). ISO 5725–1:1994(En) accuracy (Trueness and Precision) of measurement methods and results—Part 1: general principles and definitions . Available online at: https://www.iso.org/obp/ui/#iso:std:iso:5725:-1:ed-1:v1:en. Kaufman L. B. Schilling D. L. (2007 ). Implementation of a strength training program for a 5-year-old child with poor body awareness and developmental coordination disorder . Phys. Ther. 87 , 455 –467 . 10.2522/ptj.20060170 17374632 Keshavan M. S. Diwadkar V. A. DeBellis M. Dick E. Kotwal R. Rosenberg D. R. . (2002 ). Development of the corpus callosum in childhood, adolescence and early adulthood . Life Sci. 70 , 1909 –1922 . 10.1016/s0024-3205(02)01492-3 12005176 Kirby A. Sugden D. A. (2007 ). Children with developmental coordination disorders . J. R. Soc. Med. 100 , 182 –186 . 10.1258/jrsm.100.4.182 17404341 Konczak J. Corcos D. M. Horak F. Poizner H. Shapiro M. Tuite P. . (2009 ). Proprioception and motor control in Parkinson’s disease . J. Mot. Behav. 41 , 543 –552 . 10.3200/35-09-002 19592360 Konczak J. Sciutti A. Avanzino L. Squeri V. Gori M. Masia L. . (2012 ). Parkinson’s disease accelerates age-related decline in haptic perception by altering somatosensory integration . Brain 135 , 3371 –3379 . 10.1093/brain/aws265 23169922 Laget P. Raimbault J. D’Allest A. M. Flores-Guevara R. Mariani J. Thieriot-Prevost G. (1976 ). Maturation of somesthetic evoked potentials in man . Electroencephalogr. Clin. Neurophysiol. 40 , 499 –515 . 10.1016/0013-4694(76)90080-8 57038 Laszlo J. I. Bairstow P. J. (1980 ). The measurement of kinaesthetic sensitivity in children and adults . Dev. Med. Child Neurol. 22 , 454 –464 . 10.1111/j.1469-8749.1980.tb04350.x 7409337 Lebel C. Caverhill-Godkewitsch S. Beaulieu C. (2010 ). Age-related regional variations of the corpus callosum identified by diffusion tensor tractography . Neuroimage 52 , 20 –31 . 10.1016/j.neuroimage.2010.03.072 20362683 Li K. Y. Su W. J. Fu H. W. Pickett K. A. (2015 ). Kinesthetic deficit in children with developmental coordination disorder . Res. Dev. Disabil. 38 , 125 –133 . 10.1016/j.ridd.2014.12.013 25576876 Lord R. Hulme C. (1987 ). Kinesthetic sensitivity of normal and clumsy children . Dev. Med. Child Neurol. 29 , 720 –725 . 10.1111/j.1469-8749.1987.tb08816.x 3691972 Nevalainen P. Lauronen L. Pihko E. (2014 ). Development of human somatosensory cortical functions–what have we learned from magnetoencephalography: a review . Front. Hum. Neurosci. 8 :158 . 10.3389/fnhum.2014.00158 24672468 Österlund C. Liu J. X. Thornell L. E. Eriksson P. O. (2011 ). Muscle spindle composition and distribution in human young masseter and biceps brachii muscles reveal early growth and maturation . Anat. Rec. (Hoboken) 294 , 683 –693 . 10.1002/ar.21347 21370492 O’Sullivan M. C. Eyre J. A. Miller S. (1991 ). Radiation of phasic stretch reflex in biceps brachii to muscles of the arm in man and its restriction during development . J. Physiol. 439 , 529 –543 . 10.1113/jphysiol.1991.sp018680 1654417 Payne V. G. Isaacs L. D. (2008 ). Human Motor Development: A Lifespan Approach. 7th Edn. New York, NY : McGraw-Hill Pickett K. Konczak J. (2009 ). Measuring kinaesthetic sensitivity in typically developing children . Dev. Med. Child Neurol. 51 , 711 –716 . 10.1111/j.1469-8749.2008.03229.x 19207300 Piek J. P. Coleman-Carman R. (1995 ). Kinaesthetic sensitivity and motor performance of children with developmental co-ordination disorder . Dev. Med. Child Neurol. 37 , 976 –984 . 10.1111/j.1469-8749.1995.tb11952.x 8566458 Pihko E. Stephen J. Okada Y. Lauronen L. Nevalainen P. (2009 ). Maturation of somatosensory cortical processing from birth to adulthood revealed by magnetoencephalography . Clin. Neurophysiol. 120 , 1552 –1561 . 10.1016/j.clinph.2009.05.028 19560400 Visser J. Geuze R. H. (2000 ). Kinaesthetic acuity in adolescent boys: a longitudinal study . Dev. Med. Child Neurol. 42 , 93 –96 . 10.1111/j.1469-8749.2000.tb00052.x 10698325 Von Holst E. (1954 ). Relations between the central nervous system and the peripheral organs . Br. J. Anim. Behav. 2 , 89 –94 . 10.1016/s0950-5601(54)80044-x Wang T. N. Tseng M. H. Wilson B. N. Hu F. C. (2009 ). Functional performance of children with developmental coordination disorder at home and at school . Dev. Med. Child Neurol. 51 , 817 –825 . 10.1111/j.1469-8749.2009.03271.x 19416344 Zwicker J. G. Harris S. R. Klassen A. F. (2013 ). Quality of life domains affected in children with developmental coordination disorder: a systematic review . Child Care Health Dev. 39 , 562 –580 . 10.1111/j.1365-2214.2012.01379.x 22515477

PMC005xxxxxx/PMC5002404.txt

==== Front Front PharmacolFront PharmacolFront. Pharmacol.Frontiers in Pharmacology1663-9812Frontiers Media S.A. 10.3389/fphar.2016.00272PharmacologyEditorialEditorial: Molecular Mechanisms Protecting against Tissue Injury Wagener Frank A. D. T. G. 1*Immenschuh Stephan 2*1Department of Orthodontics and Craniofacial Biology, Radboud University Medical CenterNijmegen, Netherlands2Institute for Transfusion Medicine, Hannover Medical SchoolHannover, GermanyEdited by: Paola Patrignani, University of Chieti-Pescara, Italy Reviewed by: Lucia Trevisi, University of Padua, Italy; Melania Dovizio, University of Chieti-Pescara, Italy; Pietro Minuz, University of Verona, Italy *Correspondence: Frank A. D. T. G. Wagener frank.wagener@radboudumc.nlStephan Immenschuh immenschuh.stephan@mh-hannover.deThis article was submitted to Inflammation Pharmacology, a section of the journal Frontiers in Pharmacology 29 8 2016 2016 7 27228 4 2016 10 8 2016 Copyright © 2016 Wagener and Immenschuh.2016Wagener and ImmenschuhThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.The Editorial on the Research Topic Molecular Mechanisms Protecting against Tissue Injury tissue injuryinflammationcytoprotective moleculesprotective immunitycytoprotective enzymesDeutsche Forschungsgemeinschaft10.13039/501100001659IM 20/4-1Nederlandse Brandwonden Stichting10.13039/501100003503Else Kröner-Fresenius-Stiftung10.13039/501100003042 ==== Body In response to tissue injury acute inflammatory reactions occur that aim to restore homeostasis (Medzhitov, 2008). However, hampered resolution of inflammation can result in chronic inflammation and/or pathologic wound repair (Nathan and Ding, 2010). These conditions can result from excessive oxidative and inflammatory and/or overwhelmed adaptive response systems. They may also be triggered by a variety of other conditions including diabetes, infections, or aging. Targeted up-regulation of cytoprotective systems may be a therapeutic approach to ameliorate exacerbation of injury and prevent pathologic wound repair, fibrosis and/or cancer (Nathan, 2002). Such protective systems include various anti-oxidant, anti-inflammatory, anti-apoptotic molecules and also transporters or channels. In this Frontiers research topic various concepts how specific mechanisms can determine tissue damage or protection and their therapeutic potential in a number of pathological conditions and diseases are discussed. Tissue damage control is important at different levels to maintain homeostasis of cells, tissues and whole organisms. For example, when immunological reactions are primarily directed against the cause of tissue damage (e.g., pathogenic microbes), excessive collateral damage may occur. In such cases, avoiding additional tissue damage would be more important than elimination of the disease-triggering stimulus (Soares, 2014; Soares et al., 2014). Severe tissue injury can lead to chronic inflammation, fibrosis, disturbed developmental changes and cancer (Reuter et al., 2010), which underscores the need to prevent tissue damage and/or to promote regeneration. The coordinate immunological functions to maintain tissue homeostasis is orchestrated by a selected group of immune cells (Shalapour and Karin, 2015) such as dendritic cells (Mirzaee et al.) and regulatory T-cells (Lei et al.). Targeted modulation of these regulators may thus give control on the decisive machinery that determines immunity, inflammation, tissue remodeling, and cancer (Sutmuller et al., 2007). Infusion of regulatory T-cells facilitates tissue regeneration by preventing undesired immunological activity and by controlling resident non-immune tissue cells and forms an alternative strategy to dampen tissue injury (Lei et al.). Mizumura and colleagues describe how autophagy may promote tissue damage or repair and novel developments in its regulation (Mizumura et al.). In particular, the role of selective autophagy in a variety of human diseases and the therapeutic potential of this system is discussed. Surgery and other types of traumatic injury not only cause inflammatory injury, fibrosis, and scar formation (Brouwer et al., 2015), but are associated with the release of free heme (Wagener et al., 2003a). Heme is the prosthetic group of a number of physiologically important hemoproteins (e.g., hemoglobin, cytochromes or cyclooxygenase). However, when heme is not embedded in apoproteins which occurs in pathophysiological situations such as hemolysis or tissue injury, it can mediate or fuel oxidative, inflammatory, and fibrotic insults and may act as a danger signal (Nath et al., 1992; Wagener et al., 2003a,b; Lundvig et al.; Wegiel et al., 2015). High levels of free heme may contribute to or exacerbate tissue injury for example by promoting adhesion molecule expression and leukocyte recruitment (Wagener et al., 2001, 2003a; Belcher et al., 2010; Larsen et al., 2010; Gozzelino and Soares, 2011; Zenclussen et al., 2011). Therefore, protective mechanisms against free heme such as neutralization by either intra- or extra-cellular heme-binding proteins (e.g., hemopexin) or enzymatic heme-degradation by heme oxygenases (HOs) may have important protective functions (Immenschuh et al., 1995; Wagener et al., 2003b, 2013; Kartikasari et al., 2009). Previously, it has been shown that exposure to small injurious stress stimuli is protective against a follow-up stronger stress. This so-called preconditioning may thus have beneficial effects to protect tissues against injury by promoting cytoprotective signals (Murry et al., 1986). It has been shown that this tissue protection is associated with the induction of cytoprotective genes, such as HO-1, A20, hemopexin, and biliverdin reductase (BVR) (Keyse and Tyrrell, 1989; Hancock et al., 1998). A major protection system is made up by the HO-1/BVR module. HO is the rate-limiting enzyme of heme degradation, which produces carbon monoxide, iron, and biliverdin, which is converted into bilirubin by BVR. Induction of these cytoprotective proteins has been shown to prolong graft survival after solid organ transplantation by skewing toward a more tolerant immune system (Wagener et al., 2003b). Moreover, HO-activity attenuates the expression of pro-inflammatory cytokines and vascular adhesion molecules, promoting resolution of inflammation (Wagener et al., 2001; Di Francesco et al., 2009). Similar to transplanted organs, tumors are immunologically different to healthy, non-transformed tissues, suggesting that they may also benefit from these cytoprotective enzymes. Likewise, fetuses in the uterus are expressing maternal and paternal proteins, and are thus comparable with an allotransplant. In this topic, it is described how probing antioxidant genes such as HO and BVR could form novel targets for preventing pathological pregnancies (Ozen et al.) and cancer (Gibbs et al.), respectively. Induction of HO activity or exposure to its effector molecules protect against pathological pregnancies, whereas reduced HO-activity disturbs normal pregnancy (Ozen et al.). BVR interacts with various protein kinases and is involved in a complex system of regulatory pathways. The effects of BVR not only have an important impact on the pathogenesis of cancer, but targeting of this enzyme may ultimately afford the development of novel therapies in cancer (Gibbs et al.). Although, cytoprotective systems such as HO-1/BVR are generally considered to be beneficial, they may be harmful under particular circumstances. For example, it has been demonstrated that HO-1 is involved in the pathogenesis of cancer, because it can protect tumor cells against immune surveillance (Was et al., 2010). Moreover, HO-1 has recently been suggested to be involved in transforming obesity to diabetes (Jais et al., 2014). Therefore, the effects of such cytoprotective systems appear to be critically dependent on cell-type and tissue-context-specific mechanisms. Another review in this Frontiers topic by Horbach et al. addresses the role of the nuclear factors upstream stimulatory factor (USFs)-1 and -2 in the context of carcinogenesis and tissue injury. USFs have primarily been considered to be involved in the regulation of metabolism, but have also been shown to be intimately associated with tissue protection and the pathogenesis of cancer. Here, the complexity of USF-1 and -2 regulation by various kinases in carcinogenesis is discussed. Finally, a feasible approach to afford targeted protection in various pathological conditions is to trigger defined protective pathways with safe plant components. For example, induction of anti-inflammatory pathways with herbal compounds and antioxidants suppressed expression of proinflammatory cytokines in human peripheral blood mononuclear cells (Spatuzza et al.), adipose cells (Zagotta et al.), and dendritic cells (Mirzaee et al.). Interestingly, many dietary and natural compounds have been demonstrated to activate nuclear factor erythroid 2-related factor (Nrf2), which in turn induces a number of protective enzymes, such as HO-1, and promote therapeutic effects in cardiovascular diseases (Barbagallo et al., 2013). However, when translating novel protective strategies from the bench to the clinic possible differences in experimental outcome between animal models, cell lines, healthy controls and patients need to be considered (Dorresteijn et al., 2015). Better insights into the observed differences between pre-conditioning and post-conditioning in relation to tissue repair could further deepen our understanding of these regulatory pathways. It appears likely that learning more about the molecular mechanisms protecting against tissue damage will enable the development of better strategies to prevent or ameliorate wound repair and promote healthy aging. Author contributions All authors listed, have made substantial, direct and intellectual contribution to the work, and approved it for publication. Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer MD and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review. FW was supported by a grant from the Vaillant Foundation, EFRO (FlowPlast), and the Dutch Burns Foundation (#16.03). SI was funded by a grant from the “Deutsche Forschungsgemeinschaft (DFG)” (IM 20/4-1) and the “Else Kröner-Fresenius-Stiftung” (EKFS_2012_A309). ==== Refs References Barbagallo I. Galvano F. Frigiola A. Cappello F. Riccioni G. Murabito P. . (2013 ). Potential therapeutic effects of natural heme oxygenase-1 inducers in cardiovascular diseases . Antioxid. Redox Signal. 18 , 507 –521 . 10.1089/ars.2011.4360 23025298 Belcher J. D. Vineyard J. V. Bruzzone C. M. Chen C. Beckman J. D. Nguyen J. . (2010 ). Heme oxygenase-1 gene delivery by Sleeping Beauty inhibits vascular stasis in a murine model of sickle cell disease . J. Mol. Med. 88 , 665 –675 . 10.1007/s00109-010-0613-6 20306336 Brouwer K. M. Lundvig D. M. Middelkoop E. Wagener F. A. von den Hoff J. W. (2015 ). Mechanical cues in orofacial tissue engineering and regenerative medicine . Wound Repair Regen. 23 , 302 –311 . 10.1111/wrr.12283 25787133 Di Francesco L. Totani L. Dovizio M. Piccoli A. Di Francesco A. Salvatore T. . (2009 ). Induction of prostacyclin by steady laminar shear stress suppresses tumor necrosis factor-alpha biosynthesis via heme oxygenase-1 in human endothelial cells . Circ. Res. 104 , 506 –513 . 10.1161/CIRCRESAHA.108.191114 19122175 Dorresteijn M. J. Paine A. Zilian E. Fenten M. G. Frenzel E. Janciauskiene S. . (2015 ). Cell-type-specific downregulation of heme oxygenase-1 by lipopolysaccharide via Bach1 in primary human mononuclear cells . Free Radic. Biol. Med. 78 , 224 –232 . 10.1016/j.freeradbiomed.2014.10.579 25463280 Gozzelino R. Soares M. P. (2011 ).Heme sensitization to TNF-mediated programmed cell death. Adv. Exp. Med. Biol. 691 , 211 –219 . 10.1007/978-1-4419-6612-4_22 21153325 Hancock W. W. Buelow R. Sayegh M. H. Turka L. A. (1998 ). Antibody-induced transplant arteriosclerosis is prevented by graft expression of anti-oxidant and anti-apoptotic genes . Nat. Med. 4 , 1392 –1396 . 10.1038/3982 9846576 Immenschuh S. Iwahara S. Satoh H. Nell C. Katz N. Muller-Eberhard U. (1995 ). Expression of the mRNA of heme-binding protein 23 is coordinated with that of heme oxygenase-1 by heme and heavy metals in primary rat hepatocytes and hepatoma cells . Biochemistry 34 , 13407 –13411 . 10.1021/bi00041a018 7577927 Jais A. Einwallner E. Sharif O. Gossens K. Lu T. T. Soyal S. M. . (2014 ). Heme oxygenase-1 drives metaflammation and insulin resistance in mouse and man . Cell 158 , 25 –40 . 10.1016/j.cell.2014.04.043 24995976 Kartikasari A. E. Wagener F. A. Yachie A. Wiegerinck E. T. Kemna E. H. Swinkels D. W. (2009 ). Hepcidin suppression and defective iron recycling account for dysregulation of iron homeostasis in heme oxygenase-1 deficiency . J. Cell. Mol. Med. 13 , 3091 –3102 . 10.1111/j.1582-4934.2008.00494.x 18774956 Keyse S. M. Tyrrell R. M. (1989 ). Heme oxygenase is the major 32-kDa stress protein induced in human skin fibroblasts by UVA radiation, hydrogen peroxide, and sodium arsenite . Proc. Natl. Acad. Sci. U.S.A. 86 , 99 –103 . 10.1073/pnas.86.1.99 2911585 Larsen R. Gozzelino R. Jeney V. Tokaji L. Bozza F. A. Japiassu A. M. . (2010 ). A central role for free heme in the pathogenesis of severe sepsis . Sci. Transl. Med. 2 , 51r a71. 10.1126/scitranslmed.3001118 20881280 Medzhitov R. (2008 ). Origin and physiological roles of inflammation . Nature 454 , 428 –435 . 10.1038/nature07201 18650913 Murry C. E. Jennings R. B. Reimer K. A. (1986 ). Preconditioning with ischemia: a delay of lethal cell injury in ischemic myocardium . Circulation 74 , 1124 –1136 . 10.1161/01.CIR.74.5.1124 3769170 Nath K. A. Balla G. Vercellotti G. M. Balla J. Jacob H. S. Levitt M. D. . (1992 ). Induction of heme oxygenase is a rapid, protective response in rhabdomyolysis in the rat . J. Clin. Invest. 90 , 267 –270 . 10.1172/JCI115847 1634613 Nathan C. (2002 ). Points of control in inflammation . Nature 420 , 846 –852 . 10.1038/nature01320 12490957 Nathan C. Ding A. (2010 ). Nonresolving inflammation . Cell 140 , 871 –882 . 10.1016/j.cell.2010.02.029 20303877 Reuter S. Gupta S. C. Chaturvedi M. M. Aggarwal B. B. (2010 ). Oxidative stress, inflammation, and cancer: how are they linked? Free Radic. Biol. Med. 49 , 1603 –1616 . 10.1016/j.freeradbiomed.2010.09.006 20840865 Shalapour S. Karin M. (2015 ). Immunity, inflammation, and cancer: an eternal fight between good and evil . J. Clin. Invest. 125 , 3347 –3355 . 10.1172/JCI80007 26325032 Soares M. P. (2014 ). “Nuts and bolts” of disease tolerance . Immunity 41 , 176 –178 . 10.1016/j.immuni.2014.07.011 25148020 Soares M. P. Gozzelino R. Weis S. (2014 ). Tissue damage control in disease tolerance . Trends Immunol. 35 , 483 –494 . 10.1016/j.it.2014.08.001 25182198 Sutmuller R. Garritsen A. Adema G. J. (2007 ). Regulatory T cells and toll-like receptors: regulating the regulators . Ann. Rheumat. Dis. 66 (Suppl. 3 ), iii91 –iii95 . 10.1136/ard.2007.078535 17934105 Wagener F. A. Dankers A. C. van Summeren F. Scharstuhl A. van den Heuvel J. J. Koenderink J. B. . (2013 ). Heme Oxygenase-1 and breast cancer resistance protein protect against heme-induced toxicity . Curr. Pharm. Des. 19 , 2698 –2707 . 10.2174/1381612811319150004 23092328 Wagener F. A. Eggert A. Boerman O. C. Oyen W. J. Verhofstad A. Abraham N. G. . (2001 ). Heme is a potent inducer of inflammation in mice and is counteracted by heme oxygenase . Blood 98 , 1802 –1811 . 10.1182/blood.V98.6.1802 11535514 Wagener F. A. van Beurden H. E. von den Hoff J. W. Adema G. J. Figdor C. G. (2003a ). The heme-heme oxygenase system: a molecular switch in wound healing . Blood 102 , 521 –528 . 10.1182/blood-2002-07-2248 12649161 Wagener F. A. Volk H. D. Willis D. Abraham N. G. Soares M. P. Adema G. J. . (2003b ). Different faces of the heme-heme oxygenase system in inflammation . Pharmacol. Rev. 55 , 551 –571 . 10.1124/pr.55.3.5 12869663 Was H. Dulak J. Jozkowicz A. (2010 ). Heme oxygenase-1 in tumor biology and therapy . Curr. Drug Targets 11 , 1551 –1570 . 10.2174/1389450111009011551 20704546 Wegiel B. Hauser C. J. Otterbein L. E. (2015 ). Heme as a danger molecule in pathogen recognition . Free Radic. Biol. Med. 89 , 651 –661 . 10.1016/j.freeradbiomed.2015.08.020 26456060 Zenclussen M. L. Casalis P. A. El-Mousleh T. Rebelo S. Langwisch S. Linzke N. . (2011 ). Haem oxygenase-1 dictates intrauterine fetal survival in mice via carbon monoxide . J. Pathol. 225 , 293 –304 . 10.1002/path.2946 21744344

PMC005xxxxxx/PMC5002405.txt

==== Front Front NeurolFront NeurolFront. Neurol.Frontiers in Neurology1664-2295Frontiers Media S.A. 10.3389/fneur.2016.00141NeuroscienceGeneral CommentaryCommentary: The Flavonoid Baicalein Rescues Synaptic Plasticity and Memory Deficits in a Mouse Model of Alzheimer’s Disease Chirumbolo Salvatore 1*Bjørklund Geir 21Department of Neurological and Movement Science, University of Verona, Verona, Italy2Council for Nutritional and Environmental Medicine, Mo i Rana, NorwayEdited by: Irving E. Vega, Michigan State University, USA Reviewed by: Hector De Jesus-Cortes, Massachusetts Institute of Technology, USA; Manoj Kumar Jaiswal, Columbia University, USA *Correspondence: Salvatore Chirumbolo, salvatore.chirumbolo@univr.itSpecialty section: This article was submitted to Neurodegeneration, a section of the journal Frontiers in Neurology 29 8 2016 2016 7 14131 5 2016 17 8 2016 Copyright © 2016 Chirumbolo and Bjørklund.2016Chirumbolo and BjørklundThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.A commentary on The Flavonoid Baicalein Rescues Synaptic Plasticity and Memory Deficits in a Mouse Model of Alzheimer’s Disease by Gu XH, Xu LJ, Liu ZQ, Wei B, Yang YJ, Xu GG, et al. Behav Brain Res (2016) 311:309–21. doi: 10.1016/j.bbr.2016.05.052Alzheimer diseasesynaptic lossGSK-3βbaicaleinflavonoidsastrocyteglutamate receptors ==== Body A recent article by Gu et al. reported that the long-term oral administration of baicalein in mouse, inhibited the activity of lipooxygenase 12/15 LXO and glycogen synthase kinase 3β (GSK-3β) in hippocampal slices, also reducing the activity of β-secretase enzyme (BACE1) and the concentration of total beta-amyloid proteins (Aβ), leading to the in vivo restoration of spine number, synaptic plasticity, and memory deficits (1). Baicalein (5,6,7-trihydroxy-2-phenyl-chromen-4-one) is present in the plant genus Scutellaria, e.g., Scutellaria baicalensis, the flavone recently went in the spotlight for its ability in preventing Alzheimer disease (AD) and other neurodegenerative disorders (2–4). According to the authors, baicalein is able to prevent the impairment in the hippocampal long-term potentiation (LTP) induced by β-amyloids (Aβ) and improve cognitive deficit associated with AD (1). Recently, also metformin exhibited positive effects on AD quite comparable to baicalein, i.e., the inhibition of the activity of Aβ on LTP in a high fat diet (HFD) rat model (5). As evidence about the anti-obesity and anti-diabetes activity of baicalein was reported elsewhere, the activity of metformin may give insights on the role exerted by baicalein in AD. Despite the evidence reported by Asadbegi et al. some controversial result yet exists, as in type 2 diabetes models metformin increases the expression of the β-amyloid precursor protein (APP), through the activation of NF-κB, causing accumulation of Aβ in rat brain (5, 6). Whether metformin has either a beneficial or a noxious effect depending on the cellular system considered or if it acts through a typical bimodal mechanism, a hallmark shared by several bioflavonoids, is yet far to be fully elucidated. As like as flavonoids, even metformin, which is a well known anti-diabetic drug, possesses anti-oxidant and anti-inflammatory properties (7). This recent evidence suggested us a possible mechanism for baicalein action in AD prevention, as also for the flavonoid baicalein, a bimodal activity might be addressed, at least as concern its action on oxidative stress and metabolism. The main advantage respect to metformin is that baicalein is a plant-derived natural compound. As many bioflavonoids, baicalein inhibits phosphorylation of many fundamental signaling protein kinases, leading to its typical anti-inflammatory activity, e.g., it inhibits the ERK/MAPK signaling cascade, acting on the phosphorylation of MEK-1 by Raf-1 and inducing dampening of NF-κB (8). This anti-phoshorylation property was reported also by Gu et al. as baicalein should exert its action by preventing the phosphorylation of tau protein in APP/PS1 mice (1). Still, a metabolic, anti-oxidant activity from baicalein, besides its targeting glutamatergic neurotransmission, may explain some evidence reported (1). The authors showed that Aβ binding to synapse-related sites mediates the reduction of synaptic loss and also modulates the damage of the glutamatergic synaptic transmission, while baicalein can prevent this damage (1). Furthermore, the paper by Gu et al. would suggest whether the mechanism by which baicalein acts on AD may resemble some other neuro-protective compound, in order to focus onto commonly shared targets. Curiously, as like as metformin, also baicalein succeeded in rescuing hippocampal LTP impaired by the action of Aβ and the authors suggested a mechanism working through the PI3K signaling pathway (1, 5). Further bioflavonoids seem to have a comparable effect on the synaptic failure induced by Aβ, e.g., the glycosylated-flavonol troxerutin, is able also to reduce impairments in cognitive function caused by metabolic syndrome or HFD (9, 10) and the flavone nobiletin, from Citrus fruit, rescued cognitive function and memory impairment in animal fear conditioning tests by reverting the inhibition caused by Aβ on the membrane trafficking of glutamate receptors, which is required for LTP (11). And even nobiletin has an impact on obesity and metabolic syndrome (12). The question is if there is a fundamental connection between metabolic syndrome and neurodegenerative disorders in the reported action of baicalein. A possible mechanism linking these effects and shared by metformin besides flavonoids might involve their indirect action on the impaired insulin signaling and glucose metabolism, which causes AD in brain (13). For example, the paper by Gu et al. showed that baicalein inhibits the activity of GSK-3β. Brain derived insulin is expressed by the activity of the transcription factor neurogenic differentiation 1 (NeuroD1) and inhibitors of GSK-3β increase insulin gene (Ins2) and NeuroD1 in the mouse brain, through a Wnt/β-catenin pathway (14). Gu et al. also reported that baicalein increased the N-methyl-d-aspartate glutamate receptor (NMDAR)-mediated LTP in rat hippocampus and it is well known that glutamate, which is spontaneously released by astrocytes, is involved in the setting of a LTP threshold (15). Therefore, it is possible that an activity of baicalein described by Gu et al. may target the insulin signaling by inhibiting the GSK-3β pathway. The most acknowledged activity of plants bioflavonoids is their targeting oxidative stress, gene transcription factors involved in the cellular response to oxidative stress and mechanisms involved in the scavenging of oxidative stressors. Actually, it is not surprising that for the interesting evidence reported by Gu et al. the anti-oxidant property of baicalein might even play a major role. For example, the activity of membrane PI3K in astrocytes, through the up-regulation of the signaling mechanism, involving the activation of phospholipase C and protein kinase C-delta (PKC-δ), causes a weak but persistent stimulation of the N-methyl-d-aspartate receptors (NMDARs) from astrocytes and this mechanism is coupled with the phosphorylation at Thr(395), Ser(433), and Thr(439) by PKC-δ of the nuclear factor-(erythroid-derived 2)-like 2 (Nrf2), then linking glutamate transmission with anti-oxidant activity and neuron survival. Translocation to the nucleus of Nrf2 enhances the NMDAR stimulation (16). Baicalein may target this mechanism, in the system described by Gu et al. This should suggest that a possible effect of baicalein may be exerted also on the glia, besides neurons, particularly during the inflammatory response (17). In this circumstance, further research should address the widespread opinion that plant polyphenols act on more than one level, fundamentally by targeting signaling system that are involved in cell survival and stress response. This would involve the existence of complex machinery encompassing both homeostasis of the energetic balance and cell response to stressors. In this sense, the bio-molecular mechanism by which baicalein caused the evidence reported by Gu et al. may mainly depend on its anti-oxidant property, as β-secretase 1 (BACE1) activity, which increases the level of APP, is promoted by pro-oxidant factors (18). This potential should shed a light also on the neuroprotective role of those natural compounds, which are recognized as anti-diabetic and obesity-preventive molecules. Author Contributions SC conceived the paper, SC and GB read the literature, GB contribute in the editing of the document, SC wrote with GB the manuscript, SC and GB revised the first draft, SC submitted the manuscript. Conflict of Interest Statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. ==== Refs References 1 Gu XH Xu LJ Liu ZQ Wei B Yang YJ Xu GG The flavonoid baicalein rescues synaptic plasticity and memory deficits in a mouse model of Alzheimer’s disease . Behav Brain Res (2016 ) 311 :309 –21 .10.1016/j.bbr.2016.05.052 27233830 2 Wei D Tang J Bai W Wang Y Zhang Z . Ameliorative effects of baicalein on an amyloid-β induced Alzheimer’s disease rat model: a proteomics study . Curr Alzheimer Res (2014 ) 11 (9 ):869 –81 .10.2174/1567205011666141001113619 25274114 3 Zhang SQ Obregon D Ehrhart J Deng J Tian J Hou H Baicalein reduces β-amyloid and promotes nonamyloidogenic amyloid precursor protein processing in an Alzheimer’s disease transgenic mouse model . J Neurosci Res (2013 ) 91 (9 ):1239 –46 .10.1002/jnr.23244 23686791 4 Xue X Liu H Qi L Li X Guo C Gong D Baicalein ameliorated the upregulation of striatal glutamatergic transmission in the mice model of Parkinson’s disease . Brain Res Bull (2014 ) 103 :54 –9 .10.1016/j.brainresbull.2014.02.004 24576689 5 Asadbegi M Yaghmaei P Salehi I Ebrahim-Habibi A Komaki A . Neuroprotective effects of metformin against Aβ-mediated inhibition of long-term potentiation in rats fed a high-fat diet . Brain Res Bull (2016 ) 121 :178 –85 .10.1016/j.brainresbull.2016.02.005 26861514 6 Picone P Nuzzo D Caruana L Messina E Barera A Vasto S Metformin increases APP expression and processing via oxidative stress, mitochondrial dysfunction and NF-κB activation: use of insulin to attenuate metformin’s effect . Biochim Biophys Acta (2015 ) 1853 (5 ):1046 –59 .10.1016/j.bbamcr.2015.01.017 25667085 7 Porceddu PF Ishola IO Contu L Morelli M . Metformin prevented dopaminergic neurotoxicity induced by 3,4-methylenedioxymethamphetamine administration . Neurotox Res (2016 ) 30 (1 ):101 –9 .10.1007/s12640-016-9633-5 27251371 8 Nakahata N Tsuchiya C Nakatani K Ohizumi Y Ohkubo S . Baicalein inhibits Raf-1-mediated phosphorylation of MEK-1 in C6 rat glioma cells . Eur J Pharmacol (2003 ) 461 (1 ):1 –7 .10.1016/S0014-2999(02)02950-3 12568909 9 Babri S Mohaddes G Feizi I Mohammadnia A Niapour A Alihemmati A Effect of troxerutin on synaptic plasticity of hippocampal dentate gyrus neurons in a β-amyloid model of Alzheimer’s disease: an electrophysiological study . Eur J Pharmacol (2014 ) 732 :19 –25 .10.1016/j.ejphar.2014.03.018 24681055 10 Lu J Wu DM Zheng ZH Zheng YL Hu B Zhang ZF . Troxerutin protects against high cholesterol-induced cognitive deficits in mice . Brain (2011 ) 134 (Pt 3 ):783 –97 .10.1093/brain/awq376 21252113 11 Yamakuni T Nakajima A Ohizumi Y . [Preventive action of nobiletin, a constituent of AURANTII NOBILIS PERICARPIUM with anti-dementia activity, against amyloid-beta peptide-induced neurotoxicity expression and memory impairment] . Yakugaku Zasshi (2010 ) 130 (4 ):517 –20 .10.1248/yakushi.130.517 20371995 12 Lee YS Cha BY Choi SS Choi BK Yonezawa T Teruya T Nobiletin improves obesity and insulin resistance in high-fat diet-induced obese mice . J Nutr Biochem (2013 ) 24 (1 ):156 –62 .10.1016/j.jnutbio.2012.03.014 22898571 13 Chen Y Zhang J Zhang B Gong CX . Targeting insulin signaling for the treatment of Alzheimer’s Disease . Curr Top Med Chem (2016 ) 16 (5 ):485 –92 .10.2174/1568026615666150813142423 26268336 14 Lee J Kim K Yu SW Kim EK . Wnt3a upregulates brain-derived insulin by increasing NeuroD1 via Wnt/β-catenin signaling in the hypothalamus . Mol Brain (2016 ) 9 :24 .10.1186/s13041-016-0207-5 26956881 15 Bonansco C Couve A Perea G Ferradas CÁ Roncagliolo M Fuenzalida M . Glutamate released spontaneously from astrocytes sets the threshold for synaptic plasticity . Eur J Neurosci (2011 ) 33 (8 ):1483 –92 .10.1111/j.1460-9568.2011.07631.x 21395864 16 Jimenez-Blasco D Santofimia-Castaño P Gonzalez A Almeida A Bolaños JP . Astrocyte NMDA receptors’ activity sustains neuronal survival through a Cdk5-Nrf2 pathway . Cell Death Differ (2015 ) 22 (11 ):1877 –89 .10.1038/cdd.2015.49 25909891 17 Hong S Beja-Glasser VF Nfonoyim BM Frouin A Li S Ramakrishnan S Complement and microglia mediate early synapse loss in Alzheimer mouse models . Science (2016 ) 352 (6286 ):712 –6 .10.1126/science.aad8373 27033548 18 Choudhry F Howlett DR Richardson JC Francis PT Williams RJ Pro-oxidant diet enhances β/γ secretase-mediated APP processing in APP/PS1 transgenic mice . Neurobiol Aging (2012 ) 33 (5 ):960 –8 .20724034

PMC005xxxxxx/PMC5002406.txt

==== Front Front NutrFront NutrFront. Nutr.Frontiers in Nutrition2296-861XFrontiers Media S.A. 10.3389/fnut.2016.00037NutritionMethodsA Consensus Proposal for Nutritional Indicators to Assess the Sustainability of a Healthy Diet: The Mediterranean Diet as a Case Study Donini Lorenzo M. 12*Dernini Sandro 23Lairon Denis 24Serra-Majem Lluis 25Amiot Marie-Josèphe 24del Balzo Valeria 12Giusti Anna-Maria 12Burlingame Barbara 6Belahsen Rekia 7Maiani Giuseppe 8Polito Angela 8Turrini Aida 8Intorre Federica 8Trichopoulou Antonia 9Berry Elliot M. 2101Sapienza University of Rome, Rome, Italy2CIISCAM-International Inter-University Center for Mediterranean Food Culture Studies, Rome, Italy3Food and Agriculture Organization of the United Nations, Rome, Italy4INRA 1260, INSERM 1062, Aix Marseille University, Marseille, France5CIBER OBN, Instituto de Salud Carlos III, University of Las Palmas de Gran Canaria, Las Palmas, Spain6Deakin University, Melbourne, Australia7Chouaib Doukkali University, El Jadida, Morocco8Council for Agricultural Research and Economics-Research Center on Food and Nutrition (CRA-NUT, formerly INRAN), Rome, Italy9Hellenic Health Foundation, Athens, Greece10Braun School of Public Health, Hebrew University-Hadassah Medical School, Jerusalem, IsraelEdited by: Carola Strassner, Münster University of Applied Sciences, Germany Reviewed by: Youssef Aboussaleh, Ibn Tofail University, Morocco; Susanne Gjedsted Bügel, University of Copenhagen, Denmark; Adam Drewnowski, University of Washington, USA *Correspondence: Lorenzo M. Donini, lorenzomaria.donini@uniroma1.itSpecialty section: This article was submitted to Nutrition and Environmental Sustainability, a section of the journal Frontiers in Nutrition 29 8 2016 2016 3 3722 5 2016 15 8 2016 Copyright © 2016 Donini, Dernini, Lairon, Serra-Majem, Amiot, del Balzo, Giusti, Burlingame, Belahsen, Maiani, Polito, Turrini, Intorre, Trichopoulou and Berry.2016Donini, Dernini, Lairon, Serra-Majem, Amiot, del Balzo, Giusti, Burlingame, Belahsen, Maiani, Polito, Turrini, Intorre, Trichopoulou and BerryThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.Background There is increasing evidence of the multiple effects of diets on public health nutrition, society, and environment. Sustainability and food security are closely interrelated. The traditional Mediterranean Diet (MD) is recognized as a healthier dietary pattern with a lower environmental impact. As a case study, the MD may guide innovative inter-sectorial efforts to counteract the degradation of ecosystems, loss of biodiversity, and homogeneity of diets due to globalization through the improvement of sustainable healthy dietary patterns. This consensus position paper defines a suite of the most appropriate nutrition and health indicators for assessing the sustainability of diets based on the MD. Methods In 2011, an informal International Working Group from different national and international institutions was convened. Through online and face-to-face brainstorming meetings over 4 years, a set of nutrition and health indicators for sustainability was identified and refined. Results Thirteen nutrition indicators of sustainability relating were identified in five areas. Biochemical characteristics of food (A1. Vegetable/animal protein consumption ratios; A2. Average dietary energy adequacy; A3. Dietary Energy Density Score; A4. Nutrient density of diet), Food Quality (A5. Fruit and vegetable consumption/intakes; A6. Dietary Diversity Score), Environment (A7. Food biodiversity composition and consumption; A8. Rate of Local/regional foods and seasonality; A9. Rate of eco-friendly food production and/or consumption), Lifestyle (A10. Physical activity/physical inactivity prevalence; A11. Adherence to the Mediterranean dietary pattern), Clinical Aspects (A12. Diet-related morbidity/mortality statistics; A13. Nutritional Anthropometry). A standardized set of information was provided for each indicator: definition, methodology, background, data sources, limitations of the indicator, and references. Conclusion The selection and analysis of these indicators has been performed (where possible) with specific reference to the MD. Sustainability of food systems is an urgent priority for governments and international organizations to address the serious socioeconomic and environmental implications of short-sighted and short-term practices for agricultural land and rural communities. These proposed nutrition indicators will be a useful methodological framework for designing health, education, and agricultural policies in order, not only to conserve the traditional diets of the Mediterranean area as a common cultural heritage and lifestyle but also to enhance the sustainability of diets in general. sustainable dietsnutrition indicatorsMediterranean dietdietary energy densitydietary nutrient densitydietary diversityphysical activitynon-communicable chronic diseases ==== Body Introduction There is increasing evidence of the multiple effects and cost of diets on public health nutrition, society, and environment (1–4). Sustainability and food security are closely interrelated (5). Food systems around the world are changing rapidly, with profound implications for diets and nutritional outcomes. The sustainable diets concept (6) highlights the role of sustainable consumption as a driver of sustainable production, for food systems’ transformation toward more sustainable food consumption and production patterns, which are among the most important drivers of environmental pressures (7, 8). Food systems need to grow within the framework of finite and often reduced funds and need to make use of natural resources and skills in a sustainable manner to conserve the fragile ecosystem balance. There is an increasing need to develop a holistic view on sustainable food systems, from production to consumption and diets. This can be achieved through linkage to the enhancement of more sustainable dietary models. In the early 1980s, the notion of “sustainable diets” arose to recommend diets, which would be healthier for both the environment as well as for consumers (9). With food globalization and the increased industrialization of agricultural systems, the concepts of sustainable diets and agro-food systems had been neglected. In the last decade, the interest in sustainable diets has been revived by a growing body of scientific evidence of the non-sustainability of current dietary trends (10–13). However, it is not clear that high nutritional quality is always associated with low environmental impact (14). The traditional Mediterranean Diet (MD) has been studied in-depth and recognized as a healthier dietary pattern characterized by a lower environmental impact (15–18). The health benefits of the MD in preventing chronic diseases have been well recognized by the scientific community, since the pioneer Seven Countries Study, conducted by Ancel Keys, established the association of a traditional Mediterranean dietary pattern with markedly reduced coronary heart disease mortality (19–21). Later, additional benefits of the MD have been widely reported scientifically for diseases other than cardiovascular, such as obesity, diabetes, cancer, depression, cognitive decline as well as improved quality of life (22–30). But, despite these well-documented health benefits and the low environmental impact of the MD, current surveys show a decline in its adherence in Northern, Southern, and Eastern Mediterranean countries, because of multifactorial influences – lifestyle changes, globalization of food markets, and economic and sociocultural factors (26, 31–39). The three main domains of sustainability – economic, social, and environmental – need to be integrated into the dimensions of nutrition, health, and culture. During several recent international seminars, four main thematic areas of sustainability have been identified (1) nutrition, health, and lifestyle; (2) environment including agro-biodiversity; (3) economy; (4) society and culture (40). The assessment and development of sustainable diet models requires awareness among consumers, producers, and governments that agriculture, food, nutrition, health, culture, environment, and sustainability are strongly interdependent. The MD, as a case study, may guide innovative inter-sectorial efforts to counteract the degradation of ecosystems, loss of biodiversity, and homogenization of diets due to globalization through the improvement of sustainable dietary patterns with their health benefits. For this purpose, it is necessary to define the sustainability of diets through the analysis of evidence, development of methods and indicators, and the development/promotion of policy guidelines. In particular, in the context of sustainable consumption and production, indicators are necessary to monitor time-trends whether a society’s consumption and production patterns lead to more socially equitable and environmentally sustainable development. They are also necessary to evaluate the impact of dietary patterns on long-term health status and, in particular, on the pathogenesis and incidence of non-communicable chronic diseases. A number of international organizations, as well as different governments, have developed sets of indicators for sustainable consumption and production, mostly as attempts to monitor sustainable development, but also as part, or in support, of dedicated sustainable consumption and production strategies (41). However, from a methodological approach, there are at least three diverging criteria to define sustainability indicators. (1) According to the International Institute for Sustainable Development “an indicator quantifies and simplifies phenomena and helps us understand complex realities” (42). (2) According to the Organization for Economic Cooperation and Development, an indicator is “a parameter, or a value derived from parameters, which points to, provides information about, or describes the state of a phenomenon/environment/area, with a significance extending beyond that directly associated with its value” (43). (3) According to Food and Agriculture Organization of the United Nations (FAO), an indicator does not reduce to the data on which it is based; it generally comprises elements (a cutoff value, a frame of reference, a mode of expression, etc.), which allow a relatively universal appreciation of the information it supplies and also facilitate comparison in time and space (44–46). This consensus position paper attempts to define a non-exhaustive ensemble or suite of the most appropriate nutrition and health indicators for assessing the sustainability of diets, using the MD as a case study. Methods An International Working Group was informally developed in 2011 with the contribution of different national and international institutions – FAO; Sapienza University of Rome, Italy; University of Marseille, France; Hebrew University of Jerusalem, Israel; University of Las Palmas, Spain; Chouaib Doukkali University, El Jadida, Morocco; Centre International de Hautes Etudes Agronomiques Méditerranéennes (CIHEAM); International Inter-University Center for MEDITERRANEAN Food Culture Studies (CIISCAM); Council for Agricultural Research and Economics-Research Center on Food and Nutrition (formerly INRAN) (CRA-NUT), Rome, Italy; International Foundation of Mediterranean Diet (IFMED); Hellenic Health Foundation – Greece; and Forum on Mediterranean Food Cultures. Its purpose was to define the nutritional and health indicators relevant to assessing the sustainability of the diets, and in particular, of the MD. Through online and face-to-face brainstorming meetings, held from November 2011 to April 2015, a set of nutrition and health indicators of sustainability was identified. The definitions and the characteristics of the nutrition and health indicators were progressively refined based on the recursive comments of the participants (Table 1). Table 1 Milestones for the definition of nutritional indicators of sustainability of Mediterranean Diet. CIHEAM MAI–Bari International Workshop on “Guidelines for Improving the Sustainability of the Mediterranean Diet,” Bari, November 28–29, 2011 FAO/CIHEAM discussion paper on “Towards the Development of Guidelines for Improving the Sustainability of Diets and Food Consumption Patterns in the Mediterranean Area” (40, 45) CIHEAM/FAO Seminar on “Food Systems and Sustainable Diets: The Mediterranean Diet as a Pilot Study” in preparation of the ninth Meeting of CIHEAM Ministers of Agriculture, Malta, September 25–26, 2012 (46) CRA/FQH International Workshop on “Assessing Sustainable Diets within the Sustainability of Food Systems,” Rome, September 15, 2014; CIISCAM/Sapienza University of Rome, fourth Carlo Cannella Meeting, Rome, February 26, 2015. The process for the production of the present document began in the second half of 2014. The manuscript draft was then circulated among the participants of the different institutions. An agreement on the final version of the document was reached in April 2015 in Rome. Different instruments were used to have a comprehensive picture of the eating patterns (Table 2). Individual Dietary Surveys (IDS), Household Budget Surveys (HBS), and food balance sheets (FBS) vary in the methodology that leads to different levels of disaggregation and detail. Usually, for one food item (after alignment), the mathematical relation in relation to quantities is FBS > HBS > IDS, because FBS items are calculated excluding reuse and stock variation (national account budgets); they represent what food items are available per capita, but not obviously what is necessarily consumed. HBS do not include meals eaten outside the home but does include kitchen wastes and leftovers. IDS refer to edible part (excluding wastes and leftovers) of food and include eating at home and outside of home (47). Finally FBS overestimated food consumption and nutrient intake compared to IDS. Results between HBS and IDS are quite similar, except for fish, meat, pulses, and vegetables, which are underestimated by HBS, and sugar, honey, and cereals, which are overestimated. Table 2 Data sources and criteria used for the definition of nutritional indicators of Mediterranean Diet. – FBS and commodity balances provide data for domestic availability of a food, or food component in the case of protein. The contributing data include the sum of production and imports, with exports and non-food use subtracted. New modules to the FAOSTAT family of databases, including land use, emission, pesticide, fertilizer, and irrigation, will provide more data on environmental sustainability when analyzed with protein ratio data. Food consumption studies, national nutrition surveys, household budget surveys, etc., will be available in some countries to provide accurate individual data instead of FAO FBS data. It is noted that FBS data represent what food is available per capita on a national scale, but not what is actually eaten. – HBS are national surveys mainly focusing on consumption expenditure. They are conducted in all EU Member States and their primary aim (especially at national level) is to calculate weights for the consumer price index. They were launched in most EU Member States at the beginning of the 1960s, and Eurostat has been collating and publishing these survey data every 5 years since 1988. The two last collection rounds were 2005 and 2010. Although there have been continuous efforts toward harmonization, differences remain. – IDS is a class of methodologies including methods with various precision level (food record, 24-h recall, Food Frequency Questionnaire, dietary history, food propensity questionnaires, and combinations) usually not carried out at regular interval time except for some national reality as, e.g., the NHANES program in the USA (http://www.cdc.gov/nchs/nhanes.htm) or the NDNS (https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/310997/NDNS_Y1_to_4_UK_report_Executive_summary.pdf) in the UK. The European Food Safety Authority has launched the EU Menu (http://www.efsa.europa.eu/en/datexfoodcdb/datexeumenu.htm) program to push Member States of the European Union to harmonize dietary surveys. FBS and HBS for Europe are the source complying with the definition of indicators (see “Criteria for Selecting Indicators”). The worldwide data FBS1 and the European data HBS2 are current statistics. An overall view of HBS with the aim of harmonizing data codes for nutritional analysis can be found in the publication related to the DAFNE project3 (48), and the results are also published in the “European Nutrition and Health Report 2009”4 (49). Important information on Agri-environmental indicators (AEIs), HBS, and from farm to fork statistics may also be collected from Eurostat.5 Criteria for Selecting Indicators To select the most effective indicators, the following criteria were considered (41): Relevant to the question being asked. The indicator should be the best indicator currently available to answer the question; Understandable, i.e., clear, simple, and unambiguous; Graphically representable; Readily interpretable, i.e., clear, which direction the indicator should develop to lead to greater sustainability; Relevant in most Economic European Area Member and collaborating countries, i.e., not restricted to an issue, which is limited to a few member countries; Monitorable, i.e., based on data that are readily available in member and collaborating countries, or could be made available at reasonable cost–benefit ratio and with regularity within time frame of policy cycle (i.e., updated each year and with maximum 4-year time delay); Reliable and consistent, i.e., data collection and analysis methodologies should preferably be consistent from country to country, and at very least, be consistent within a given country from year to year; Representative, i.e., can be taken to represent current sustainable consumption and production trends within a given sector, final consumption cluster, etc. Results A set of nutrition indicators of sustainability was identified by the Working Group, and thirteen indicators, from A1 to A13, were finalized. Biochemical characteristics of food A1. Vegetable/animal protein consumption ratios A2. Average dietary energy adequacy A3. Dietary Energy Density Score A4. Nutrient density of diet Food quality A5. Fruit and vegetable consumption/intakes A6. Dietary Diversity Score Environment A7. Food biodiversity composition and consumption A8. Rate of local/regional foods and seasonality A9. Rate of eco-friendly food production and/or consumption Lifestyle A10. Physical activity/physical inactivity prevalence A11. Adherence to the Mediterranean dietary pattern Clinical aspects A12. Diet-related morbidity/mortality statistics A13. Nutritional anthropometry For each indicator, the following set of information is provided: definition, methodology, background, data sources, limitations of the indicator, and references. A1. Plant and Animal Protein Consumption Ratios Definition This indicator is a ratio of the relative intakes of protein from plant and animal sources, assessing adherence to an optimal dietary pattern, and a proxy for environmental impact of diets. Methodology Parameter considered is as follows: – ratio of plant (cereals, vegetables, pulses, fruit) and animal (meat, fish, eggs, dairy products) proteins in the diet using existing data. Adherence to an optimal ratio, including the MD, can be judged by simple comparison, and the trend can be monitored over the time series of available data, regardless of the data source. Data Sources FAOSTAT FBS and commodity balances provide data for domestic availability of a food, and food component in the case of protein (50). The contributing data include the sum of production and imports, with exports and non-food use subtracted. New modules to the FAOSTAT family of databases, including land use, emission, pesticide, fertilizer, and irrigation, will provide more data on environmental sustainability when analyzed with protein ratio data. Food consumption studies, National Nutrition Surveys, HBS, etc., will be available in some countries to augment or replace FAOSTAT data. Limitations of the Indicator FAOSTAT data from FBS and commodity balances reflect domestic availability of foods, not consumption or production per se. While these data have proven useful for assessing nutritional adequacy of diets, with a long history of use, they may significantly misrepresent sustainability issues. For example, livestock production has a greater role in GHG emission than livestock consumption. If meat is imported rather than domestically produced, the calculation of environmental impact may be skewed if using food balance or commodity balance datasets. Similarly, National Nutrition Surveys do not address the issue of production. Food losses and waste not accounted for in the datasets will affect the calculations and interpretation. Additionally, the advantages of using plant:animal protein ratio, as opposed to plant:animal dietary energy ratio or plant:animal ratio in grams per person per day, need to be elaborated (51, 52). A2. Average Dietary Energy Adequacy Definition The indicator expresses the dietary energy supply (DES) as a percentage of the average dietary energy requirement (ADER) in the country. Each country’s or region’s average supply of calories for food consumption is normalized by the ADER estimated for its population, to provide an index of adequacy of the food supply in terms of calories. This indicator was proposed by FAO as new approach for the measurement of food security (53). Methodology Parameters considered are as follows: – dietary energy supply (kilocalories/capita/day): average supply available for each individual in the total population (it does not indicate what is actually consumed by individuals); – average dietary energy requirement (kilocalories/capita/day): the amount of food energy needed to balance energy expenditure in order to maintain body size, body composition, and a level of necessary and desirable physical activity consistent with long-term good health. Data Sources Data can be downloaded from http://www.fao.org/economic/ess/ess-fs/ess-fadata/en/#.VPhu3y7K1i0. Otherwise, DES can be obtained from National IDS, HBS, FBS, and ADER from National Energy Requirements, FAO Human Energy Requirements. Limitations of the Indicator National IDS using food diaries or dietary recalls estimate the actual consumption (i.e., the dietary energy intake), provide the best evidence on food consumption, and constitute the best method for assessing energy intake, and more generally, dietary patterns and evaluating diet–disease associations. Being expensive and labor-intensive, these surveys are undertaken only in a limited number of countries, often at regional or local level or in specific population groups; furthermore, it is difficult to accomplish comparability at the international level, because the assessment methods are variable, self-reported, and consequently subject to considerable measurement errors. In order to overcome these problems, the per capita DES can be used instead of the dietary energy intake. Data on DES can be gathered from FBS and HBS. FBS and HBS overestimate energy intake, although not in a linear way; while FBS includes eating out, HBS does not; food losses and waste should be considered. In order to reduce the impact of possible errors in estimated DES, due to the difficulties in properly accounting of stock variations in major food, the indicator is calculated as an average over 3 years (for example, 2010–2012, 2011–2013, and 2012–2014). Average dietary energy requirement is a reference for adequate nutrition in the population. The recommended level of dietary energy intake for a population group is the mean energy requirement of the healthy, well-nourished individuals who constitute that group. The estimates of requirements derived from measurements of a collection of individuals of the same age, gender, body size, presumed body composition, and physical activity are grouped to give the average energy requirements for a class of people or a population group. These requirements are used together to predict the requirements of other individuals with similar characteristics, but on whom measurements have not been made. Consequently, application of these results to any one individual for clinical or other purposes may lead to errors of diagnosis and improper management (54). On the other hand, data on the size of consumed food portion that influence energy intake need to be evaluated (55), as positive relationships between portion size and energy intake have been demonstrated in adults (56–58). A3. Dietary Energy Density Score Definition This indicator measures the amount of energy (kcal or kJ) in a given weight (g, 100 g, or kg) of diet as a proxy for healthy dietary patterns. Methodology Parameter considered is as follows: – dietary energy density (kilocalories/gram) calculated by dividing total dietary energy by the edible weight of foods and caloric beverages consumed. The primary data are as follows: the mean amounts of various foods/beverages or food groups consumed daily. the energy provided by weight unit of the foods/beverages or food groups as provided by food composition databases. It will be expressed as the amount of energy (kilocalories or kilojoules) in 100 g or 1 k of daily diet. Data Sources National IDS, HBS, and FAO FBS. Limitations of the Indicator Individual Dietary Surveys provide the most accurate figures for actual daily food consumption. Data obtained from FBS do not reflect the effective food intake, because they relate to the food quantities theoretically available for consumption; the amount of food consumed is lower than those reported in FBS, due to the degree of losses of edible food and nutrients in the household, e.g., during storage, in preparation, and cooking, as plate waste or quantities fed to domestic animals and pets, or thrown away. Depending on data sources and studies, the level of accuracy and units used can vary. Also, the data obtained even from National Dietary Surveys do not reflect the portion size. Indeed, there is evidence that larger portion size of energy-dense and nutrient-poor foods is involved in the increase of overweight and obesity that accompany the changes in dietary patterns in children and adults (57, 58). A4. Nutrient Density of Diet and Foods Definition The nutrient density of a composite diet is the amount of various necessary nutrients and fibers present in a given daily diet expressed in weight (or evenly in energy). Methodology Parameters considered are as follows: – a daily diet: mean adequacy ratio (MAR) based on the mean percentage of the recommended intakes for 29 key needed nutrients, alone or in combination with the mean excess ratio (MER) for nutrients to limit. – general purpose: the amount of every nutrient present in a unit of a given food/beverage or food group as provided by food composition databases; Nutrient Density Scores referring to either 100 g, 100 kcal/kJ, or cost/kg or L of a given food: ex simplified SAIN/LIM scoring (Score d’Adéquation Individuel aux Recommandations Nutritionnelles – SAIN; nutrients to be limited – LIM) (10). Some publications cited in the reference list provide examples of calculations and interpretations (12, 59). Data Sources National IDS, HBS, and FAO FBS (see indicator A2 for a comprehensive description). Limitations of the Indicator All scores designed to evaluate the nutrient density of either individual foods or whole diet have advantages and limitations. They must be taken into account depending on the precise context and objective considered. The limitations are (i) the need for accurate and quantitative dietary intake data and food composition databases; (ii) comparisons between countries are limited by possibly different daily recommended intakes (energy, nutrients, and fiber); and (iii) comparisons between studies need the use of the same nutrients and total number of nutrients. It has to be considered that the MAR normally should be 29, but because of the lack of composition tables, the number is usually less. In France, for example, 23 includes the different lipids. A5. Fruit and Vegetable Consumption/Intakes Definition This indicator is the measure of the consumption (grams/capita/day) of fruits and vegetables, including pulses, nuts, and seeds (12), directly applicable to assessing adherence to MD, and as a proxy for a healthy diet and specific micronutrient intakes. Methodology Parameter considered is as follows: – measure of the consumption (supply, availability, intake) of fruits and vegetables (grams/capita/day), including pulses, nuts, and seeds. Data Sources National IDS, HBS, and FAO FBS. Limitations of the Indicator Data obtained from FBS do not reflect the effective food intake, because they relate to the food quantities available to the consumer (but not necessarily consumed). Thus, the amount of food consumed is usually lower than those reported in FBS (59), due to the degree of losses of edible food and nutrients in the household/catering, e.g., during storage, in preparation, and cooking, as plate waste or quantities fed to domestic animals and pets, or thrown away. However, when National IDS are not available, the HBS and/or FBS provide good indicators by which to compare several countries and different time periods. Although it is not specified in official documents, considering the high proportion of waste often present in preparations of plant foods, it should be specified that the weight of 400 g daily refers to edible product net of waste (12, 60). When using national supply data, the reference value could be increased to take into account that goods include inedible parts. Moreover, 500 g per day are also recommended in some dietary guidelines6 and for ischemic heart disease prevention (61, 62). A6. Dietary Diversity Score Definition Dietary diversity is a qualitative measure of the household access and consumption of a wide variety of foods. It is an indicator that reflects the households’ diet quality and is also a proxy for the adequacy of nutrient intake of the diet for individuals (63–65). This concept is based on the fact that the needs in nutrients are not covered by a single food, but by a diet composed of several foods. It is associated with household socioeconomic status and food security (energy availability at the household level). A greater dietary diversity was also reported to protect different households against the double burden of malnutrition known in countries in nutrition transition (66–68). Methodology Parameters considered are as follows: – Dietary Diversity Scores (DDS) defined as the number of food groups consumed over a reference period: Individual Dietary Diversity Score (IDDS) used as proxy of the nutritional quality of individual diet has for aim to assess the adequacy of nutrient intake; Dietary Diversity Score at the household level (HDDS) is used, on the other hand, as proxy of the socioeconomic level of the household and intends to reflect the economic ability of a household to consume a variety of foods. – Dietary Variety Score (DVS) (69) corresponding to the number of foods consumed among a list of foods. – US Healthy Food Diversity index (70), a tool for the simultaneous measurement of dietary variety, quality, and proportionality at individual level. Data Sources Usually, specific questionnaires are administered. The use of National IDS, HBS, and FAO FBS need to be experimented. Limitations of the Indicator Even if there is a preference for the DDS based on food groups, the issue of the number and the choice of these food groups has not yet been resolved. The selection of food groups can be guided by the objectives for which the scores are used. For example, if the score of diversity is used to identify populations at risk of micronutrient deficiency, the classification used should distinguish food groups depending on their content in micronutrients. In this case, it is obvious that comparisons between studies or countries are more difficult. Moreover, it has to be considered that the diversity scores have been designed specifically for developing countries without regularly carried out national statistics about this topic. A7. Food Biodiversity Composition and Consumption Definition Biodiversity covers diversity within species, between species, and of ecosystems; synonyms are biological diversity and ecological diversity. For the purposes of human nutrition, biodiversity refers to foods identified at the taxonomic level below species (e.g., cultivar, breed) or by local varietal name, and wild, neglected, and/or underutilized species. Biodiversity is distinctly different from “dietary diversity,” which reflects intake at the level of aggregate food groups. Methodology Parameters considered are as follows: – food composition: a count of the number of foods: at variety/cultivar/breed level for common foods with at least one value for component found in published and unpublished sources. at species level for wild/indigenous/underutilized foods with at least one value for component found in published and unpublished sources. – food Consumption: the taxonomic diversity of foods, as for food composition, reported in food consumption/dietary intake surveys. Data collected and reported include: ▪ the study instrument (e.g., diet history, food frequency) with details (scope, date, number, and description of subjects, geographical/ethnic coverage; reference, total number of studies examined); ▪ the qualifying biodiverse foods reported (number of foods, food lists); the number of surveys with at least one reported food counting for biodiversity. Data Sources Food and Agriculture Organization of the United Nations/International Network of Food Data Systems (INFOODS) compile data and report periodically (71). For food composition, data are obtained by searching peer-reviewed journals using the search engines Scopus and Science Direct, and through a call for data conducted via INFOODS. These data are then compiled in a Biodiversity Food Composition Database (72). For food consumption, data are obtained from all surveys, including National Nutrition Surveys, market surveys, ethnobiological investigations, and inventory studies. All published and unpublished available resources are searched, including peer-reviewed journals, official international/regional/national/subnational survey reports, conference presentations, and published matter, including posters, abstracts published from meetings, and theses. Limitations of the Indicators The development and reporting on the indicators are recent, and only two to three time points are available. The usefulness of the indicators should be assessed in the future and judged against market survey data as well as nutritional outcomes. For the moment, the results represent a reflection of the attention being paid to biodiversity by researchers designing food composition studies and dietary surveys. Monitoring and reporting on the biodiversity indicators is the responsibility of FAO/INFOODS. It is a time-consuming activity, and, for the 2014–2015 biennium, FAO has put few or no resources into the continuation of this effort. A8. Local/Regional Foods and Seasonality Definition The term “local food system” (or “regional food system”) is used to describe a method of food production and distribution that is geographically localized, rather than national and/or international. Food is grown (or raised) and harvested close to consumers’ homes, then distributed over much shorter distances than is usual in the conventional globalized industrial food system with long-distance transportation. In general, local/regional food systems are associated with the sustainable agriculture concept, but not systematically. In particular, it is based on purchases at short distance from the producer (from few to 100 km or miles) and directly from the producer or with one intermediate between the consumer and producer. Production “in season” means that minimum artificial conditions are used to grow the products (essentially plant products: vegetables and fruits), without heated greenhouses in the local agro-environmental conditions and no long-term cold storage). Methodology Parameters considered are as follows: – the distance between consumer purchase location and producing area; it is usually considered that it should be at maximum 150 km (around 100 miles). – the number of intermediates between producer and consumer with zero when direct from producer, one when one intermediate is present (one can be considered as a cut point for discrimination). – the consumer choice: directly to local/regional producers (on-farm, farmer’s market/shop, food baskets made of local foods) as a share of total food purchases, share of fresh vegetables or fruits consumed coming from open field or unheated greenhouse cultivation. – the duration between fruit harvest (known or estimated from agriculture statistics of the concerned growing location or country) and purchase of fresh fruit, as a direct reflect of distance from seasonal production (and cold storage duration). Data Sources The information necessary to assess these indicators can be only obtained from dedicated studies where such specific questions are addressed. It is the case in some national human cohort surveys or more local/regional consumption studies. The growing interest in such consumption approaches will stimulate more investigations in this domain. Limitations The parameters to use are still under debate and need further testing. The present availability of data can be restricted to a limited number of studies, but this figure is expected to markedly improve in the next future. A9. Organic/Eco-friendly Production and Consumption Definition Nowadays, most agro-food productions based on agro-ecological principles are called as “organic” and are certified and labeled at national and continental levels (73). These well-characterized, controlled, and certified methods of food productions exclude the use of chemical fertilizers, pesticides, GMOs, and intensive animal husbandry (74–79). These methods are acknowledged to better protect environment, biodiversity, and potentially, health consumers. Methodology Parameters considered are as follows: – the percentage of consumers buying organic foods and the frequency of consumption. – organic food consumption in percentage of total food amount or money per capita (e.g., Bionutrinet Cohort Survey in France7). – the percentage of the organic market volume. – the percentage of land use under organic certification. Data Sources In most industrialized countries, data on the organic market volume as well as the market shares are available as well as recorded (73). Detailed data for specific food types can be available too. During some consumer cohort surveys or in national consumption surveys, individual data are collected on organic food consumption (e.g., Germany, France). Furthermore, national yearly data are now available and continuously recorded (73) regarding the importance of organic food production (number of farms, acreage, volume of foods produced) and share of total. In some countries, agricultural production data (at local/regional/national) are also available along with organic food import/export data (73). Limitations of the Indicator In some countries, organic production can be marginal only or data on organic production or consumption are not available at national or regional level. But, the availability of data has been and will be increasing (73). A10. Physical Activity/Physical Inactivity Prevalence Definition As physical activity is a key determinant of energy expenditure, it is fundamental to energy balance and weight control. Although there are doubts on considering it as a nutritional indicator or a cofactor of nutritional status, the Working Group decides to consider it in the list of nutritional indicators of sustainability. Anyway, it is important to underline that MD, and more in general, diet needs to be considered a lifestyle, and the regular practice of physical activity is a key component of it. Physical activity is defined as any bodily movement produced by skeletal muscles that require energy expenditure. The term “physical activity” should not be mistaken with “exercise.” Exercise, is a subcategory of physical activity that is planned, structured, repetitive, and purposeful in the sense that the improvement or maintenance of one or more components of physical fitness is the objective. Physical activity includes exercise as well as other activities, which involve bodily movement and are done as part of working, active transportation, house chores, and recreational activities (80). Several physical activity indicators have been proposed (81). On the basis of available data, the physical inactivity prevalence has been selected as an indicator of physical activity, using the definition of not meeting any of the following criteria: at least 30 min of moderate-intensity activity per day on at least 5 days per week, or at least 20 min of vigorous-intensity activity per day on at least 3 days per week, or an equivalent combination (82). Methodology Parameters considered are as follows: – Attributable disability-adjusted life years (DALYs) from physical inactivity; – Physical Activity Questionnaires [e.g., WHO Global Physical Activity Questionnaire (GPAQ); International Physical Activity Questionnaire (IPAQ), etc.]. Data Sources National surveys and WHO Global Infobase. Limitations of Indicator It is difficult to use questionnaires across that are comparable across cultures. All the questionnaires dealing with physical activity presents some limitations in particular, considering the shorter-forms and the versions to be used without personal interview (83). Moreover, data on population-based physical inactivity may be limited in some countries. As an indicator should be, by the way, “monitorable,” it should be based on the data that are readily available, but most of the available data may be difficult to interpret due to differences in the way physical inactivity is measured. The use of objective methods, such as pedometers, is becoming more feasible, especially with the use of mobile technology and apps that provide such information. However, at the moment, for surveillance activity, the objective methods were rarely used, and there are not available data. Moreover, pedometer is specifically designed to assess walking only; it is enabled to record non-locomotor movements and to examine the rate or intensity of movement. On the other hand, accelerometer is suitable for all populations and is an objective indicator of body movement (acceleration) but is an inaccurate assessment of a large range of activities, and the financial cost may prohibit assessment of large numbers of participants. Finally, considering the cost–benefit ratio, the Working Group suggests to promote the collection of using harmonized methodologies, such as the same physical activity questionnaire, in all countries (i.e., GPAQ) supported by pedometer or accelerometer. A11. Adherence to the Mediterranean Dietary Pattern Definition Adherence to the traditional MD, or, to diets that resemble the Mediterranean pattern, have been expressed through indexes or scores, defined a priori, which operate by combining conceptually and computationally the dietary components that capture the essence of this dietary pattern. Methodology Parameter considered is as follows: – Mediterranean Diet Score (MDS) (it ranges from 0 – minimal adherence to the traditional MD – to 9 – maximal adherence): for the five components, which are representative of the MD and are presumed to be consumed in large quantities (vegetables, legumes, fruits and nuts, cereal, and fish), a value of 1 is assigned to persons whose consumption is at or above the components’ sex-specific medians based on the considered sample, and 0 otherwise; a sixth component of the score is the ratio of monounsaturated lipids to saturated lipids, reflecting the principal role of olive oil consumption in the traditional MD: a value of 1 is assigned to persons whose consumption is at or above the sample-specific median of this lipid ratio and 0 otherwise; for the following two components that are presumed to be consumed in low/moderate quantities in the MD (all meats and all dairy products, which are rarely non-fat or low-fat in Mediterranean countries), persons whose consumption is below the median are granted with a value of 1, and persons whose consumption is at or above the median are penalized with a value of 0; for alcohol, a value of 1 is assigned to men who consume between 10 and 50 g of ethanol per day and to women who consume between 5 and 25 g per day, expressing the moderate ethanol consumption in the MD. Data Sources The MD indexes were estimated in their majority from information collected through detailed Food Frequency Questionnaires (FFQ) or repeated measures of 24-h recall dietary questionnaires, which are not easy to be dealt with, especially from the general public. Limitations The previous-indicated approach has been very valuable in order to express the whole of a dietary pattern, and specifically of MD. The limitation of the approach is that usually cutoff points used in most scores are sample-dependent, making the interpretation of any identified association of this pattern with health outcomes difficult to generalize. Second, since many MD indexes exist, a natural question is whether some work better than others with respect to capturing the adherence to MD, as well as, to identifying associations of this diet with a specific health outcome. However, to decide which of these numerous MD indexes is “optimal” is rather difficult, since such a decision would require one to evaluate the predictive ability of the various indexes with respect to different outcomes using one population, and then validate the results to different populations. The issue becomes even more complicated due to the population-specific and not universal cutoff values that have been used for discriminating the low/high consumptions for each of the MD components, as described previously. Notwithstanding the scientific value of such an approach, any “optimal” MD scale should also be characterized by its simplicity in the construction of the index as well as in the use of this index widely in public health as well as in clinical practice. Such an investigation would be very important for future studies, which wish to assess the association of MD with health. A12. Diet-Related Morbidity/Mortality Statistics Definition This indicator monitors mortality and morbidity (occurrence of cardiovascular events, type II diabetes, dyslipidemia, hypertension, osteoporosis, neurodegenerative diseases, and some types of cancer) as a proxy for the consumption of healthy diets. Methodology Parameters considered are as follows: – prevalence of individuals having physician-diagnosed obesity, cardiovascular diseases (CHD, stroke, and hypertension), type II diabetes, osteoporosis, neurodegenerative diseases, and obesity-related cancers. – disability-adjusted life year as a measure of overall disease burden expressed of years lost due to illness, disability or early death associated with nutrition-related factors: high blood pressure, high cholesterol (total and LDL), high blood sugar (insulin resistance and/or diabetes). Data Sources National surveys and WHO World Health Statistics (84, 85). Limitations of the Indicator Some pathologies can be undiagnosed or underreported in some countries. Data may not be available for the same age groups. If data are not available, mortality prevalence will be used. A13. Nutritional Anthropometry Definition This indicator is based on the body mass index (BMI) and the waist circumference (WC), which are used in a wide variety of contexts as simple methods to assess how much an individual’s body weight departs from what is normal or desirable for a person of his or her height (86). Body mass index in both men and women represents a measure of underweight (<18.5 kg/m2) or different levels of overweight (25–29.9, 30–34.9, 35–39.9, ≥40 kg/m2). Waist circumference in both men and women represents a measure of visceral adiposity (>88 cm in women and 102 cm in men). Increased WC can be a marker for increased risk, even in persons of normal weight associated with insulin resistance. Overnutrition and undernutrition frequently coexist. Weight loss, real to ideal weight ratio, and specific nutritional assessment tools (Mini Nutritional Assessment – MNA, Just a Nutritional Screening – JANUS) (87, 88) may be useful to detect the presence of malnutrition. Methodology Parameters considered are as follows: – undernutrition: prevalence of individuals having a BMI <18.5 kg/m2 calculated from self-reported weight and height; – overweight or obesity: prevalence of individuals having a BMI ≥25.0 kg/m2 calculated from self-reported weight and height and/or WC >88 cm in women and 102 cm in men. Classification of overweight and obesity by BMI, waist circumference, and associated disease risks. Disease riska relative to normal weight and waist circumference BMI (kg/m2) Obesity class Men 102 cm or less Women 88 cm or less Men >102 cm Women >88 cm Underweight <18.5 – – Normal 18.5–24.9 – – Overweight 25.0–29.9 Increased High Obesity 30.0–34.9 I High Very high 35.0–39.9 II Very high Very high Extreme obesity 40.0+ III Extremely high Extremely high aDisease risk for type 2 diabetes, hypertension, and CVD (http://www.nhlbi.nih.gov/health/public/heart/obesity/lose_wt/bmi_dis.htm). Data Sources WHO Global Database and data locally available through National surveys (89–93). Limitations of Indicator Individuals tend to overestimate their height and underestimate their weight, leading to underestimation of BMI and of the prevalence of overweight and obesity. Moreover, anthropometric measurements have to be performed by skilled personnel according to a standardized procedure. Self-reported national surveys might be subject to systematic error (lower reported weight and higher reported height) resulting from non-coverage (e.g., lower telephone coverage among populations of low socioeconomic status), non-response (e.g., refusal to participate in the survey or to answer specific questions), or measurement (e.g., social desirability or recall bias). Data could not be available for some countries. It is hoped that data of BMI less than 18.5 kg/m2 will be available soon from the WHO to measure the rates of undernutrition in these populations. Discussion This paper has identified and summarized some of the most relevant nutritional indicators to measure the sustainability of food consumption. The purpose is, together with additional indicators for the other three sustainability dimensions (environment, economic, and sociocultural), to formulate recommendations for cross-sectoral policy instruments, allowing the comparability and improvement of the sustainability of the diets and food systems in different countries. The selection and analysis of each indicator has been done (wherever possible) with specific reference to the MD as a case study. The sustainability of food systems represents an urgent action area for governments and international organizations to tackle the serious socioeconomic and environmental consequences of short-sighted behaviors and practices involving agricultural land and rural communities (94). It requires developing a set of comprehensive, coherent, integrated, and holistic policies that simultaneously consider the relative priorities (and trade-offs) between different sectors: nutrition, health, lifestyle, society, culture, economy, environment, and agro-biodiversity. The proposed nutrition indicators outlined in this paper will be useful for further developing a methodological framework for designing policies in order, not only to conserve and preserve the traditional diet, such as the MD, as a common cultural heritage and lifestyle but also to enhance the sustainability of dietary models. The MD, in its various national forms, may be used as a case study: a model to describe, understand, and improve the sustainability of current food consumption because of the high and increasing pressure on its fragile natural resources exacerbated by the changes of Mediterranean food consumption patterns (15, 95). A medium term research and action framework needs to be implemented to analyze the sustainability of the diets in the Mediterranean area (40, 45). The use of the selected indicators and their validation may represent a first step of a “pilot sustainability laboratory” aimed at the definition of a validated procedure that will help governments and policy makers to formulate sustainability-sensitive policies in the promotion of sustainable food systems development in different areas. The Mediterranean area can be considered as a case study because of its passage through a “nutritional transition” in which problems of undernutrition coexist with overweight, obesity, and food-related chronic diseases (37). Undernutrition is still significant in the South of the Mediterranean: 9.2 million people in 2001–03, 3.9% of the population of the zone, compared with 7.3 million people in 1990–92, 3.8% of the population (96). The rate of stunting among children less than 5 years of age is also very high in many countries in the South: 18% in Algeria, 21% in Egypt, 12% in Lebanon, 24% in Morocco, 12% in Tunisia, and 16% in Turkey. At the same time, according to WHO, overweight and obesity rates in Mediterranean countries continue to rise. Currently reported rates for overweight and obesity range, respectively, from 45.5 and 16.0% in Algeria to 67.9 and 33.1% in Egypt (51). The indicators that were selected can be attributed to five domains related to nutritional aspects of the diet: biochemical quality of food (A3. Vegetable/animal protein consumption ratios; A4. Average dietary energy adequacy; A6. Dietary Energy Density Score; A7. Nutrient density of diet); food quality (A2. Fruit and vegetable consumption/intakes; A5. Dietary Diversity Score); environment (A8. Food biodiversity composition and consumption; A12. Rate of Local/regional foods and seasonality; A13. Rate of eco-friendly food production and/or consumption); lifestyle (A10. Physical activity/Physical inactivity prevalence; A11. Adherence to the Mediterranean dietary pattern), and clinical aspects (A1. Diet-related morbidity/mortality statistics; A9. Nutritional Anthropometry). The choice of the indicators is indeed a compromise between what is desirable and what is practical and available in which countries. In one sense, these indicators represent an “ideal” list; it remains to be seen how useful they are for practical application. In this paper, we refer to databases: FAOSTAT, HBS, EUROSTAT, etc. The quality of the data can be highly variable, and this may represent a limit of the procedure. Anyhow, there is no means to avoid these difficulties, and the presence of different databases may minimize, at least partly, the effects of the deficiencies of each single database. Thus, the next phase of this work will be: – Collecting data sets from individual countries to ascertain what figures are available for analysis. – The validation of the indicators that were selected by the expert group. Thus these indicators will be performed versus sustainability or versus an outcome variable that can be considered a proxy of it. – The definition of a mathematical model that will be able to combine all the indicators belonging to a single area (e.g., nutritional indicator). In this phase, it will be necessary to verify that all the indicators add some new information to the model, to attribute to each indicator a weight, to avoid collinearity (when a variable can be linearly predicted from the others, it has to be omitted respecting the lex parsimoniae). – Following these steps, we might consider organizing the groups of indicators into a composite index to quantify and monitor sustainability over time. The methodology has been established as a two stage approach to determine first, within each dimension, the relative weightings of the indicators selected and then the weightings between each dimension to enable building a composite index, which may be easily disaggregated into the four dimensions of sustainable diets (45, 97). This methodological approach is sufficiently flexible to allow modifying the type and number of indicators in each dimension as new data accrue. Adherence Recent surveys show that many countries in the Mediterranean area are drifting away from the traditional MD healthy pattern, and current food consumption habits show a decline in their adherence to the MD (26, 31–34, 36). Because of such waning in adherence to the MD, there are major concerns, including health risks [due to an increase in the consumption of lipids (e.g., meat, dairy products, etc.), an increase in the consumption of processed foods, simple carbohydrates (e.g., beverages and foodstuffs with a high carbohydrate content), and a decrease in the consumption of complex carbohydrates (e.g., cereals and legumes) leading to chronic nutrition-related diseases, disability, and increased mortality]; environmental issues (due to an exacerbate ecological footprint as a consequence of a more prevalent consumption of foods from animal sources), and loss of biodiversity (due to the globalization of food production/consumption and the homogenization of eating patterns; the dietary diversity is linked to nutrient composition diversity between foods and among varieties/cultivars/breeds of the same food; dietary diversity may guarantee healthy diet through an adequate presence of nutrients and bioactive molecules) (98). Also, in a time of abundance, larger portion sizes in terms of energy, nutrient, and diversity lead to unhealthy diets, diabesity, and diet-associated morbidity/mortality. Studies on these trends in the Mediterranean area can be useful in promoting more efficacious interventions to preserve healthy eating patterns, to reduce the environmental impact of food production, and to conserve biodiversity. Conclusion The Working Group has selected thirteen indicators from a broader framework of indicators. In this wider group were listed some other indicators, such as food composition, frugality, household food security, level of food processing, nutrient profile, global nutritional index, food losses, and waste, that could be considered a proxy of those selected or overlapping, at least in part, with them. The Working Group will attempt to consider them in the future. Author Contributions All authors listed have made substantial, direct, and intellectual contribution to the work and approved it for publication. Conflict of Interest Statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. 1http://faostat3.fao.org/browse/FB/*/E 2http://gesis.org 3http://www.nut.uoa.gr/dafneENG.html 4http://www.european-nutrition.org/images/uploads/pub-pdfs/European_Nutrition_and_Health_Report_2009.pdf 5http://ec.europa.eu/eurostat/web/main/home 6e.g., http://www.fao.org/nutrition/education/food-based-dietary-guidelines/regions/countries/finland/en/ 7http://bionutrinet.etude-nutrinet-sante.fr Abbreviations ADER, average dietary energy requirement; BMI, body mass index; CIHEAM, Centre International de Hautes Etudes Agronomiques Méditerranéennes; CIISCAM, International Inter-University Studies Centre on Mediterranean food culture; DALY, disability-adjusted life years; DDS, Dietary Diversity Score; DES, dietary energy supply; DGA, dietary guidelines for Americans; DVS, Dietary Variety Score; EUROSTAT, statistical office of the European Union; FAO, Food and Agriculture Organization of the United Nations; FBDG, food-based dietary guidelines; FBS, food balance sheets; FFQ, Food Frequency Questionnaire; GHG, green house gas; GPAQ, Global Physical Activity Questionnaire; HBS, Household Budget Surveys; HDDS, Dietary Diversity Score at the household level; IDDS, Individual Dietary Diversity Score; IDS, Individual Dietary Surveys; IFMED, International Foundation of Mediterranean diet; INRAN, Italian National Institute of Food and Nutrition; IPAQ, International Physical Activity Questionnaire; JaNUS, just a nutritional screening; LIM, nutrients to be LIMited; MAR, mean adequacy ratio; MDS, Mediterranean Diet Score; MED, Mediterranean diet; MER, mean excess ratio; MNA, mini nutritional assessment; NDS, Nutrient Density Score; SAIN, Score d’Adéquation Individuel aux recommandations Nutritionnelles; UNESCO, United Nations Educational, Scientific and Cultural Organization; WC, waist circumference; WHO, World Health Organization. ==== Refs References 1 Tilman D Clark M . Global diets link environmental sustainability and human health . Nature (2014 ) 515 :518 –22 .10.1038/nature13959 25383533 2 Heller MC Keoleian GA Willett WC . Toward a life cycle-based, diet-level framework for food environmental impact and nutritional quality assessment: a critical review . Environ Sci Technol (2013 ) 47 (22 ):12632 –47 .10.1021/es4025113 24152032 3 O’Kane G . What is the real cost of our food? Implications for the environment, society and public health nutrition . Public Health Nutr (2012 ) 15 (2 ):268 –76 .10.1017/S136898001100142X 21733281 4 Clonan A Holdsworth M The challenges of eating a healthy and sustainable diet . Am J Clin Nutr (2012 ) 96 (3 ):459 –60 .10.3945/ajcn.112.044487 22875711 5 Berry EM Dernini S Burlingame B Meybeck A Conforti P . Food security and sustainability: can one exist without the other? Public Health Nutr (2015 ) 18 (13 ):2293 –302 .10.1017/S136898001500021X 25684016 6 FAO/Bioversity . Sustainable Diets and Bioversity. Directions and Solutions for Policy, Research and Action . Rome (2012 ). Available from: www.fao.org/docrep/016/i3004e/i3004e00.htm 7 Kearney J . Food consumption trends and drivers . Philos Trans R Soc Lond B Biol Sci (2010 ) 365 (1554 ):2793 –807 .10.1098/rstb.2010.0149 20713385 8 Reisch L Eberle U Lorek S Sustainable food consumption: an overview of contemporary issues and policies . Sust Sci Pract Policy (2013 ) 9 (2 ):7 –25 . 9 Gussow JD Clancy KL Dietary guidelines for sustainability . J Nutr Educ (1986 ) 18 :1 –5 .10.1016/S0022-3182(86)80255-2 10 Masset G Soler LG Vieux F Darmon N . Identifying sustainable foods: the relationship between environmental impact, nutritional quality, and prices of foods representative of the French diet . J Acad Nutr Diet (2014 ) 114 (6 ):862 –9 .10.1016/j.jand.2014.02.002 24703928 11 Macdiarmid JI Loe J Kyle J McNeill G . “It was an education in portion size”. Experience of eating a healthy diet and barriers to long term dietary change . Appetite (2013 ) 71 :411 –9 .10.1016/j.appet.2013.09.012 24076020 12 Vieux F Soler LG Touazi D Darmon N . High nutritional quality is not associated with low greenhouse gas emissions in self-selected diets of French adults . Am J Clin Nutr (2013 ) 97 (3 ):569 –83 .10.3945/ajcn.112.035105 23364012 13 Garnett T Appleby MC Balmford A Bateman IJ Benton TG Bloomer P Agriculture. Sustainable intensification in agriculture: premises and policies . Science (2013 ) 341 (6141 ):33 –4 .10.1126/science.123448 23828927 14 Vieux F Darmon N Touazi D Soler LG Greenhouse gas emissions of self-selected individual diets in France: changing the diet structure or consuming less? Ecol Econ (2012 ) 75 :91 –101 .10.1016/j.ecolecon.2012.01.003 15 Sáez-ciheam S Obrador B Bach-Faig A Serra-Majem LL . Environmental footprints of Mediterranean versus Western dietary patterns: beyond the health benefits of the Mediterranean diet . Environ Health (2013 ) 12 :118 .10.1186/1476-069X-12-118 24378069 16 Baroni L Cenci L Tettamanti M Berati M . Evaluating the environmental impact of various dietary patterns combined with different food production systems . Eur J Clin Nutr (2007 ) 61 :279 –86 .10.1038/sj.ejcn.1602522 17035955 17 Duchin F Sustainable consumption of food: a framework for analyzing scenarios about changes in diets . J Ind Ecol (2005 ) 9 :99 –114 .10.1162/1088198054084707 18 Gussow JD . Mediterranean diets: are they environmentally responsible? Am J Clin Nutr (1995 ) 61 :1383S –9S .7754992 19 Keys A A practical, palatable and prudent way of eating . J Med Assoc Ga (1970 ) 59 (9 ):355 –9 .5452772 20 Keys A Coronary heart disease – the global picture . Atherosclerosis (1975 ) 22 (2 ):149 –92 .10.1016/0021-9150(75)90001-5 1103902 21 Keys A Coronary heart disease, serum cholesterol, and the diet . Acta Med Scand (1980 ) 207 (3 ):153 –60 .10.1111/j.0954-6820.1980.tb09697.x 7368981 22 Willett WC Sacks F Trichopoulou A Drescher G Ferro-Luzzi A Helsing E Mediterranean diet pyramid: a cultural model for healthy eating . Am J Clin Nutr (1995 ) 61 (6 Suppl ):1402S –6S .7754995 23 Nestle M . Mediterranean diets: historical and research overview . Am J Clin Nutr (1995 ) 61 (6 Suppl ):1313S –20S .7754981 24 Serra-Majem L Roman B Estruch R . Scientific evidence of interventions using the Mediterranean diet: a systematic review . Nutr Rev (2006 ) 64 (2 Pt 2 ):S27 –47 .10.1301/nr.2006.feb.S27-S47 16532897 25 Sofi F Cesari F Abbate R Gensini GF Casini A . Adherence to Mediterranean diet and health status: meta-analysis . BMJ (2008 ) 337 :a1344 .10.1136/bmj.a1344 18786971 26 Bach-Faig A Berry EM Lairon D Reguant J Trichopoulou A Dernini S Mediterranean Diet Foundation Expert Group: Mediterranean diet pyramid today. Science and cultural updates . Public Health Nutr (2011 ) 14 :2274 –84 .10.1017/S1368980011002515 22166184 27 Estruch R Ros E Martínez-González MA Mediterranean diet for primary prevention of cardiovascular disease . N Engl J Med (2013 ) 369 (7 ):676 –7 .10.1056/NEJMc1306659 23944307 28 Trichopoulou A Kouris-Blazos A Wahlqvist ML Gnardellis C Lagiou P Polychronopoulos E Diet and overall survival in elderly people . BMJ (1995 ) 311 (7018 ):1457 –60 .10.1136/bmj.311.7018.1457 8520331 29 Trichopoulou A Costacou T Bamia C Trichopoulos D Adherence to a Mediterranean diet and survival in a Greek population . N Engl J Med (2003 ) 348 (26 ):2599 –608 .10.1056/NEJMoa025039 12826634 30 Trichopoulou A Bamia C Trichopoulos D . Anatomy of health effects of Mediterranean diet: Greek EPIC prospective cohort study . BMJ (2009 ) 338 :b2337 .10.1136/bmj.b2337 19549997 31 León-Muñoz LM Guallar-Castillón P Graciani A López-García E Mesas AE Aguilera MT Adherence to the Mediterranean diet pattern has declined in Spanish adults . J Nutr (2012 ) 142 (10 ):1843 –50 .10.3945/jn.112.164616 22875552 32 Da Silva R Bach-Faig A Raidó Quintana B Buckland G Vaz de Almeida MD Serra-Majem L Worldwide variation of adherence to the Mediterranean diet, in 1961–1965 and 2000–2003 . Public Health Nutr (2009 ) 12 :1676 –84 .10.1017/S1368980009990541 19689839 33 Vareiro D Bach-Faig A Raidó Quintana B Bertomeu I Buckland G Vaz de Almeida MD Availability of Mediterranean and non-Mediterranean foods during the last four decades: comparison of several geographical areas . Public Health Nutr (2009 ) 12 (9A ):1667 –75 .10.1017/S136898000999053X 19689838 34 Garcia-Closas R Berenguer A González CA . Changes in food supply in Mediterranean countries from 1961 to 2001 . Public Health Nutr (2006 ) 9 (1 ):53 –60 .10.1079/PHN2005757 16480534 35 Alexandratos N The Mediterranean diet in a world context . Public Health Nutr (2006 ) 9 (1A ):111 –7 .10.1079/PHN2005932 16512957 36 Belahsen R Rguibi M . Population health and Mediterranean diet in southern Mediterranean countries . Public Health Nutr (2006 ) 9 (8A ):1130 –5 .10.1017/S1368980007668517 17378952 37 Belahsen R . Nutrition transition and food sustainability . Proc Nutr Soc (2014 ) 73 (3 ):385 –8 .10.1017/S0029665114000135 24824339 38 Roccaldo R Censi L D’Addezio L Toti E Martone D D’Addesa D Adherence to the Mediterranean diet in Italian school children (The ZOOM8 Study) . Int J Food Sci Nutr (2014 ) 65 (5 ):621 –8 .10.3109/09637486.2013.873887 24527679 39 Bonaccio M Di Castelnuovo A Bonanni A Costanzo S De Lucia F Persichillo M Decline of the Mediterranean diet at a time of economic crisis. Results from the Moli-sani study . Nutr Metab Cardiovasc Dis (2014 ) 24 (8 ):853 –60 .10.1016/j.numecd.2014.02.014 24819818 40 FAO/CIHEAM . Towards the Development of Guidelines for Improving the Sustainability of Diets and Food Consumption Patterns in the Mediterranean Area . Rome (2012 ). Available from: www.fao.org/docrep/016/ap101e/ap101e.pdf 41 Watson D Lorenz U St. Hansen M Szlezak J Zoboli R Kuhndt M European Environment Agency (EEA); towards a Set of Indicators on Sustainable Consumption and Production (scp) for EEA Reporting. European Topic Centre on Sustainable Consumption and Production (ETC/SCP) Working Paper 1/2010 . Copenhagen (2010 ). 42 IISD . Assessing Sustainable Development – Principles in Practice . Winnipeg, Canada : International Institute for Sustainable Development (1997 ). Available from: https://www.iisd.org/pdf/bellagio.pdf 43 OECD (Organisation for Economic Co-operation and Development) . Environmental Indicators – Development, Measurement and Use. Organisation for Economic Co-operation and Development Reference Paper . Paris (2003 ). 44 FAO . Nutrition Indicators for Development . Rome (2005 ). 45 Dernini S Meybeck A Burlingame B Gitz G Lacirignola C Debs P Developing a methodological approach for assessing the sustainability of diets: the Mediterranean diet as a case study . New Medit (2013 ) 12 (3 ):28 –36 . 46 CIHEAM . Final Declaration. 9th Meeting of the Ministers of Food, Agriculture and Fisheries of the Member Countries of CIHEAM . Valletta : International Centre for Advanced Mediterranean Agronomic Studies (CIHEAM) (2012 ). 47 Available from: http://ec.europa.eu/eurostat/web/household-budget-surveys 48 Naska A Fouskakis D Oikonomou E Almeida MDV Berg MA Gedrich K Dietary patterns and their socio-demographic determinants in ten European countries. Data from the DAFNE databank . Eur J Clin Nutr (2006 ) 60 :181 –90 .10.1038/sj.ejcn.1602284 16278696 49 Elmadfa I Meyer AL Nowak V Hasenegger V Putz P Verstraeten R European Nutrition and Health Report 2009 . Ann Nutr Metab (2009 ) 55 (2 Suppl ):1 –40 .10.1159/000244607 20104000 50 FAO . FAOSTAT . (2015 ). Available from: http://faostat3.fao.org/browse/FB/*/E 51 World Bank . Global Consumption Database . Available from: http://datatopics.worldbank.org/consumption/ 52 Yoon GA Hwang HJ . Effect of soy protein/animal protein ratio on calcium metabolism of the rat . Nutrition (2006 ) 22 (4 ):414 –8 .10.1016/j.nut.2005.07.019 16458484 53 FAO . Statistical Yearbook . (2013 ). 54 FAO . Human Energy Requirements. Report of a Joint FAO/WHO Expert Consultation . Rome (2001 ). 55 Rolls BJ Morris EL Roe LS . Portion size of food affects energy intake in normal-weight and overweight men and women . Am J Clin Nutr (2002 ) 76 :1207 –13 .12450884 56 Diliberti N Bordi PL Conklin MT Roe LS Rolls BJ . Increased portion size leads to increased energy intake in a restaurant meal . Obes Res (2004 ) 12 :562 –8 .10.1038/oby.2004.64 15044675 57 Ello-Martin JA Ledikwe JH Rolls BJ . The influence of food portion size and energy density on energy intake: implications for weight management . Am J Clin Nutr (2005 ) 82 :236S –41S .16002828 58 Lioret S Volatier JL Lafay L Trouvier M Maire B . Is food portion size a risk factor of childhood overweight? Eur J Clin Nutr (2009 ) 63 :382 –91 .18030311 59 Nishida C Uauy R Kumanyika S Shetty P The joint WHO/FAO expert consultation on diet, nutrition and the prevention of chronic diseases: process, product and policy implications . Public Health Nutr (2004 ) 7 (1A ):245 –50 .14972063 60 Cialfa E Turrini A Lintas C A national food survey. Food balance sheets and other methodologies: a critical overview (chapter 4) . In: MacDonald I , editor. ILSI Monographs – Monitoring Dietary Intakes . (1991 ). p. 24 –44 . 61 EURODIET . Nutrition and Diet for Healthy Lifestyles in Europe. Kafatos AG (Coordinator) . (2001 ). Available from: http://ec.europa.eu/health/archive/ph_determinants/life_style/nutrition/report01_en.pdf 62 Tse H-F Lip GY Coats AJS , editors. Diets and ischaemic heart disease (chapter 5.2) . Oxford Desk Reference: Cardiology . (2011 ). p. 139 –212 . 63 Willet WC . The Mediterranean diet: science and practice . Public Health Nutr (2006 ) 9 :105 –10 .16512956 64 FAO . Guide pour la mesure de la diversité alimentaire au niveau des individus et des ménages . Rome : FAO (2007 ). 65 Hoddinott J Yohannes Y Dietary Diversity as a Household Food Security Indicator . Washington, DC : Projet d’Assistance Technique en Alimentation et Nutrition, Académie pour le Développement de l’Education (2002 ). Available from: www.fantaproject.org/publications/dietdiversity1.shtml 66 US Department of Agriculture . Food Guide Pyramid . Washington, DC : USDA (1996 ). 67 Deleuze Ntandou Bouzitou G Fayomi B Delisl H Cahiers Santé vol. 15, n° 4, octobre-novembre-décembre . (2005 ). 68 Savy M Martin-Prevel Y Sawadogo P Kameli Y Delpeuch F . Use of variety/diversity scores for diet quality measurement: relation with nutritional status of women in a rural area in Burkina Faso . Eur J Clin Nutr (2005 ) 59 :703 –16 .10.1038/sj.ejcn.1602135 15867942 69 Bezerra IN Sichieri R . Household food diversity and nutritional status among adults in Brazil . Int J Behav Nutr Phys Act (2011 ) 8 :22 .10.1186/1479-5868-8-22 21439090 70 Drewnowski A Henderson SA Shore AB Fischler C Preziosi P Hercberg S . Diet quality and dietary diversity in France: implications for the French paradox . J Am Diet Assoc (1996 ) 96 (7 ):663 –9 .10.1016/S0002-8223(96)00185-X 8675909 71 Vadiveloo M Dixon LB Mijanovich T Elbel B Parekh N . Development and evaluation of the US Healthy Food Diversity index . Br J Nutr (2014 ) 112 :1562 –74 .10.1017/S0007114514002049 25242619 72 FAO/INFOODS . FAO/INFOODS Food Composition Database for Biodiversity Version 2.1 – BioFoodComp2.1 . Rome : FAO (2013 ). Available from: http://www.fao.org/infoods/infoods/food-biodiversity/en/ 73 Willer . The World of Organic Agriculture 2013 and 2014 . Available from: www.organic-world.net 74 Guidelines for the Production, Processing, Labelling and Marketing of Organically Produced Foods (GL 32-1999) . Codex Alimentarius : Rome, Italy , 2010 ; Adopted 1999, Revisions 2001, 2003, 2004 and 2007, Amendments 2008, 2009 and 2010. 75 Available from: http://www.ifoam.org/en/ifoam-standard Council Regulation (EC) No 834/2007 of 28 June 2007 on Organic Production and Labelling of Organic Products and Repealing Regulation (EEC) No 2092/91. Commission Regulation (EC) No 889/2008 of 5 September 2008 Laying Down Detailed Rules for the Implementation of Council Regulation (EC) No 834/2007 of 28 June 2007 on Organic Production and Labelling of Organic Products with Regard to Organic Production, Labelling of Organic Products and Control. 78 Available from: http://www.maff.go.jp/e/jas/specific/criteria_o.html 79 Available from: http://www.usda.gov/wps/portal/usda/usdahome?navid%3DORGANIC_CERTIFICATIO 80 SACN (Scientific Advisory Committee on Nutrition) . Dietary Reference Values for Energy . London (2011 ). 81 WHO (World Health Organization) . Report of the Workshop on Integration of Data on Physical Activity Patterns Report n°4 . (2009 ). Available from: http://www.euro.who.int/__data/assets/pdf_file/0007/87433/E93707.pdf 82 WHO Global Infobase . Available from: https://apps.who.int/infobase/ 83 Pereira MA FitzGerald SJ Gregg EW Joswiak ML Ryan WJ Suminski RR A collection of Physical Activity Questionnaires for health-related research . Med Sci Sports Exerc (1997 ) 29 (6 Suppl ):S1 –205 .9243481 84 Available from: http://www.who.int/mediacentre/factsheets/fs311/en/ 85 Available from: http://www.who.int/healthinfo/global_burden_disease/GlobalHealthRisks_report_full.pdf 86 WHO . Physical Status: The Use and Interpretation of Anthropometry . Geneva : World Health Organization (1995 ). 87 Guigoz Y Vellas B Garry PJ Assessing the nutritional status of the elderly: the Mini Nutritional Assessment as part of the geriatric evaluation . Nutr Rev (1996 ) 54 :S59 –65 .10.1111/j.1753-4887.1996.tb03793.x 8919685 88 Donini LM Ricciardi LM Neri B Lenzi A Marchesini G . Risk of malnutrition (over and under-nutrition): validation of the JaNuS screening tool . Clin Nutr (2014 ) 33 :1087 –94 .10.1016/j.clnu.2013.12.001 24373664 89 Available from: http://www.cdc.gov/nchs/data/nhanes/nhanes3/cdrom/nchs/manuals/anthro.pdf 90 Available from: http://www.fao.org/fileadmin/templates/food_composition/documents/Nutrition_assessment/c2.pdf 91 Available from: http://www.measuredhs.com 92 Available from: http://www.childinfo.org/mics.html 93 Available from: http://www.who.int/bmi/index.jsp 94 HLPE . Food Losses and Waste in the Context of Sustainable Food Systems. A Report by the High Level Panel of Experts on Food Security and Nutrition of the Committee on World Food Security, Rome 2014. HLPE. Food Losses and Waste in the Context of Sustainable Food Systems. Report HLPE . Rome (2014 ). Available from: http://www.fao.org/3/a-i3901e.pdf 95 De Marco A Velardi M Camporeale C Screpanti A Vitale M The adherence of the diet to Mediterranean principle and its impacts on human and environmental health . Int J of Environ Prot Policy (2014 ) 2 (2 ):64 –75 .10.11648/j.ijepp.20140202.14 96 Mediterra . The Future of Agriculture and Food in Mediterranean Countries/International Centre for Advanced Mediterranean Agronomic Studies . Paris : Presses de Sciences Po (2008 ). 97 Lobstein T Rigby N Leach R International Obesity Task Force EU Platform Briefing Paper Prepared in Collaboration with the European Association for the Study of Obesity . Brussels (2005 ). 98 Rivera D Obón C Heinrich M Inocencio C Verde A Fajardo J Gathered Mediterranean food plants – ethnobotanical investigations and historical development . Forum Nutr (2006 ) 59 :18 –74 .10.1159/000095207 16917173

PMC005xxxxxx/PMC5002407.txt

==== Front Front NeurosciFront NeurosciFront. Neurosci.Frontiers in Neuroscience1662-45481662-453XFrontiers Media S.A. 10.3389/fnins.2016.00402NeuroscienceOriginal ResearchMaternal Forced Swimming Reduces Cell Proliferation in the Postnatal Dentate Gyrus of Mouse Offspring Wasinski Frederick 12Estrela Gabriel R. 2Arakaki Aline M. 2Bader Michael 13Alenina Natalia 1Klempin Friederike 1*†Araújo Ronaldo C. 2†1Max Delbrueck Center for Molecular Medicine, Molecular Biology of Hormones in the Cardiovascular SystemBerlin, Germany2Department of Biophysics, Federal University of São PauloSão Paulo, Brazil3Charité - University Medicine BerlinBerlin, GermanyEdited by: Paolo Peretto, University of Turin, Italy Reviewed by: Juan Manuel Encinas, Achucarro Basque center for Neuroscience/Ikerbasque, Spain; Claudio Giachino, University of Basel, Switzerland *Correspondence: Friederike Klempin friederike.klempin@mdc-berlin.deThis article was submitted to Neurogenesis, a section of the journal Frontiers in Neuroscience †These authors have contributed equally to this work. 29 8 2016 2016 10 40222 7 2016 16 8 2016 Copyright © 2016 Wasinski, Estrela, Arakaki, Bader, Alenina, Klempin and Araújo.2016Wasinski, Estrela, Arakaki, Bader, Alenina, Klempin and AraújoThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.Physical exercise positively affects the metabolism and induces proliferation of precursor cells in the adult brain. Maternal exercise likewise provokes adaptations early in the offspring. Using a high-intensity swimming protocol that comprises forced swim training before and during pregnancy, we determined the effect of maternal swimming on the mouse offspring's neurogenesis. Our data demonstrate decreased proliferation in sublayers of the postnatal dentate gyrus in offspring of swimming mother at postnatal day (P) 8 accompanied with decreased survival of newly generated cells 4 weeks later. The reduction in cell numbers was predominantly seen in the hilus and molecular layer. At P35, the reduced amount of cells was also reflected by a decrease in the population of newly generated immature and mature neurons of the granule cell layer. Our data suggest that forced maternal swimming at high-intensity has a negative effect on the neurogenic niche development in postnatal offspring. exerciseneurogenesisBrdUpregnancyhippocampusDeutsche Forschungsgemeinschaft10.13039/501100001659DFG award KL 2805/1-1 ==== Body Introduction Continued neuron generation in the adult brain contributes to plasticity that allows immediate or long-term adaptations to neurogenic stimuli (Kempermann et al., 1997; van Praag et al., 1999). Physical exercise is an external factor that strongly induces the proliferation of precursor cells in the adult mouse hippocampus (van Praag et al., 1999; Kronenberg et al., 2003). The beneficial effect of physical activity lies in its proposed use as non-pharmacotherapy to increase mood and to ameliorate the outcome of neurological diseases (Ahlskog et al., 2011; Klempin et al., 2013; Gomes et al., 2014). In animal models, forced or voluntary exercise can be distinguished. Several studies reveal positive effects of voluntary running such as improved metabolism, memory, or increased adult neurogenesis in rodents (van Praag et al., 1999; Muller et al., 2011; Inoue et al., 2015). Furthermore, maternal running exercise reveals beneficial effects on metabolic health of offspring (Carter et al., 2013; Stanford et al., 2015). However, inadequate duration or intensity may lead to excessive stress associated with increased corticosterone levels (Girard and Garland, 2002; Contarteze et al., 2008) that may in turn decrease adult neurogenesis (Wong and Herbert, 2006). More recently, studies focus on swimming exercise as it improves the immune system and has neuroprotective effects (Wasinski et al., 2013; Goes et al., 2014); the effect on hippocampal neurogenesis has not yet been shown. In humans, one advantage of moderate swimming as compared to running is continued training during pregnancy. In animal models, we and others have shown that maternal swimming exercise is beneficial for the mother as well as the offspring's metabolism, it increases glucose tolerance and induces mitochondrial biogenesis (Damasceno et al., 2013; Marcelino et al., 2013; Wasinski et al., 2015). However, the effect of maternal swimming on brain plasticity in the offspring has not yet been studied. In humans, continued physical activity during pregnancy showed advanced neurobehavioral maturation in children (reviewed Clapp, 2000). A pro-mitotic effect in the postnatal hippocampus of mouse offspring following maternal voluntary wheel running has been shown (Bick-Sander et al., 2006). Here, we studied the effect of maternal swimming exercise on early postnatal cell proliferation in the dentate gyrus (DG) of mouse offspring. We found that maternal swimming as opposed to voluntary wheel running decreased the number of mitotic cells in the DG at P8, that led to decreased numbers of newly generated cell survival at P35. The reduced amount of cells was predominantly seen in the immature and mature neuron population. Materials and methods Animals Female C57BL/6 mice (8–12 weeks old) were obtained from the Animal Care Facility at the Federal University of São Paulo (UNIFESP). Animals (n = 14, 7 per group) were single housed under standard conditions with access to water and food ad libitum. All procedures were previously reviewed and approved by the internal ethical committee of the institution. One group of mice was subjected to physical swimming through a period of 6 weeks; the other group did not exercise (sedentary group / Ctl). Briefly, after a 2-week-adaptation period for the exercising group, female animals were mated with males of the same strain; the swimming assigned group continued training in the meantime, and during pregnancy (4 weeks including mating; Figure 1A). Figure 1 Quantitative analysis of corticosterone levels in mothers, and absolute numbers of BrdU- positive cells in the pup's postnatal dentate gyrus. (A) Experimental design. Female mice are either subjected (Swim group) or not subjected (Sedentary group, Ctl) to a training adaptation phase for 2 weeks followed by mating (approx. 1–5 days). The swimming group continues exercise during pregnancy (1 h per day, 5 days per week) for 18–21 days (depending on the length of pregnancy). Offspring of both groups is randomly split into two groups to determine I) proliferation and II) survival. (B) Corticosterone analysis in mothers at day 20 of pregnancy reveals significant increased levels following maternal swimming. (C) When determining proliferation (P8) and survival (P35), the number of BrdU+ cells is significantly reduced in the dentate gyrus of pups from swimming mothers. Student's t test *p < 0.05, **p < 0.01 between Ctl and Swim; data are presented as mean ± SEM. To study postnatal neurogenesis in the offspring, male and female pups from either sedentary or trained mothers were split randomly into two groups to determine cell proliferation (at P8) and survival of newly generated cells (P35). Bromodeoxyuridine (BrdU, 50 mg/Kg body weight; Sigma) was injected intraperitoneal on P7 to assess cell numbers 24 h later (P8, n = 24), and ones per day on P7, P8, and P9 to examine cell survival and differentiation 4 weeks later (P35, n = 22). Swimming exercise One group of female mice was subjected to swimming sessions in a swimming system adapted for mice (at 30°C) (Evangelista et al., 2003). Briefly, the 300-liter tank consists of 10 lanes fitted with air pumps to maintain mice in constant motion. Swimming was done during consecutive weekdays with 2 days off over the weekend. The adaptation period comprised 14 days with initial swimming sessions of 20 min per day during the first week (day 1–5). The time gradually increased until mice were able to swim for 60 min per day at “moderate intensity” induced by a load (3% of the body weight attached to the tail (Wasinski et al., 2013). Following 1–5 days of mating, pregnant mice were exercising 60 min per day (18–21 days, corresponding to the last 3 weeks of the total exercise period, Figure 1A); however, during the last week of pregnancy the load was removed. The presence of a vaginal plug was used to indicate pregnancy and considered as the first day of gestation (Figure 1A). Swimming was stopped immediately when the pups were born. Corticosterone levels In the morning of day 20 of pregnancy, blood was collected from the heart shortly before perfusion of ketamine/xylazine-anesthesized mothers (another set of animals, n = 6 per group for swimming and sedentary control). Blood serum was obtained by centrifugation at 10,000 rpm for 15 min at −80 degrees; corticosterone levels were measure with a commercial enzyme-linked immunosorbent assay (ELISA, IBL international, Germany). Immunohistochemistry For proliferation, mice were decapitated at P8 and the brains were transferred into 4% paraformaldehyde (PFA) over night. For survival of newly generated cells, P35 offspring was deeply anesthetized with ketamine/xylazine (10 ml/Kg body weight) and perfused transcardially with 0.9% sodium chloride followed by 4% PFA. Brains were removed from the skulls, postfixed in PFA at 4°C over night, and transferred into 30% sucrose (dissolved in 0.1 M phosphate buffer). Sequential 40 μm coronal sections were cut on a cryostat (Leica CM 3050, Germany) and cryoprotected. For BrdU staining, DNA was denatured in 2N HCl for 20 min at 37°C. Sections were then rinsed in 0.1 M borate buffer and washed in Tris-buffered saline (TBS). Sections were stained free-floating with all antibodies diluted in TBS containing 3% donkey serum and 0.1% Triton X-100. Primary antibodies were applied in the following concentrations: anti-BrdU (rat, 1:500; Biozol), anti-calbindin D-28k (rabbit, 1:000; Swant), anti-doublecortin (DCX; goat, 1:250; Santa Cruz Biotechnology), anti-NG2 (rabbit, 1:1000; Santa Cruz Biotechnology), anti-S100β (mouse, 1:2000; Acris Antibodies). For immunofluorescence, Alexa488-conjugated, Cy3-conjugated, or Alexa647-conjugated secondary antibodies (Jackson ImmunoResearch Laboratories) were used at a concentration of 1:250. Fluorescent sections were coverslipped in Vector shield (Vector Laboratories). Quantification One-in-six series of sections of each brain were stained, and BrdU-immunoreactive cells were counted throughout the rostro-caudal extent of the DG, and with respect to the different layers (40x, Keyence microscope). Results were multiplied by six to obtain the total number of BrdU-positive cells per DG. A one-in-six series of adjacent sections was further used to measure DG volume at P35. First, the area of the DG was traced using a semiautomatic area measurement system (10x, Keyence microscope). The volume was determined by summing the traced areas per slice multiplied by the distance between slices (240 μm). For phenotypic analysis, fifty to 100 randomly selected cells per animal were co-stained for markers of specific lineages and evaluated by examining orthogonal views from a series of focal planes using a Leica TCS SP5 (Leica) confocal microscope. Statistical analysis Student's t test was used for individual pair-wise comparisons. To detect differences between group means, one-way analysis of variance (ANOVA) was performed (PRISM software). All values are expressed as mean ± SEM. P < 0.05 were considered statistically significant. Results We applied daily swimming to pregnant mice to determine the effect on the offspring's postnatal neurogenesis (Figure 1A). Beforehand, corticosterone was measured in mothers (20th day of pregnancy) in the morning after the last swimming session to detect stress levels following continued exercise. Blood serum of exercising mice revealed a significant increase in corticosterone levels compared to sedentary pregnant controls (Ctl 180.6 ± 14.1 ng/ml vs. Swim 307.6 ± 88.9 ng/ml, p = 0.012; Figure 1B). When examining the offspring, no changes in pup size, sex ratio, or bodyweight of the pups at P8 and P35 were observed (data not shown). We next assessed the number of proliferating precursor cells at P8, 24 h after one injection of BrdU. Our data revealed significantly less BrdU-labeled cells in the entire DG of mice from swimming mothers (Ctl 16,730 ± 776 vs. Swim 12,540 ± 918, p = 0.0030; Figure 1C). At P35, the reduction in the number of BrdU-positive cells was maintained in mice of swimming mothers compared to control (p = 0.010; Figure 1C). Although, less proliferating cells were found after maternal swimming, no difference has been observed in the volume of the DG at P35 (Ctl 0.806 ± 0.071 vs. Swim 0.855 ± 0.063 mm3, p = 0.631). We next determined which layer in the postnatal DG of P8 mice might reflect the decrease and found significantly reduced numbers (approximately one-third) of proliferating precursor cells in the hilus (p = 0.0002), and molecular layer (ML, p = 0.0014); fewer cells were also found in the subgranular zone (SGZ, p = 0.0156) with unchanged numbers in the granule cell layer (GCL; Figure 2A). Notably, the majority of BrdU-positive cells in the sedentary group was found in ML (One-way ANOVA to SGZ, GCL, and hilus, respectively, p < 0.001) with similar cell numbers observed for all other layers in the control groups. Spatial distribution of surviving precursor cells at P35, 4 weeks after three daily injections of BrdU, revealed significantly fewer cells in the hilus (p = 0.006) and ML (p = 0.022) of pups born of swimming mothers as was observed at P8 (Figure 2B). Noticeable, at P35 the number of BrdU-positive cells in the GCL was highly reduced in mice of exercising mothers in comparison to sedentary controls (p = 0.027; Figure 2B). Majority of proliferating cells for both experimental groups was found in the SGZ with respect to the neurogenic niche in the adult. Cell numbers of the SGZ were reduced in comparison to P8 by only a third while survival of proliferating cells was highly decreased in all other sublayers from P8 to P35. Figure 2 Quantification of BrdU-positive cells in sublayers of the postnatal dentate gyrus (DG) at P8, 24 h after one injection of BrdU (A), and at P35, 4 weeks after three daily injections of BrdU (B). (A,B) The hilus and molecular layer (ML) of pups of swimming mothers (Swim) reveal significant less numbers compared to sedentary control (Ctl). At P8 the number is also reduced in the subgranular zone (SGZ), at P35 in the granule cell layer (GCL) of the swimming group. Scale bar = 100 μm. Student's t test *p < 0.05, **p < 0.01, ***p < 0.001 indicates statistical significance relative to Ctl; One-way ANOVA ###p < 0.001 to all layers of sedentary group Ctl; in (A), and $p < 0.05, $$$p < 0.001 relative to SGZ for each group in (B); data are presented as mean ± SEM. Next, we assessed the phenotype of newly generated cells in the SGZ and GCL at P35 by co-labeling with markers of the neuronal and glial lineages. Our data revealed that only a very small population of cells coexpress BrdU and the oligodendrocyte precursor marker Neuron-Glia (NG)2 or the astroglia marker S100β as expected, with no difference seen between the experimental groups (Table 1). However, the decrease in the number of BrdU-positive cells was reflected by a significant reduction in the fraction of immature (BrdU/DCX-positive) and mature (BrdU/calbindin-positive) neurons in the SGZ and GCL of offspring of swimming mothers (p = 0.031 and p = 0.023, respectively; Table 1). Table 1 Number and phenotypes of BrdU-positive cells in the SGZ/GCL at P35. Cell number DCX (%) Calbindin (%) S100ß (%) NG2 (%) Other (%) Control 3382 (535) 23.1 (3.4) 35.6 (1.2) 2.7 (0.8) 5.2 (1.0) 33.4 Swim 2499 (586) 12.9 (1.3)* 27.8 (2.4)* 1.9 (0.2) 5.0 (1.7) 52.4 The percentages of BrdU-positive cells co-labeled for DCX (immature neurons), Calbindin (mature granule cells), S100β (astrocytes), NG2 (oligodendrocyte precursors) or neither marker are presented. All data are presented as mean (standard error). * p < 0.05 statistical significance between Control and Swim. Discussion Our data demonstrate that maternal “high-intensity” swimming exercise decreases proliferation and survival of newly generated cells in the DG of postnatal offspring. Reduced numbers predominantly found in the hilus and ML reflected the overall decrease. At P35, the reduction was mainly seen in the number of newly generated cells of the GCL expressing DCX and calbindin. Our data suggest that forced swimming exercise during pregnancy has a negative effect on the neurogenic niche development in postnatal offspring. Voluntary wheel running has been established to augment neurogenesis in rodents (van Praag et al., 1999; Kronenberg et al., 2003). Soluble factors accompany this effect such as growth- and neurotrophic factors, or neuromodulators (Trejo et al., 2001; Fabel et al., 2003; Moon et al., 2012; Klempin et al., 2013). An earlier study also showed stimulation of postnatal DG development in the offspring of maternal voluntary wheel running (Bick-Sander et al., 2006). Maternal swimming in contrast, appears to be a bad exercise for the pup's brain development. A positive outcome on brain maturity as well as metabolism depends on frequency, duration, and intensity of the exercise (Ploughman et al., 2007; Speck et al., 2014). Swimming is rather forced work out for rodents and studied as aerobic exercise to challenge the metabolism and immune system (Wasinski et al., 2013; Cunha et al., 2015). Only few studies address the effect of swimming exercise on the brain revealing a neuroprotective role in Parkinson's disease, and increased neurotrophin levels (Goes et al., 2014; Jiang et al., 2014). There is no literature on “voluntary” swimming; however, different swimming protocols have been used comprising of either low- or high-intensity exercise distinguished by the length and intensity of the swimming (10, 30, 60, or 120 min per day for 2 to 8 weeks) (Lee et al., 2006; Radak et al., 2006; Marcelino et al., 2013; Wasinski et al., 2015) that revealed beneficial effects on both, brain and periphery. We have recently shown that swimming prior and during pregnancy up to 1 h per day/5 days per week positively affect glucose tolerance (Wasinski et al., 2015). Nevertheless, it seems no benefits occur for the postnatal hippocampus in the offspring using the same high-intensity protocol, as the population of immature and mature neurons is decreased considering more DCX-, and calbindin-positive cells are better. Exhaustive exercise may lead to stress and increased corticosterone levels (Contarteze et al., 2008). Our protocol forces exercise at consistent times and intensity and we found elevated serum corticosterone levels in highly pregnant female mice following 6 weeks of swimming. Pregnancy itself is associated with the attenuation of the HPA-Axis (reviewed in Slattery and Neumann, 2008) affecting the offspring development (reviewed in Douglas, 2010). Prenatal stress can also reduce postnatal neurogenesis (Lucassen et al., 2009); yet, there are mechanisms that protect the offspring against stress-induced maternal corticosterone levels (Thomas et al., 2007). We did not determine glucocorticoid levels in the pups as they may also affect postnatal cell genesis. Other studies in contrary show that early life-stress like maternal deprivation did not affect adult neurogenesis (Herpfer et al., 2012). Furthermore, when (maternal) exercise was done at lower duration and speed, a similar increase in corticosterone levels was observed which did not negatively affect neurotrophic factor levels and even increased the offspring's performance in the Morris water maze in rats (Ploughman et al., 2007; Akhavan et al., 2008). These data reveal that forced exercise at lower intensity does not harm brain plasticity. A new study claims that treadmill running has a positive effect on the offspring brain development by increased BDNF levels (da Silva et al., 2016). The hilus is the source of granule cells in late development of the neurogenic niche (Altman and Bayer, 1990). In pups of swimming mothers, BrdU cells in the hilus were reduced at both time points together with decreased cell numbers in the SGZ (at P8). This might explain the reduced number of cells in the GCL 4 weeks later (at P35) where to newly generated cells migrate out of the SGZ. However, the majority of newly generated cells at P35 for both experimental groups (w/o statistical difference) was found aligned in the SGZ shaping the neurogenic niche; thus this continues source of granule cells is still intact. In addition to being less, the distribution of immature and mature newborn neurons among newly generated cells was decreased. We tested other cell fate markers and observed no changes for NG2 or S100β expression; Sox2 was not visible. Considering the protocol in our study being stressful, it largely affected the number of proliferating cells in the hilus and ML with considerable effects on the postnatal neurogenic niche. Further experiments will define the impact on brain function, and cognitive processes later in life since changes in hippocampal neurogenesis due to stress maybe recovered in adulthood (Herpfer et al., 2012). These data are the first to determine the effect of maternal swimming on the offspring's postnatal hippocampal neurogenesis. Using the same high-intensity swimming exercise protocol, it beneficially influences the metabolism of adult offspring (Wasinski et al., 2015) whereas it negatively affects newly generated cells in the postnatal neurogenic niche DG. Author contributions FW, FK, and RA designed research; FW, GE, AA, and FK performed research and analyzed the data; FW and FK wrote the paper. FW, GE, AA, MB, NA, FK, RA reviewed the manuscript. Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. FAPESP 2013/04757-9 to FW, CAPES (DAAD/PROBRAL) 427/15 to RA and MB, and DFG award KL 2805/1-1 to FK supported this work. ==== Refs References Ahlskog J. E. Geda Y. E. Graff-Radford N. R. Petersen R. C. (2011 ). Physical exercise as a preventive or disease-modifying treatment of dementia and brain aging . Mayo Clin. Proc. 86 , 876 –884 . 10.4065/mcp.2011.0252 21878600 Akhavan M. M. Emami-Abarghoie M. Safari M. Sadighi-Moghaddam B. Vafaei A. A. Bandegi A. R. . (2008 ). Serotonergic and noradrenergic lesions suppress the enhancing effect of maternal exercise during pregnancy on learning and memory in rat pups . Neuroscience 151 , 1173 –1183 . 10.1016/j.neuroscience.2007.10.051 18207332 Altman J. Bayer S. A. (1990 ). Migration and distribution of two populations of hippocampal granule cell precursors during the perinatal and postnatal periods . J. Comp. Neurol. 301 , 365 –381 . 10.1002/cne.903010304 2262596 Bick-Sander A. Steiner B. Wolf S. A. Babu H. Kempermann G. (2006 ). Running in pregnancy transiently increases postnatal hippocampal neurogenesis in the offspring . Proc. Natl. Acad. Sci. U.S.A. 103 , 3852 –3857 . 10.1073/pnas.0502644103 16537457 Carter L. G. Qi N. R. De Cabo R. Pearson K. J. (2013 ). Maternal exercise improves insulin sensitivity in mature rat offspring . Med. Sci. Sports Exerc. 45 , 832 –840 . 10.1249/MSS.0b013e31827de953 23247711 Clapp J. F. IIIrd. (2000 ). Exercise during pregnancy. A clinical update . Clin. Sports Med. 19 , 273 –286 . 10.1016/S0278-5919(05)70203-9 10740759 Contarteze R. V. Manchado Fde B. Gobatto C. A. De Mello M. A. (2008 ). Stress biomarkers in rats submitted to swimming and treadmill running exercises . Comp. Biochem. Physiol. Part A. Mol. Integr. Physiol. 151 , 415 –422 . 10.1016/j.cbpa.2007.03.005 17428717 Cunha V. N. de Paula Lima M. Motta-Santos D. Pesquero J. L. de Andrade R. V. de Almeida J. A. . (2015 ). Role of exercise intensity on GLUT4 content, aerobic fitness and fasting plasma glucose in type 2 diabetic mice . Cell Biochem. Funct. 33 , 435 –442 . 10.1002/cbf.3128 26467261 Damasceno D. C. Silva H. P. Vaz G. F. Vasques-Silva F. A. Calderon I. M. Rudge M. V. . (2013 ). Diabetic rats exercised prior to and during pregnancy: maternal reproductive outcome, biochemical profile, and frequency of fetal anomalies . Reprod. Sci. 20 , 730 –738 . 10.1177/1933719112461186 23171681 da Silva S. G. de Almeida A. A. Fernandes J. Lopim G. M. Cabral F. R. Scerni D. A. . (2016 ). Maternal exercise during pregnancy increases bdnf levels and cell numbers in the hippocampal formation but not in the cerebral cortex of adult rat offspring . PLoS ONE 11 :e0147200 . 10.1371/journal.pone.0147200 26771675 Douglas A. J. (2010 ). Baby on board: do responses to stress in the maternal brain mediate adverse pregnancy outcome? Front. Neuroendocrinol. 31 , 359 –376 . 10.1016/j.yfrne.2010.05.002 20546772 Evangelista F. S. Brum P. C. Krieger J. E. (2003 ). Duration-controlled swimming exercise training induces cardiac hypertrophy in mice . Braz. J. Med. Biol. Res . 36 , 1751 –1759 . 10.1590/S0100-879X2003001200018 14666261 Fabel K. Fabel K. Tam B. Kaufer D. Baiker A. Simmons N. . (2003 ). VEGF is necessary for exercise-induced adult hippocampal neurogenesis . Eur. J. Neurosci. 18 , 2803 –2812 . 10.1111/j.1460-9568.2003.03041.x 14656329 Girard I. Garland T. Jr. (2002 ). Plasma corticosterone response to acute and chronic voluntary exercise in female house mice . J. Appl. Physiol. (1985) 92 , 1553 –1561 . 10.1152/japplphysiol.00465.2001 11896022 Goes A. T. Souza L. C. Filho C. B. Del Fabbro L. De Gomes M. G. Boeira S. P. . (2014 ). Neuroprotective effects of swimming training in a mouse model of Parkinson's disease induced by 6-hydroxydopamine . Neuroscience 256 , 61 –71 . 10.1016/j.neuroscience.2013.09.042 24090962 Gomes F. G. Gomes Da Silva S. Cavalheiro E. A. Arida R. M. (2014 ). Beneficial influence of physical exercise following status epilepticus in the immature brain of rats . Neuroscience 274 , 69 –81 . 10.1016/j.neuroscience.2014.05.024 24857853 Herpfer I. Hezel H. Reichardt W. Clark K. Geiger J. Gross C. M. . (2012 ). Early life stress differentially modulates distinct forms of brain plasticity in young and adult mice . PLoS ONE 7 :e46004 . 10.1371/journal.pone.0046004 23071534 Inoue K. Hanaoka Y. Nishijima T. Okamoto M. Chang H. Saito T. . (2015 ). Long-term mild exercise training enhances hippocampus-dependent memory in rats . Int. J. Sports Med. 36 , 280 –285 . 10.1055/s-0034-1390465 25429548 Jiang P. Dang R. L. Li H. D. Zhang L. H. Zhu W. Y. Xue Y. . (2014 ). The impacts of swimming exercise on hippocampal expression of neurotrophic factors in rats exposed to chronic unpredictable mild stress . Evid. Based Complement. Alternat. Med. 2014 :729827 . 10.1155/2014/729827 25477997 Kempermann G. Kuhn H. G. Gage F. H. (1997 ). More hippocampal neurons in adult mice living in an enriched environment . Nature 386 , 493 –495 . 10.1038/386493a0 9087407 Klempin F. Beis D. Mosienko V. Kempermann G. Bader M. Alenina N. (2013 ). Serotonin is required for exercise-induced adult hippocampal neurogenesis . J. Neurosci. 33 , 8270 –8275 . 10.1523/JNEUROSCI.5855-12.2013 23658167 Kronenberg G. Reuter K. Steiner B. Brandt M. D. Jessberger S. Yamaguchi M. . (2003 ). Subpopulations of proliferating cells of the adult hippocampus respond differently to physiologic neurogenic stimuli . J. Comp. Neurol. 467 , 455 –463 . 10.1002/cne.10945 14624480 Lee H. H. Kim H. Lee J. W. Kim Y. S. Yang H. Y. Chang H. K. . (2006 ). Maternal swimming during pregnancy enhances short-term memory and neurogenesis in the hippocampus of rat pups . Brain Dev. 28 , 147 –154 . 10.1016/j.braindev.2005.05.007 16368211 Lucassen P. J. Bosch O. J. Jousma E. Kromer S. A. Andrew R. Seckl J. R. . (2009 ). Prenatal stress reduces postnatal neurogenesis in rats selectively bred for high, but not low, anxiety: possible key role of placental 11beta-hydroxysteroid dehydrogenase type 2 . Eur. J. Neurosci. 29 , 97 –103 . 10.1111/j.1460-9568.2008.06543.x 19032587 Marcelino T. B. Longoni A. Kudo K. Y. Stone V. Rech A. de Assis A. M. . (2013 ). Evidences that maternal swimming exercise improves antioxidant defenses and induces mitochondrial biogenesis in the brain of young Wistar rats . Neuroscience 246 , 28 –39 . 10.1016/j.neuroscience.2013.04.043 23639877 Moon H. Y. Kim S. H. Yang Y. R. Song P. Yu H. S. Park H. G. . (2012 ). Macrophage migration inhibitory factor mediates the antidepressant actions of voluntary exercise . Proc. Natl. Acad. Sci. U.S.A. 109 , 13094 –13099 . 10.1073/pnas.1205535109 22826223 Muller A. P. Gnoatto J. Moreira J. D. Zimmer E. R. Haas C. B. Lulhier F. . (2011 ). Exercise increases insulin signaling in the hippocampus: physiological effects and pharmacological impact of intracerebroventricular insulin administration in mice . Hippocampus 21 , 1082 –1092 . 10.1002/hipo.20822 20824731 Ploughman M. Granter-Button S. Chernenko G. Attwood Z. Tucker B. A. Mearow K. M. . (2007 ). Exercise intensity influences the temporal profile of growth factors involved in neuronal plasticity following focal ischemia . Brain Res. 1150 , 207 –216 . 10.1016/j.brainres.2007.02.065 17382914 Radak Z. Toldy A. Szabo Z. Siamilis S. Nyakas C. Silye G. . (2006 ). The effects of training and detraining on memory, neurotrophins and oxidative stress markers in rat brain . Neurochem. Int. 49 , 387 –392 . 10.1016/j.neuint.2006.02.004 16564605 Slattery D. A. Neumann I. D. (2008 ). No stress please! Mechanisms of stress hyporesponsiveness of the maternal brain . J. Physiol. 586 , 377 –385 . 10.1113/jphysiol.2007.145896 17974588 Speck A. E. Tromm C. B. Pozzi B. G. Paganini C. S. Tuon T. Silveira P. C. . (2014 ). The dose-dependent antioxidant effects of physical exercise in the hippocampus of mice . Neurochem. Res. 39 , 1496 –1501 . 10.1007/s11064-014-1339-6 24858240 Stanford K. I. Lee M. Y. Getchell K. M. So K. Hirshman M. F. Goodyear L. J. (2015 ). Exercise before and during pregnancy prevents the deleterious effects of maternal high-fat feeding on metabolic health of male offspring . Diabetes 64 , 427 –433 . 10.2337/db13-1848 25204976 Thomas R. M. Hotsenpiller G. Peterson D. A. (2007 ). Acute psychosocial stress reduces cell survival in adult hippocampal neurogenesis without altering proliferation . J. Neurosci. 27 , 2734 –2743 . 10.1523/JNEUROSCI.3849-06.2007 17360895 Trejo J. L. Carro E. Torres-Aleman I. (2001 ). Circulating insulin-like growth factor I mediates exercise-induced increases in the number of new neurons in the adult hippocampus . J. Neurosci. 21 , 1628 –1634 . 11222653 van Praag H. Kempermann G. Gage F. H. (1999 ). Running increases cell proliferation and neurogenesis in the adult mouse dentate gyrus . Nat. Neurosci. 2 , 266 –270 . 10.1038/6368 10195220 Wasinski F. Bacurau R. F. Estrela G. R. Klempin F. Arakaki A. M. Batista R. O. . (2015 ). Exercise during pregnancy protects adult mouse offspring from diet-induced obesity . Nutr. Metab. (Lond). 12 , 56 . 10.1186/s12986-015-0052-z 26690877 Wasinski F. Bacurau R. F. Moraes M. R. Haro A. S. Moraes-Vieira P. M. Estrela G. R. . (2013 ). Exercise and caloric restriction alter the immune system of mice submitted to a high-fat diet . Mediators Inflamm. 2013 :395672 . 10.1155/2013/395672 23576853 Wong E. Y. H. Herbert J. (2006 ). Raised circulating corticosterone inhibits neuronal differentiation of progenitor cells in the adult hippocampus . Neuroscience 137 , 83 –92 . 10.1016/j.neuroscience.2005.08.073 16289354

PMC005xxxxxx/PMC5002408.txt

==== Front Front Plant SciFront Plant SciFront. Plant Sci.Frontiers in Plant Science1664-462XFrontiers Media S.A. 10.3389/fpls.2016.01281Plant ScienceOriginal ResearchTranscriptome Analysis of Pepper (Capsicum annuum) Revealed a Role of 24-Epibrassinolide in Response to Chilling Li Jie 1†Yang Ping 2†Kang Jungen 3Gan Yantai 45Yu Jihua 1Calderón-Urrea Alejandro 6Lyu Jian 1Zhang Guobin 1Feng Zhi 1Xie Jianming 1*1Department of Facility Horticulture Science, College of Horticulture, Gansu Agricultural UniversityLanzhou, China2Department of Crop Cultivation and Farming System, College of Agronomy, Gansu Agricultural UniversityLanzhou, China3Department of Vegetable Genetics and Breeding, Beijing Vegetable Research Center, Beijing Academy of Agriculture and Forestry SciencesBeijing, China4Semiarid Prairie Agricultural Research Centre, Agriculture and Agri-Food CanadaSwift Current, SK, Canada5Gansu Provincial Key Lab of Aridland Crop Science, Gansu Agricultural UniversityLanzhou, China6Department of Biology, California State University FresnoFresno, CA, USAEdited by: Claudio Bonghi, University of Padua, Italy Reviewed by: Antonio Ferrante, University of Milan, Italy; Rui Fan, Chinese Academy of Tropical Agricultural Science, China *Correspondence: Jianming Xie gsau23@163.comThis article was submitted to Crop Science and Horticulture, a section of the journal Frontiers in Plant Science †These authors have contributed equally to this work. 29 8 2016 2016 7 128110 5 2016 11 8 2016 Copyright © 2016 Li, Yang, Kang, Gan, Yu, Calderón-Urrea, Lyu, Zhang, Feng and Xie.2016Li, Yang, Kang, Gan, Yu, Calderón-Urrea, Lyu, Zhang, Feng and XieThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.Brassinosteroids (BRs) have positive effects on many processes during plant growth, development, and various abiotic stress responses. However, little information is available regarding the global gene expression of BRs in response to chilling stress in pepper. In this study, we used RNA sequencing to determine the molecular roles of 24-epibrassinolide (EBR) during a chilling stress response. There were 39,829 transcripts, and, among them, 656 were differently-expressed genes (DEGs) following EBR treatment (Chill+EBR) compared with the control (Chill only), including 335 up-regulated and 321 down-regulated DEGs. We selected 20 genes out of the 656 DEGs for RT-qPCR analysis to confirm the RNA-Seq. Based on GO enrich and KEGG pathway analysis, we found that photosynthesis was significantly up-enriched in biological processes, accompanied by significant increases in the net photosynthetic rate (Pn), Fv/Fm, and chlorophyll content. Furthermore, the results indicate that EBR enhanced endogenous levels of salicylic acid (SA) and jasmonic acid (JA) while suppressing the ethylene (ETH) biosynthesis pathway, suggesting that BRs function via a synergistic cross-talk with SA, JA, and ETH signaling pathways in response to chilling stress. In addition, EBR induced cellulose synthase-like protein and UDP-glycosyltransferase, suggesting a contribution to the formation of cell wall and hormone metabolism. EBR also triggered the calcium signaling transduction in cytoplasm, and activated the expression of cellular redox homeostasis related genes, such as GSTX1, PER72, and CAT2. This work, therefor, identified the specific genes showed different expression patterns in EBR-treated pepper and associated with the processes of hormone metabolism, redox, signaling, transcription, and defense. Our study provides the first evidence of the potent roles of BRs, at the transcription level, to induce the tolerance to chilling stress in pepper as a function of the combination of the transcriptional activities, signaling transduction, and metabolic homeostasis. pepperBrassinosteroidchill-stresstranscriptomeRNA sequencingNational Natural Science Foundation of China10.13039/50110000180931260493 ==== Body Introduction Chilling stress affects the geographical distribution of many important vegetables such as pepper (Capsicum annuum L.; Sanghera et al., 2011). This abiotic stress adversely affects plant growth and development and the yield and quality of many crops (Janská et al., 2010; Ren et al., 2014). In Northern China, where greenhouse cultivation is used for vegetable production, it is very common for plants suffer a chilling injury during the fall-to-winter transition period when a sudden temperature drops often leads to early frost damage. In vegetables, low temperatures lead to the arrest of growth or abortion of flower buds, resulting in significant yield and economic losses (Chinnusamy et al., 2007). Plants have evolved pleiotropic and intricate regulatory functions to defend against environmental stresses (Xia et al., 2009). Under stress, plants produce a number of phytohormones, such as salicylic acid (SA), brassinosteroids (BRs), and abscisic acid (ABA), which play a critical role in the perception of external signals and the activation of defense mechanism within plants. Brassinosteroids, a group of naturally occurring plant steroids, have been shown to provide positive effects on the regulation of plant growth and a broad spectrum of physiological responses to abiotic stresses, such as high and low temperature stress (Mazorra et al., 2002; Bajguz, 2009), drought (Yuan et al., 2010), and salinity injury (Liu et al., 2014). For example, BRs increase thermotolerance of plants by inducing heat shock protein synthesis and gene expression for heat tolerance (Dhaubhadel et al., 2002; Dhaubhadel and Krishna, 2008). BRs induce cold-related gene expression in Brassica napus and Arabidopsis thaliana (Kagale et al., 2007). Ubiquitin-conjugating enzyme (UBC32) as a critical gene involved in BRI1 biosynthesis and ER-associated protein degradation (ERAD) pathway positively regulates BR-induced salt tolerance (Cui et al., 2011). A synergistic interaction among BR signaling and the production of reactive oxygen species (ROS) induces the gene expression of respiratory burst oxidase homolog (RBOH), whereas encoding NADPH oxidase and NADPH oxidase trigger apoplastic ROS accumulation, activating MAPKs to increase plant tolerance to stress (Hao et al., 2013). RNA sequencing techniques have been used to investigate global expression profiles and reveal the signal transduction pathways involved in the resistance network under various stresses (Liu et al., 2015; Wang J. et al., 2015). The genome of pepper has been recently sequenced (Qin et al., 2014), which provides a valuable resource for molecular-based investigations for stress tolerance in plants. In previous studies, we found that exogenous BRs alleviated low temperature stress in pepper by enhancing antioxidant capacity and maintenance of photosystem II (Li et al., 2015a,b). However, it is unknown regarding the specific gene expression profile of BRs-induced chilling tolerance in pepper, and the genomic characteristics of the BRs-induced tolerance were undefined. Here, we reveal the genes associated with chilling stress and the associated signaling pathways mediated by BRs using RNA-seq analysis. The goals of the present study were to (i) provide insights into the pepper leaf transcriptome response to BRs under chilling stress; and (ii) uncover the genes and pathways that are associated with BRs-induced stress tolerance in pepper. Materials and methods Plant material and stress treatment Pepper seeds (cv. “Xiangyan NO.16”) were germinated in the dark for 72 h at 28°C before being transplanted into plastic pots containing a mixture of vermiculite and peat (1:2, v: v) for subsequent growth. The seedlings were grown in an intelligent greenhouse with 25/15°C (day/night), photon flux density of 350–400 μmol m−2 s−1, 12-h photoperiod, and relative humidity of 60–80%. We used six different concentrations (0, 10, 1, 0.1, 0.01, and 0.001 μM) to manipulate BRs levels in pepper plants as described by Li et al. (2015b). We used 0.1 μM EBR as an optimum EBR concentration based on the results of the prior experiment where various concentrations were studied and the optimum level was identified. At the 6–7 leaf stage (50 days after planting), the seedlings were divided into the following two groups: one was sprayed with 0.1 μM EBR solution, and the other was sprayed with the same volume of double distilled water. Twenty-four hours later, seedlings in both groups were transferred to a controlled growth chamber with temperature at 15/5°C (day/night), photon flux density at 100 μmol m−2 s−1, 12-h photoperiod, and relative humidity of 80%. Each treatment was replicated three times at the growth chamber, and each replicate had 20 plants. At day 7 of chilling, we collected two biological replicates of each treatment for sequencing, and each biological replicate had three plants. All samples were collected at the same time, ground into powder in liquid nitrogen and stored at −80°C for further use. These samples were labeled Chill+EBR (EBR treatment under chilling stress) and Chill (control). RNA extraction and RNA-seq Total RNA was extracted from the Chill and Chill+EBR samples (leaves) with Trizol reagent (Invitrogen, Carlsbad, CA, USA) using method described by Hu et al. (2012). The total RNA extraction was divided into two aliquots; one aliquote was used for RNA-sequencing, and the other was used for real-time PCR. The RNA-sequencing was performed using an Illumina HiSeq 4000 platform (Illumina, San Diego, CA, USA) at Novogene Bioinformatics Technology Co., China. Each sample generated more than 6 gigabyte of data. The clean reads were filtered from raw sequencing data and the low-quality reads containing unknown nucleotides or adaptor sequences were removed; this procedure was performed in accordance with the method of Chen et al. (2014). The filtered clean reads were aligned to C. annuum reference genome (http://peppersequence.genomics.cn/page/species/download.jsp). Differential expression analysis We performed differential expression analysis for both Chill and Chill+EBR treatments based on the DESeq R package, which allowed for statistical analysis using the negative binomial distribution model (Wang et al., 2010). To control the false discovery rate, we adjusted the resulting p-values according to Benjamini and Hochberg's approach (Benjamini and Hochberg, 1995), where an adjusted p < 0.05 is accepted to represent differentially expressed genes (DEGs). We performed gene ontology (GO) enrichment analysis of the DEGs according to the GOseq R package, and GO terms with q < 0.05 were regarded as significantly enriched (Young et al., 2010). We carried out the statistical enrichment of the differential expression genes in Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways using KOBAS software (Xie et al., 2011). Validation of DEGs by real-time quantitative PCR Twenty transcript genes were selected for the qRT-PCR assay; the genes and gene-specific primers used are summarized in Table S1. Actin was used as an internal reference. qRT-PCR was performed using SYBR-Green (ABI-Invitrogen, California, USA) on an ABI 7900 Fast Real-Time PCR Detection System (Applied Biosystems, Carlsbad, USA). A real-time RT-PCR reaction (20 μl) included 10 μl of 2 × SuperReal PreMix Plus, 2 μl cDNA, 1 μl of each primer and 6 μl ddH2O, and it proceeded for 40 cycles. The relative expression levels of the selected 20 genes normalized to the expression level of actin (internal reference control) were calculated from cycle threshold values using the 2−ΔΔCt method (Livak and Schmittgen, 2001). Chlorophyll content and net photosynthetic rate The chlorophyll a and b concentrations were determined according to the method of Arnon (1949). Net photosynthetic rate (Pn) in the fully expanded leaves of pepper plants was measured by Ciras-2 portable photosynthesis system (PP Systems, USA). Chlorophyll fluorescence determination The chlorophyll fluorescence parameters in pepper leaves were measured using a pulse-modulated fluorometer (FMS-2, Hansatech, Norfolk, UK). Pepper seedlings were placed in the dark for 30 min and were prepared for the determination of the Fv/Fm. Analysis of hormones and hormone metabolites Pepper leaf tissues treated with DDH2O or 0.1 μM EBR under chilling stress were collected in liquid nitrogen and stored at −80°C. For analysis of IAA and ETH metabolites, SA and JA, leaf tissues were lyophilized using a freeze dryer; these procedures were in accordance with the methods of published research (Chiwocha et al., 2003; Peng and Zhou, 2009). Indole acetic acid oxidase (IAAO) activity was measured on acetone extract leaves by measuring residual IAA following incubation and agitation in the dark at 30°C. One unit of IAAO activity was expressed as 1 mg of IAA destroyed per milligram of protein per minute. ACC content and ACC synthase (ACS) activity was determined according to the method of Prasad and Cline (Prasad and Cline, 1987). Results Mapping and quantitative assessment of iiiumina sequence We constructed two libraries from Chill and Chill+EBR for RNA-Seq. A total of 32.49 million (Chill) and 39.86 million (Chill+EBR) reads were generated. After removing low-quality regions, adapters, and possible contamination, we obtained more than 3 giga base clean bases with a Q20 percentage over 92%, Q30 percentage over 86%, and a GC percentage between 42.9 and 43.7% (Table 1). Table 1 Summary of sequence assembly after illumina sequencing. Sample namea Raw reads Clean reads Clean bases Gb Error rate (%) Q20b(%) Q30c(%) GC contentd(%) Chill1-1 33,040,705 32,108,541 4.01 0.04 94.63 89.54 42.91 Chill1-2 33,040,705 32,108,541 4.01 0.04 93.08 87.4 42.92 Chill2-1 31,938,987 30,969,153 3.87 0.04 94.44 89.22 43.38 Chill2-2 31,938,987 30,969,153 3.87 0.04 92.48 86.51 43.41 Chill+EBR1-1 41,024,790 40,319,076 5.04 0.04 94.22 88.8 43.63 Chill+EBR1-2 41,024,790 40,319,076 5.04 0.04 92.22 86.11 43.71 Chill+EBR2-1 38,701,387 37,690,638 4.71 0.04 95.02 90.2 42.95 Chill+EBR2-2 38,701,387 37,690,638 4.71 0.04 92.44 86.35 42.97 a The numbers 1 and 2 at the end of the sample name represent left and right ends (pair-end sequencing), respectively. Chill-samples treated with water; Chill+EBR-samples treated with 0.1 μM EBR. b Percentage of bases with a Phred value of at least 20. c Percentage of bases with a Phred value of at least 30. d Proportion of guanidine and cytosine nucleotides among total nucleotides. Each library that produced the clean reads was aligned to the recently released C. annuum reference genome, release_2.0 (Qin et al., 2014). The proportion of clean reads in the two pepper transcriptome libraries that mapped to C. annuum reference genome ranged from 85.34 to 87.60% (Table 2). A total of 39,829 genes were confirmed from the mapped libraries, including the locations of exons and introns (Table S2). All of the RNA-sequence data in this article have been deposited in the NCBI-SRA database and are accessible in SRX1959970. Table 2 Number of reads sequenced and mapped to the pepper genome. Sample name Chill1 Chill2 Chill+EBR1 Chill+EBR2 Total reads 64,217,082 61,938,306 80,638,152 75,381,276 Total mapped 56,251,137 (87.6%) 53,309,676 (86.07%) 68,819,267 (85.34%) 65,985,060 (87.54%) Multiple mapped 2,129,037 (3.32%) 2,064,848 (3.33%) 2,861,761 (3.55%) 28,11,891 (3.73%) Uniquely mapped 54,122,100 (84.28%) 51,244,828 (82.74%) 65,957,506 (81.79%) 63173169 (83.8%) Read-1 27,305,549 (42.52%) 25,918,892 (41.85%) 33,351,826 (41.36%) 32,030,587 (42.49%) Read-2 26,816,551 (41.76%) 25,325,936 (40.89%) 32605680 (40.43%) 31,142,582 (41.31%) Reads map to “+” 27,008,832 (42.06%) 25,526,323 (41.21%) 32,830,110 (40.71%) 31,512,085 (41.8%) Reads map to “−” 27,113,268 (42.22%) 25,718,505 (41.52%) 33127396 (41.08%) 31,661,084 (42%) Non-splice reads 36,188,485 (56.35%) 33853641 (54.66%) 43,682,428 (54.17%) 41,602,031 (55.19%) Splice reads 17,933,615 (27.93%) 17,391,187 (28.08%) 22,275,078 (27.62%) 21,571,138 (28.62%) Transcriptome profiles of the leaves from the two groups of pepper seedlings The 39,829 genes from the mapped libraries were normalized (Table S3) using the reads per kilo bases per million reads (RPKMs) method (Mortazavi et al., 2008). With RPKMs in 0 to 1, the genes were regarded as having a low expression level; genes with RPKMs of 3–15 were regarded as having a medium expression level; and genes with RPKM beyond 60 were regarded as having a very high expression level (Table 3). Table 3 Statistics of genes in different expression-level interval. RPKMa interval Chill Chill+EBR 0 ~ 1 17,437 (43.78%)b 17,267 (43.35%) 1 ~ 3 3859 (9.69%) 3925 (9.85%) 3 ~ 15 8283 (20.80%) 8415 (21.13%) 15 ~ 60 6932 (17.41%) 6954 (17.46%) >60 3319 (8.33%) 3269 (8.21%) a Reads per kilo bases per million reads. b Ratios of gene number to total gene number are presented in parentheses. Differentially expressed genes in the pepper leaves We identified 656 DEGs between Chill and Chill+EBR treatments (Table S4; Figure 1A). We used hierarchical clustering of all the DEGs to observe the gene expression patterns, and it was evaluated by log10 RPKMs for the two groups (Figure 1B). Compared to Chill treatment, genes in Chill+EBR treatment contained 335 up-regulated genes and 321 down-regulated genes. These results suggest that EBR had a markedly effect on the transcription of a subset of genes response for chilling stress. Figure 1 Transcriptome analysis of differentially expressed genes in Chill and Chill+EBR treatment of pepper leaves. (A) Volcano plot showed the DEGs between two different libraries. The q < 0.05 was used as thresholds to determine the significance of DEGs. Red dots represent up-regulated genes, green dots show down-regulated genes, and blue dots indicate transcripts that did not change significantly in the Chill+EBR library compared Chill. (B) Hierarchical clustering of all the DEGs based on log10 RPKM values. The color (from blue to red) represents gene expression intensity from low to high. Chill and Chill+EBR represent two treatments under alone chilling stress and chilling stress with 0.1 μM EBR. Real-time qPCR analysis To validate the DEG data from RNA-sequencing, we randomly selected 20 DEG for qRT-PCR assay in EBR-mediated chilling stress. The qRT-PCR results showed a strong positive correlation with the RNA-seq (R2 = 0.947), indicating that the RNA-seq data were validated (Figure 2). Figure 2 Correlation of RNA-seq (y axis) and qRT-PCR data (x axis), and the assay is carried out for 20 randomly selected DEGs. GO and KEGG enrichment analyses. We evaluated the gene functions of DEGs using GO enrichment analysis, and it revealed the biological process, molecular function and cellular component categories for the 656 DEGs of the two groups tested (Table S5). A total of 335 up-regulated DEGs were enriched significantly in 10 functional terms. Among these, six terms were significantly enriched in cellular component, there was one term significantly enriched in molecular function, and two terms were significantly enriched in biological process. Within in the cellular component domain, the terms that were significantly enriched Photosystem I reaction center (GO: 0009538; 8 genes), photosystem (GO: 0009521; 15 genes), photosynthetic membrane (GO: 0034357; 15 genes), thylakoid (GO: 0009579;15 genes), thylakoid part (GO: 0044436; 15 genes), and photosystem I (GO: 0009522; 8 genes). Within the molecular function domain, the term that was significantly enriched intransferring hexosyl groups (GO: 0016758) with 16 genes. Within the biological process domain, the terms that were significantly enriched included Photosynthesis (GO: 0015979) and DNA-dependent transcription (GO: 0006352) with 19 and 9 genes, respectively. For down-regulated DEGs, three terms were markedly enriched in molecular function, and 41 terms were markedly enriched in biological process (Figure 3). Figure 3 GO enrichment analysis of DEGs between Chill and Chill+EBR treatment. (A) up-regulation; (B) down-regulation. The 30 most enriched GO terms are shown. Asterisks indicate significantly enriched GO terms (q < 0.05). Among DGEs of the two groups tested the following 9 pathways, with a KEGG pathway annotation, were affected: carbon fixation in photosynthetic organisms, photosynthesis, photosynthesis-antenna proteins, carbon metabolism, pentose phosphate pathway, glyoxylate and dicarboxylate metabolism, metabolic pathways, zeatin biosynthesis, glycine, serine, and threonine metabolism pathway (q < 0.05; Table S6; Figure 4). Figure 4 KEGG pathway enrichment analysis of DEGs between Chill and Chill+EBR treatment. The left Y-axis shows the KEGG pathway. The X-axis shows the Rich factor. A high q-value is represented by blue and a low q-value is represented by red (q < 0.05). Effect of EBR on chlorophyll content and net photosynthetic rate The exposure to chilling stress treatment influenced the morphological traits in pepper seedlings, and EBR application significantly alleviated the inhibited growth (Figure 5A). The net photosynthetic rate (Pn) was increased significantly 7 days after EBR-treatment compared to without EBR; this was accompanied by a remarkable increase in chlorophyll content (Figures 5D,E). EBR (Chill+EBR) attenuated the inhibitory effect of chilling on photosynthesis. Low temperature treatments resulted in the reduction of Fv/Fm in pepper seedlings, while the application of EBR significantly increased Fv/Fm (Figure 5C). Figure 5 Phenotypic changes (A), net photosynthetic rate (Pn) (B), Fv/Fm (C), chlorophyll a and chlorophyll b (D,E) in chill-stressed pepper seedlings with or without exogenous application of EBR. Asterisks above the histograms indicated significant differences between Chill and Chill +EBT by Student's t-test (**P < 0.01; *P < 0.05). Transcriptome profiles of photosynthesis-related genes The genes involved in the photosynthesis in the EBR-treated chilling responses were all up-regulated, and among them 8 genes encoded photosystem I reaction center subunit. It is known that the chlorophyll a/b-binding protein is an important component of light-harvesting complex II (Jansson, 1999, 1994). We found that EBR induced the expression of 10 genes encoding chlorophyll a/b-binding protein. ATP-dependent zinc metalloproteases, known to take part in chloroplast protein modification and the metabolism of extracellular matrix, were also up-regulated by EBR. Additionally, the genes involved in ATP synthase chain, oxygen-evolving enhancer protein, ABC transporter I family member, thylakoid lumenal protein, and PsbP domain-containing protein were all up-regulated, suggesting that BRs might play a significant and positive role in the photosynthesis processes under chilling stress (Table 4). Table 4 Photosynthesis related genes expression in pepper leaves as influenced by chilling alone or in combination EBR treatment. Gene ID Log2FC q-value Symbol Description Capana06g001915 0.786 6.73E-03 psaD Photosystem I reaction center subunit II Capana02g001321 1.198 2.76E-02 PSAF Photosystem I reaction center subunit III Capana06g000155 1.229 3.90E-02 PSAEA Photosystem I reaction center subunit IV A Capana04g000192 0.798 6.55E-03 PSAEA Photosystem I reaction center subunit IV A Capana06g001274 1.170 1.58E-02 PSAH Photosystem I reaction center subunit VI Capana03g000290 0.700 2.66E-02 PSAH Photosystem I reaction center subunit VI Capana06g000231 0.733 1.60E-02 PSAL Photosystem I reaction center subunit XI Capana06g000232 1.184 2.77E-06 PSAL Photosystem I reaction center subunit XI Capana10g002492 1.216 2.83E-03 PSBY Photosystem II core complex proteins psbY Capana09g002353 1.130 2.36E-02 psbW Photosystem II reaction center W protein Capana09g000146 1.218 4.15E-02 CB12 Chlorophyll a/b binding protein Capana07g001245 1.480 2.70E-10 CAB4 Chlorophyll a/b binding protein 4 Capana08g000250 1.106 1.82E-03 CAB7 Chlorophyll a/b binding protein 7 Capana08g001648 1.626 9.29E-08 CAB8 Chlorophyll a/b binding protein 8 Capana08g001647 1.834 4.99E-05 CAB8 Chlorophyll a/b binding protein 8 Capana00g002799 3.279 3.52E-02 CAB21 Chlorophyll a/b binding protein 21 Capana00g002801 2.960 1.58E-02 CAB21 Chlorophyll a/b binding protein 21 Capana09g000473 2.058 4.26E-03 CAB37 Chlorophyll a/b binding protein 37 Capana09g001520 0.956 2.74E-04 LHCB4.2 Chlorophyll a/b binding protein CP29.2 Capana01g002398 1.342 1.11E-03 LHCB5 Chlorophyll a/b binding protein CP26 Capana03g002052 0.915 1.06E-02 FTSH11 ATP-dependent zinc metalloprotease FTSH 11 Capana02g002193 0.690 2.66E-02 ATPC ATP synthase gamma chain Capana05g001562 0.768 2.19E-02 ATPD ATP synthase delta chain Capana00g001213 0.908 7.37E-04 PSBP Oxygen-evolving enhancer protein 2 Capana02g002133 1.140 7.25E-04 PSBQ2 Oxygen-evolving enhancer protein 3-2 Capana06g001078 1.313 1.42E-02 ABCI17 ABC transporter I family member 17 Capana00g004571 1.295 1.83E-04 At1g03600 Thylakoid lumenal protein At1g03610 Capana04g000971 1.510 4.25E-03 CLEB3J9 Thylakoid lumenal 29 kDa protein Capana06g001397 1.348 8.68E-05 PPD6 PsbP domain-containing protein 6 Shown are the q-value (< 0.05) for genes expression (comparison between EBR+Chill and Chill). Effect of EBR on the endogenous levels of other hormones and their metabolites To determine whether BR influenced the endogenous levels of other hormones, we assessed the concentrations of IAA, ABA, JA, and SA and ETH metabolites in the leaves of pepper seedlings grown in the absence or presence of 0.1 μM EBR under chilling stress. The concentrations of ABA, SA and JA were significantly increased by 37.5, 189.1, and 132.3%, respectively, in the EBR-treated seedlings compared to the untreated seedlings (Figure 6A). There was an increase in the level of IAA, but a significant decrease in the activity of acetic acid oxidase (IAAO) in the EBR-treated tissue (Figure 6B). There was a significant increase in the activity of ACC synthase (ACS) in the EBR-treated seedlings compared to chilling stress without EBR (Figure 6B). Figure 6 Endogenous hormone profiles of pepper seedlings grown in the only chill stress (Chill) or chill stress with 0.1 μM EBR (Chill+EBR). (A) Contents of ABA, SA and JA in leaves of pepper seedlings. (B) Contents and metabolites of IAA and ETH in leaves of pepper seedlings. Error bars represent the standard error (SE) of the mean for three replicates. Asterisks above the histograms indicated significant differences between Chill and Chill +EBT by Student's t-test (**P < 0.01; *P < 0.05). Transcriptome profiles of hormone metabolism and signaling-related genes EBR treatment under chilling stress induced changes in the expression of some genes involved in auxin signaling pathways, including auxin-induced protein, auxin-responsive protein, and auxin-binding protein (Table 5). Genes involved in cytokinin dehydrogenase, which plays a significant role in maintaining the well-organized cytokinin functions (Werner et al., 2003), were down-regulated by EBR in pepper. Protein phosphatase type 2C24 (P2C24) was identified as the second component of the ABA signaling pathway, which was also down-regulated by EBR. JA biosynthesis gene, linoleate 13S-lipoxygenase 2-1, and SA-related isochorismate synthase gene were all up-regulated, while four gibberellin-responsive genes were down-regulated. EBR treatment under chilling stress down-regulated all of the genes in the ETH signaling pathways. These genes included the ETH biosynthesis gene ACC synthase (ACS1), the induction of the ETH receptor (ETR), 8 ETH responsive transcription factors and one ETH insensitive 3-like 1 protein. Table 5 Hormone metabolism and signaling-related gene expression in pepper leaves as influenced by chilling alone or in combination EBR treatment. Gene ID Log2FC q-value Symbol Description AUXIN Capana09g001771 −1.705 9.66E-03 Unknown Auxin-induced protein 6B Capana03g004288 1.079 6.65E-03 Unknown Auxin-induced protein PCNT115 Capana03g000310 −0.888 1.67E-03 IAA17 Auxin-responsive protein IAA17 Capana07g000961 1.786 2.76E-02 ABP19A Auxin-binding protein ABP19a CYTOKININ Capana00g003808 −1.856 1.99E-02 CKX6 Cytokinin dehydrogenase 6 ABSCISIC ACID Capana03g002801 −1.084 2.66E-02 P2C24 Probable protein phosphatase 2C 24 JASMONIC ACID Capana01g001577 6.419 2.41E-76 LOX2.1 Linoleate 13S-lipoxygenase 2-1 SALICYLIC ACID Capana06g001004 1.021 4.65E-02 ICS Isochorismate synthase GIBBERELLIN Capana03g001010 −2.357 5.88E-11 GASA14 Gibberellin-regulated protein 14 Capana01g002809 −1.463 0.011357 GA2OX2 Gibberellin 2-beta-dioxygenase 2 Capana03g001010 −2.357 5.88E-11 GASA14 Gibberellin-regulated protein 14 Capana05g000709 −2.415 7.10E-05 GA2OX7 Gibberellin 2-beta-dioxygenase 8 ETHYLENE Capana01g000175 −1.589 1.73E-02 ACS1 1-aminocyclopropane-1-carboxylate synthase Capana03g004532 −1.636 4.05E-02 ETR2 Ethylene receptor 2 Capana06g000678 −0.805 7.97E-03 EIL1 Ethylene insensitive 3-like 1 protein Capana00g003399 −0.945 2.40E-03 ERF061 Ethylene-responsive transcription factor ERF061 Capana01g000662 −1.770 2.69E-02 ERF5 Ethylene-responsive transcription factor 5 Capana03g000085 −0.718 2.11E-02 RAP2-12 Ethylene-responsive transcription factor RAP2-12 Capana04g001107 −0.941 8.75E-04 RAP2-4 Ethylene-responsive transcription factor RAP2-4 Capana05g001701 −1.143 2.51E-05 ERF1B Ethylene-responsive transcription factor 1B Capana05g001951 −1.478 6.45E-08 ERF106 Ethylene-responsive transcription factor ERF106 Capana07g001714 −1.603 4.93E-03 ERF4 Ethylene-responsive transcription factor 4 Capana10g000499 −1.353 3.63E-08 ERF4 Ethylene-responsive transcription factor 4 OTHER SIGNALING-RELATED GENE Capana02g002718 −3.045 2.50E-04 LEA29 Late embryogenesis abundant protein D-29 Capana00g002896 −0.777 1.00E-02 POX2 Proline dehydrogenase 2, mitochondrial Capana01g002720 −0.997 3.08E-04 PERK1 Proline-rich receptor-like protein kinase PERK1 Capana04g000460 1.357 1.36E-04 PERK10 Proline-rich receptor-like protein kinase PERK10 Capana01g004123 −2.109 3.57E-03 Unknown 14 kDa proline-rich protein DC2.15 Capana03g000595 −1.306 7.23E-05 RBOHC Respiratory burst oxidase homolog protein C Capana12g000877 −1.355 3.59E-02 CRK25 Cysteine-rich receptor-like protein kinase 25 Capana12g000878 −0.936 4.87E-02 CRK10 Cysteine-rich receptor-like protein kinase 10 Capana00g004816 2.524 2.82E-06 CRK25 Cysteine-rich receptor-like protein kinase 25 Capana06g002351 0.851 4.51E-03 ATL72 RING-H2 finger protein ATL72 Capana00g003879 −1.036 4.84E-02 ATL78 RING-H2 finger protein ATL78 Capana01g002369 −1.149 5.76E-05 ATL2 RING-H2 finger protein ATL2 Capana10g001323 −1.6687 9.52E-06 HSP18.1 18.1 kDa class I heat shock protein (Fragment) Capana05g002063 −0.9636 0.013016 DREB2C Dehydration-responsive element-binding protein Capana01g002345 −0.89067 0.02905 DI19-3 Protein Dehydration-induced 19 homolog 3 Capana00g003859 −1.3668 0.02239 At5g39030 Probable receptor-like protein kinase At5g39030 Capana08g002140 −0.8267 0.01623 IKU2 Receptor-like protein kinase HAIKU2 Capana12g002105 0.941 5.94E-06 TMK1 Probable receptor protein kinase TMK1 Shown are the q-value (< 0.05) for genes expression (comparison between EBR+Chill and Chill). Late embryogenesis abundant protein D-29, which is involved in the tolerance to water stress resulting from desiccation or chilling injury was down-regulated by EBR. The genes involved in proline responses including proline dehydrogenase and proline-rich receptor-like protein kinase PERK1 (PERK1) were down-regulated, while PERK10 was up-regulated. RBOHC which is associated with ROS production in plants, was down-regulated by EBR. There are three genes related to cysteine-rich receptor-like protein kinase, among them two genes were down-regulated, and the other gene was up-regulated. The RING-H2 finger protein ATL72 was up-regulated, while the RING-H2 finger protein ATL78 and RING-H2 finger protein 2 were down-regulated. The genes may be involved in the early steps of the plant defense signaling pathway. Dehydration-responsive element-binding protein (DREB2C) and protein dehydration-induced 19 homolog (3DI19-3) which mediate cold-inducible transcription, were down-regulated by BRs (Table 5). Transcriptome profiles of calcium signaling and redox-related genes We found that EBR treatment under chilling stress up-regulated calcium-dependent protein kinase (CAS, Capana00g001365), while CBL-interacting serine/threonine-protein kinase, calcium uniporter protein, probable calcium-binding protein CML15, calcium-binding protein CML38, and calmodulin-related protein were down-regulated (Table 6). Table 6 Calcium signaling and redox-related genes expression in pepper leaves as influenced by chilling alone or in combination EBR treatment. Gene ID Log2FC q-value Symbol Description CALCIUM Capana00g001365 0.727 1.88E-02 CAS Calcium sensing receptor Capana06g000284 −1.279 1.77E-06 CIPK11 CBL-interacting serine/threonine-protein kinase 11 Capana02g000787 −1.128 1.06E-05 MCU Calcium uniporter protein Capana03g000955 −1.945 4.74E-04 CML15 Probable calcium-binding protein CML15 Capana11g000436 −1.920 3.75E-07 CML38 Calcium-binding protein CML38 Capana10g002124 −1.084 8.67E-05 CAM53 Calmodulin-related protein REDOX Capana09g001740 1.765 8.55E-03 GSTX1 Probable glutathione S-transferase Capana02g002747 2.8381 5.52E-06 PER72 Peroxidase 72 Capana02g002452 1.149 4.89E-02 CAT2 Catalase isozyme 2 Capana04g000138 0.884 1.36E-03 CDSP32 Thioredoxin-like protein CDSP32 Capana01g004227 2.526 6.43 E-05 FAO4A Long-chain-alcohol oxidase FAO4A Capana00g002845 3.276 4.01E-02 Unknown NAD(P)H:quinone oxidoreductase Capana03g001927 1.616 1.49E-04 At1g06690 Uncharacterized oxidoreductase, chloroplastic Shown are the q-value (< 0.05) for genes expression (comparison between EBR+Chill and Chill). All of the genes involved in redox homeostasis were up-regulated by EBR (Table 6). GSTX1, PER72, and CAT2 inactivate endogenous epoxides and hydroperoxides, and are associated with secondary metabolites during oxidative stress. CDSP32, which encodes thioredoxin as a physiological electron donor to the BAS1 peroxiredoxin, participated in the defense against lipid peroxidation in photosynthetic membranes (Broin and Rey, 2003). FAO4A was involved in the omega-oxidation pathway of lipid degradation. At1g06690 and NAD(P)H quinone oxidoreductase played a positive role in the antioxidant defense by generating vitamin E and ubiquinone. Transcriptome profiles of transcription factors and post transcription We analyzed the function of transcription factors (TFs) in pepper based on their annotations in NCBI database. The majority of WRKY family TFs were down-regulated, such as WRKY11 [a negative regulator of resistance to Pst (Journot-Catalino et al., 2006)], WRKY41, WRKY6, WRKY40 [a transcriptional repressor in plant cells response to abscisic acid and abiotic stress (Chen et al., 2010)], WRKY33, WRKY28, and WRKY20. Only WRKY51 (Capana12g001826), which positively mediated the signaling transduction of SA and JA (Gao et al., 2011), was up-regulated by EBR under chilling stress. The MYB family, which is involved in a diversity of gene regulation (Jackson et al., 1991), and NAC family members (NAC100, NAC002, and NAC072) were all down-regulated by EBR under chilling stress. bHLH family members were up-regulated except bHLH130. bHLH transcription factors positively regulated ABA-responsive kinase substrate (AKS) and facilitated stomatal opening by triggering the phosphorylation of AKS family transcription factors (Takahashi et al., 2013). Heat-stress transcription factor A-2/B-1, the key regulator alleviating oxidative damage caused by heat stress negatively regulated by cold stress in Arabidopsis (Zhang et al., 2009), was down-regulated by EBR in pepper. COL2, a circadian clock that controls many rhythmic processes (Kim et al., 2013), was up-regulated. MGP, which regulated tissue boundaries and asymmetric cell division and controlling SHORT-ROOT activity in a transcriptional and protein interaction network (Welch et al., 2007), was down-regulated by EBR. Additionally, ATHB-7 mediates a drought response via transcriptional regulation in an ABA-dependent manner. ATHB-52 and ATHB-21, which responded to auxin, ETH, and ABA or water deficits (Henriksson et al., 2005), were down-regulated by EBR. TCP19, involved in the orchestrated regulation of ICS1 expression (Wang X. et al., 2015), was down-regulated by EBR. GATA22, which participates in the GA-mediated signaling pathway (Richter et al., 2010) and chlorophyll biosynthetic process (Hudson et al., 2011), was up-regulated by EBR in pepper under chilling stress (Table 7). Table 7 Transcription factors and post-transcription related genes expression in pepper leaves as influenced by chilling alone or in combination EBR treatment. Gene ID Log2FC q-value Symbol Description TRANSCRIPTION FACTORS Capana00g003083 −1.152 4.21E-04 WRKY11 Probable WRKY transcription factor 11 Capana01g004472 −1.627 5.85E-07 WRKY41 Probable WRKY transcription factor 41 Capana02g002230 −1.123 9.79E-05 WRKY6 WRKY transcription factor 6 Capana03g000473 −1.909 7.33E-10 WRKY40 Probable WRKY transcription factor 40 Capana06g001110 −1.726 2.04E-12 WRKY40 Probable WRKY transcription factor 40 Capana06g001506 −1.139 5.55E-03 WRKY33 Probable WRKY transcription factor 33 Capana09g001251 −1.221 2.94E-05 WRKY33 Probable WRKY transcription factor 33 Capana07g001968 −1.071 8.10E-03 WRKY28 Probable WRKY transcription factor 28 Capana07g002350 −1.064 8.71E-05 WRKY20 Probable WRKY transcription factor 20 Capana12g001826 1.562 3.98E-02 WRKY51 Probable WRKY transcription factor 51 Capana02g002068 −1.643 8.43E-05 MYB39 Transcription factor MYB 39 Capana02g003369 −1.485 2.37E-05 MYB306 Myb-related protein 306 Capana03g000766 −1.099 4.99E-05 MYB306 Myb-related protein 306 Capana01g000650 −1.645 8.28E-05 NAC100 NAC domain-containing protein 100 Capana03g000802 −0.837 2.61E-02 NAC100 NAC domain-containing protein 100 Capana05g000569 −1.148 1.47E-05 NAC002 NAC domain-containing protein 2 Capana07g002219 −0.891 1.24E-03 NAC072 NAC domain-containing protein 72 Capana09g000936 −0.804 4.78E-03 NAC072 NAC domain-containing protein 72 Capana01g002561 −0.695 4.11E-02 bHLH130 Transcription factor bHLH130 Capana02g000436 1.652 1.35E-02 bHLH64 Transcription factor bHLH64 Capana03g002644 1.181 7.40E-03 bHLH51 Transcription factor bHLH51 Capana10g000647 1.036 1.36E-03 bHLH14 Transcription factor bHLH14 Capana02g001490 −1.427 1.75E-02 HSFB1 Heat stress transcription factor B-1 Capana07g000898 −1.210 6.80E-05 HSFA2 Heat stress transcription factor A-2 Capana02g003199 1.028 3.06E-02 COL2 Zinc finger protein CONSTANS-LIKE 2 Capana04g000388 −0.939 1.23E-02 MGP Zinc finger protein MAGPIE Capana01g000046 −1.858 2.34E-08 ATHB-7 Homeobox-leucine zipper protein ATHB-7 Capana03g001926 −0.701 3.71E-02 ATHB-7 Homeobox-leucine zipper protein ATHB-7 Capana03g002675 −0.865 2.69E-02 ATHB-52 Homeobox-leucine zipper protein ATHB-52 Capana04g002305 −1.170 1.44E-05 ATHB-21 Homeobox-leucine zipper protein ATHB-21 Capana01g000353 −0.935 2.89E-02 TCP19 Transcription factor TCP19 Capana09g000221 2.084 6.49E-04 GATA22 Putative GATA transcription factor 22 POST- TRANSCRIPTION Capana05g002235 1.140 3.19E-03 CSLE6 Cellulose synthase-like protein E6 Capana07g001101 −0.727 2.42E-02 CSLG1 Cellulose synthase-like protein G1 Capana07g001384 0.906 1.45E-02 CSLH1 Cellulose synthase-like protein H1 Capana02g001405 −1.297 1.03E-04 UGT91C1 UDP-glycosyltransferase 91C1 Capana06g000316 1.161 3.07E-02 UGT87A2 UDP-glycosyltransferase 87A2 Capana10g001522 1.364 1.56E-07 UGT73C3 UDP-glycosyltransferase 73C3 Capana10g001627 1.965 5.76E-05 UGT76E1 UDP-glycosyltransferase 76E1 Capana11g001875 1.222 2.33E-02 GT6 UDP-glucose flavonoid 3-O-glucosyltransferase 6 Capana12g000737 1.058 3.71E-02 GT7 UDP-glucose flavonoid 3-O-glucosyltransferase 7 Capana12g002768 0.968 1.53E-03 GT3 UDP-glucose flavonoid 3-O-glucosyltransferase 3 Capana06g002960 1.187 1.08E-05 GT4 Putative UDP-rhamnose:rhamnosyltransferase 1 Capana10g000708 −0.976 6.90E-04 UPTG2 Alpha-1,4-glucan-protein synthase [UDP-forming] 2 Shown are the q-value (< 0.05) for genes expression (comparison between EBR+Chill and Chill). EBR treatment induced CSLE6 and CSLH1 up-regulation, but down-regulated CSLG1, and it is known that they encode various non-cellulosic β-linked polysaccharides synthesis enzymes involved in the backbone of the cell wall (Doblin et al., 2009). Four genes encoding glycosyltransferase, which catalyzed the receptor substrates of cytokinin, auxin and ABA, were up-regulated. Similarly, three genes encoding UDP-glucose flavonoid 3-O-glucosyltransferase, GT6, GT7, and GT3, were up-regulated by EBR (Table 7), and it is known that they are also involved in the detoxification of xenobiotics. Discussion BRs, a group of naturally occurring plant steroids, are involved in many important cellular and physiological processes of plants in response to environmental stresses (Xia et al., 2009). Additionally, BRs mediate the response and signal transduction pathways of multiple hormones such as salicylic acid (SA), ethylene (ETH), jasmonic acid (JA), or abscisic acid (ABA) to abiotic stresses (U. K. Divi et al., 2010). In a previous study, we demonstrated that EBR treatment increased the basal chill tolerance of pepper via physiological and biochemical methods (Li et al., 2015a). To identify BR-mediated changes in gene expression, in the present study, we revealed the potent role of BRs in the response to chilling stress at the transcriptome level, and the findings provide the groundwork for elucidating the mechanism of BRs-induced chilling tolerance in pepper plants. We identified 656 DEGs, including 335 up-regulated and 321 down-regulated genes in the EBR-treated tissues compared to the chilling stress without EBR. Photosynthesis is among the primary processes in plants that are often affected by chilling stress (Allen and Ort, 2001). In the present study, we used GO enrichment analysis and found that the 29 genes involved in photosynthesis were up-regulated by EBR under chilling stress conditions. Indeed, EBR-treated seedlings under chilling stress significantly enhanced net photosynthetic rate (Pn) and chlorophyll content when compared to controls (Figure 5). This suggests that BRs induction of chlorophyll and the net photosynthetic rate is a major factor contributing to the distinct photosynthetic characteristics in the BRs transcriptome. We identified several up-regulated genes associated with chloroplast organization (Capana02g002385) and the photosynthetic apparatus including the chloroplast and thylakoid luminal. In addition, we found that BRs also significantly increased Fv/Fm under chilling stress. These results suggest that EBR treatment contributes to the pepper leaves increased ability to absorb and transfer light energy in chilling conditions. Indeed, the 5 up-regulated genes were associated with redox regulation of photosystem II (Capana06g001397), transport activity (Capana06g001078, Capana03g002052), and photosystem II reaction center (Capana09g002353, Capana10g002492). KEGG pathway analysis also showed that the photosynthesis (ko00195) and photosynthesis-antenna proteins (ko00196) were significantly enriched up-regulated pathway terms (Figure 4). As small molecules, plant hormones mediate many cellular processes in plants, including plant morphogenesis, and responses to changing environmental conditions (Kang et al., 2005; Pantin et al., 2013; Colebrook et al., 2014). Signal transduction components perceive plant hormone signals and transmit to the nuclear to induce gene expression synergistically with other signals to induce gene expression synergistically with other signals to influence response to environmental stress via a series of physiological processes (Lu et al., 2014). In our study, some unigenes were markedly enriched in the hormone biosynthesis and signaling components processes. These hormones including auxin, JA, ETH, ABA, and SA, and their signaling components were involved in BR-induced plant defense. Our results showed that the auxin, JA, and SA pathways appeared to act synergistically with BRs in mediating the response to chilling stress (Table 5). It is known that BRs can induce the SA-perceptive pathway or ABA-dependent pathways for resistance to heat and salt stress in A. thaliana or Chlorella vulgaris (Bajguz, 2009; Divi et al., 2010). We demonstrated the effects of BRs on chilling tolerance via the transcriptome profiles of hormone metabolism and signaling-related components in pepper. Accordingly, the transcription involved in hormone signaling components may function as important mediators of BR-induced resistance to chilling stress. We found that four DEGs were associated with auxin biosynthesis and signaling pathways, and the BRs mediated the down-regulation of auxin-induced protein B6 and IAA17. It is known that auxin-induced protein B6 induced the auxin-activated signaling pathway, whereas IAA17 functioned as a repressor of early auxin response genes (Liscum and Reed, 2002). Furthermore, we found that EBR-treated seedlings had significantly enhanced IAA content, but decreased the activity of acetic acid oxidase (IAAO). It was speculated that BRs reduced sensitivity to IAA by down-regulating genes encoding IAA conjugates. We found the down-regulation of genes involved in cytokinin; EBR treatment did not affect the endogenous levels of cytokinins (data not shown), but it induced cytokinin signaling in pepper. We speculated that BRs have a negative feedback loop to ensure the proper regulation of cytokinin functions, supported by previous observations (Werner et al., 2003). In addition, we found that ABA levels were increased by 37.5% in EBR-treated seedlings (Figure 6). Intriguingly, the gene P2C24, encoding the second component of the ABA signaling pathway, was down regulated. Indeed, Yu et al. (2011) demonstrated some ABA response genes as the direct targets of BES1 in ChIP-chip analysis. Additionally, BRs can inhibit ABA effects. For example, BRs acted as an opposing factor, providing inhibitory effects of ABA on seed germination (Divi and Krishna, 2010). The molecular mechanism is that BRs inhibit BRASSINOSTEROID INSENSITIVE2 (BIN2), thus, repressing the BIN2-ABI5 (ABSCISIC ACID INSENSITIVE5) cascade and antagonizing ABA inhibitory effect on germination (Hu and Yu, 2014). We found that JA and SA levels were significantly higher in the EBR-treated seedlings (Figure 6). Likewise, the JA biosynthesis gene, linoleate 13S-lipoxygenase 2-1, and SA-related gene, isochorismate synthase (ICS), were up-regulated. These results suggest that the possible cross-talk of BRs with JA and SA signaling pathways plays a positive role of mediating chilling stress responses in pepper. In the present study, we found that EBR treatment decreased ETH levels, reduced the content of ACC which is an ETH synthetic substance, and increased ACS activity. We also identified genes encoding ETH signaling components (ACS1) and encoding ETH-insensitiive 3-like protein and ethylene-responsive transcription factors (ERF), which were down-regulated by EBR. This finding indicates that BRs and ETH have an antagonistic relationship under chilling stress and that ETH contributes to the activation of the BR signaling pathway and increased chill tolerance. As secondary signaling molecules, ROS and calcium are crucial for plant defense against abiotic stresses. Here, we found that EBR up-regulated genes associated with cellular redox homeostasis, including glutathione S-transferase (GSTX1), peroxidase, catalase isozyme, and ferredoxin related genes (Table 6). It was reported that RBOHs are associated with ROS production in plants (Marino et al., 2012). We found that BRs triggered RBOHC down-regulation in pepper, which suggested that BRs reduced apoplastic ROS accumulation generated by NADPH oxidase so that the plants increased stress tolerance. Calcium is a key element in many cellular processes in plants, and calcium signaling is crucial for plant defense against abiotic stresses (Yuan et al., 2007; Boudsocq and Sheen, 2013). In our study, EBR up-regulated CAS, which modulated cytoplasmic Ca2+ concentrations for the induction of a series of biochemical reactions to chilling stress (Table 6). In addition, EBR-treated down-regulated the genes that encoded CBL-interacting serine/threonine-protein kinase, calcium-binding protein CML, and calmodulin-related protein. It was most likely that BR induced calmodulin-related signaling transduction, and Ca2+ was pumped quickly out of the cells and then returned to the Ca2+ pools through intracellular Ca2+-ATPase. Transcriptional facts (TFs) are very important for the combination of cis acting element in gene promoter elements and the mediation of the signaling pathway in response to stresses, such as WRKYs, MYBs, NACs, BHLHs, and ZFPs (Liu et al., 2008; Fujita et al., 2011; Shi and Chan, 2014; Shi et al., 2014). Several studies have shown that the expressions of C-repeat binding factors (CBFs) as master molecular switches are positively regulated by MYB56, ZFP1/182, and CAMTA1/2/3 and are negatively regulated by MYB15, MYBS3, WRKY34, and EIN3. The TFs specifically bind to the DRE/CRT (dehydration-responsive element/C-repeat element) cis-acting regulatory element of the promoter region of the cold-responsive genes, such as DHN (dehydrin) and RD (responsive to dehydration). In the present study, we found that BRs down-regulated WRKY11 and WRKY40 which mediated the metabolic pathway of ABA, whereas WRKY51, which positively mediated JA- and SA-signaling was the up-regulated by EBR under chilling stress. Additionally, BRs induced up-regulation of TFs bHLH involved in AKS, which facilitated stomatal opening by triggering the phosphorylation of AKS family transcription factors. We also found that BRs positively mediated TCP19 TF, which was associated with the orchestrated regulation of ICS1 expression and the GA-mediated signaling pathway (GATA22). A study on blueberry demonstrated that the genes encoding zinc finger proteins are associated with cold acclimation (Die and Rowland, 2014). In pepper, we found that EBR up-regulated COL2 and down-regulated MGP under chilling stress. These results indicate that BRs induced different expressions of zinc finger proteins in response to chilling stress. Although it was is unknown why some TFs are up-regulated while the others are down-regulated in DEGs, we assume that TFs coordinated the network regulation of transcriptional activities in multiple pathways to increase chilling tolerance in EBR-treated pepper. Glycosylation in plants is crucial not only for the regulation of cellular metabolism including plant hormones, secondary metabolites, and xenobiotics (Li et al., 2001), but also for the activity of several signaling molecules and defense compounds (Vogt and Jones, 2000). Sun et al. (2013) reported that glycosyltransferase could protect tobacco against salt stress, and the corresponding glycosyltransferase gene is essential for modifying cellular redox homeostasis under abiotic stress. In the present study, many glycosyltransferase genes were up-regulated by EBR (Table 7), which suggested that the up-regulation of the genes may confer cold tolerance in EBR-treated pepper. Cellulose synthase was also up-regulated, providing primary interface for plant environment interactions, and was associated the formation of cell wall for performing sophisticated strategies to respond to different environmental stresses. Taken together, the BR-induced cold stress signals are perceived by several receptors at the cell membrane, followed by calcium and hormones molecules transduction to activate downstream stress-responsive genes in response to chilling stress. Among these, BR-induced signaling pathways might regulate transcription or directly/indirectly interact with several other signaling networks. BR promoted the change of Ca2+ in the cytoplasm- and induced a series of biochemical reactions to regulate cellular redox homeostasis related genes, such as GSTX1, PER72, and CAT2. The key transcriptional factor bHLH of the JA signaling pathway interacted with MaICE1, and then activated the expression of downstream CBF related genes in response to chilling stress. JA, as the key upstream signal of ICE-CBF/DREB1 pathways, positively regulated cold-related gene expression in EBR-treated pepper and regulated GSTX1, PER72, and CAT2 expression. We also found that BR induced SA signaling for antistress effectsby positively regulating the expression of the CBF upstream gene ICE and that WRKY51 may be a potential point of cross-talking among JA, SA, and BR, where BR induces a subset of SA- or JA-responsive genes (Figure 7). BR negatively regulated the ETH signaling components and the TFs AP2/ERF, which indicated that BR induces the expression of cold related genes and the expression depends on the ETH signaling pathway. In addition, BR activated cellulose synthase-like protein (CSLE6 and CSLH1), which regulated the formation of a cell wall, and BR also activated UDP-glycosyltransferase genes associated with hormone metabolism, such BR glucoside, and other metabolic enzymes (e.g., PERK10, HSP, and LEA29). Figure 7 Model depicting BR-mediated chilling tolerance via Ca2+, JA, SA, ETH signaling. LOX, ICS, and ACS are jasmonic acid biosynthesis gene, salicylic acid-related gene isochorismate synthase, and ethylene biosynthesis genes ACC synthase, respectively. WRKY51, bHLH, and AP2/ERF are all transcriptional facts. CSL is cellulose synthase-like protein; UGT is UDP-glycosyltransferase. Cellular redox homeostasis related genes: GSTX1, PER72, and CAT2. CAMTA: CaM-binding transcription activator. Conclusion In the study, we analyzed the gene expression profiles of BRs induced chilling tolerance in pepper using RNA-seq analysis. Our results showed that EBR induced 656 differently expressed genes, including 335 up-regulated and 321 down-regulated DEGs. Using GO and KEGG pathway analysis, we found that EBR application under chilling stress positively regulated photosynthesis-related genes, cellulose synthase-like protein, UDP-glycosyltransferase, and cellular redox homeostasis-related genes (GSTX1, PER72, and CAT2). Moreover, we present a model to explain the possible cross-talk of BR with SA, ETH, and JA signaling pathways under BR-induced cold tolerance. Our study provides the first evidence of the potent roles of exogenous EBR at the transcriptional level, and the response to chilling stress in pepper involved the activation of extensive transcriptional activities, signaling transduction, and modulation of metabolic homeostasis. Author contributions JK and JX conceived and designed the experiments. Jie L and PY performed the experiments; Jie L, ZF, and Jia L analyzed the data. GZ and JY contributed reagents/materials/analysis tools. YG and AC helped perform the analysis with constructive discussions and language polished. JX and YG approved the final version. All authors have read and approved the final manuscript. Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. We are grateful to the funding provided by National Natural Science Foundation of China (NSFC) (project # 31260493). We also thank Mr. Dawuda M. M. and Dr. John Constable for critically reading the manuscript. Supplementary material The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2016.01281 Table S1 Gene primers used for qPCR validation in this study. Click here for additional data file. Table S2 Genes information. Click here for additional data file. Table S3 RPKM of all genes. Click here for additional data file. Table S4 Differential expressed genes in chilling stress with EBR-treated plant compared to only chilling treated plant. Click here for additional data file. Table S5 Gene ontology analyses of differential expressed genes regulated by the EBR + chilling and only chilling stress. Click here for additional data file. Table S6 KEGG pathway enrichment of differential expressed genes regulated by the EBR + chilling and only chilling stress. Click here for additional data file. ==== Refs References Allen D. J. Ort D. R. (2001 ). Impacts of chilling temperatures on photosynthesis in warm-climate plants . Trends Plant Sci. 6 , 36 –42 . 10.1016/S1360-1385(00)01808-2 11164376 Arnon D. I. (1949 ). Copper enzymes in isolated chloroplasts. Polyphenoloxidase in Beta vulgaris . Plant Physiol . 24 , 1 –15 . 10.1104/pp.24.1.1 16654194 Bajguz A. (2009 ). Brassinosteroid enhanced the level of abscisic acid in Chlorella vulgaris subjected to short-term heat stress . J. Plant Physiol. 166 , 882 –886 . 10.1016/j.jplph.2008.10.004 19062130 Benjamini Y. Hochberg Y. (1995 ). Controlling the false discovery rate: a practical and powerful approach of multiple testing . J. R. Stat. Soc. Ser. B (Methodological) 57 , 289 –300 . Boudsocq M. Sheen J. (2013 ). CDPKs in immune and stress signaling . Trends Plant Sci. 18 , 30 –40 . 10.1016/j.tplants.2012.08.008 22974587 Broin M. Rey P. (2003 ). Potato plants lacking the CDSP32 plastidic thioredoxin exhibit overoxidation of the BAS1 2-cysteine peroxiredoxin and increased lipid Peroxidation in thylakoids under photooxidative stress . Plant Physiol. 132 , 1335 –1343 . 10.1104/pp.103.021626 12857815 Chen H. Lai Z. Shi J. Xiao Y. Chen Z. Xu X. (2010 ). Roles of arabidopsis WRKY18, WRKY40 and WRKY60 transcription factors in plant responses to abscisic acid and abiotic stress . BMC Plant Biol. 10 :281 . 10.1186/1471-2229-10-281 21167067 Chen J. Hou K. Qin P. Liu H. Yi B. Yang W. . (2014 ). RNA-Seq for gene identification and transcript profiling of three Stevia rebaudiana genotypes . BMC Genomics 15 :571 . 10.1186/1471-2164-15-571 25001368 Chinnusamy V. Zhu J. Zhu J. K. (2007 ). Cold stress regulation of gene expression in plants . Trends Plant Sci. 12 , 444 –451 . 10.1016/j.tplants.2007.07.002 17855156 Chiwocha S. D. S. Abrams S. R. Ambrose S. J. Cutler A. J. Mary L. Ross A. R. S. . (2003 ). A method for profiling classes of plant hormones and their metabolites using liquid chromatography-electrospray ionization tandem mass spectrometry: an analysis of hormone regulation of thermodormancy of lettuce (Lactuca sativa L.) seeds . Plant J. 35 , 405 –417 . 10.1046/j.1365-313X.2003.01800.x 12887591 Colebrook E. H. Thomas S. G. Phillips A. L. Hedden P. (2014 ). The role of gibberellin signalling in plant responses to abiotic stress . J. Exp. Biol. 217(Pt 1) , 67 –75 . 10.1242/jeb.089938 24353205 Cui J. X. Zhou Y. H. Ding J. G. Xia X. J. Shi K. Chen S. C. . (2011 ). Role of nitric oxide in hydrogen peroxide-dependent induction of abiotic stress tolerance by brassinosteroids in cucumber . Plant Cell Environ. 34 , 347 –358 . 10.1111/j.1365-3040.2010.02248.x 21054437 Dhaubhadel S. Browning K. S. Gallie D. R. Krishna P. (2002 ). Brassinosteroid functions to protect the translational machinery and heat-shock protein synthesis following thermal stress . Plant J. 29 , 681 –691 . 10.1046/j.1365-313X.2002.01257.x 12148527 Dhaubhadel S. Krishna P. (2008 ). Identification of differentially expressed genes in brassinosteroid-treated Brassica napus Seedlings . J. Plant Growth Regul. 27 , 297 –308 . 10.1007/s00344-008-9056-5 Die J. V. Rowland L. J. (2014 ). Elucidating cold acclimation pathway in blueberry by transcriptome profiling . Environ. Exp. Bot. 106 , 87 –98 . 10.1016/j.envexpbot.2013.12.017 Divi U. K. Krishna P. (2010 ). Overexpression of the brassinosteroid biosynthetic gene AtDWF4 in arabidopsis seeds overcomes abscisic acid-induced inhibition of germination and increases cold tolerance in transgenic seedlings . J. Plant Growth Regul. 29 , 385 –393 . 10.1007/s00344-010-9150-3 Divi U. K. Rahman T. Krishna P. (2010 ). Brassinosteroid-mediated stress tolerance in Arabidopsis shows interactions with abscisic acid, ethylene and salicylic acid pathways . BMC Plant Biol. 10 :151 . 10.1186/1471-2229-10-151 20642851 Doblin M. S. Pettolino F. A. Wilson S. M. Campbell R. Burton R. A. Fincher G. B. . (2009 ). A barley cellulose synthase-like CSLH gene mediates (1,3;1,4)-β-d-glucan synthesis in transgenic Arabidopsis . Proc. Natl. Acad. Sci. U.S.A. 106 , 5996 –6001 . 10.1073/pnas.0902019106 19321749 Fujita Y. Fujita M. Shinozaki K. Yamaguchi-Shinozaki K. (2011 ). ABA-mediated transcriptional regulation in response to osmotic stress in plants . J. Plant Res. 124 , 509 –525 . 10.1007/s10265-011-0412-3 21416314 Gao Q. M. Venugopal S. Navarre D. Kachroo A. (2011 ). Low oleic acid-derived repression of jasmonic acid-inducible defense responses requires the WRKY50 and WRKY51 proteins . Plant Physiol. 155 , 464 –476 . 10.1104/pp.110.166876 21030507 Hao J. Yin Y. Fei S. Z. (2013 ). Brassinosteroid signaling network: implications on yield and stress tolerance . Plant Cell Rep. 32 , 1017 –1030 . 10.1007/s00299-013-1438-x 23568410 Henriksson E. Olsson A. S. Johannesson H. Johansson H. Hanson J. Engström P. . (2005 ). Homeodomain leucine zipper class I genes in Arabidopsis. Expression patterns and phylogenetic relationships . Plant Physiol . 139 , 509 –518 . 10.1104/pp.105.063461 16055682 Hu L. Huang Z. Liu S. Fu J. (2012 ). Growth response and gene expression in antioxidant-related enzymes in two bermudagrass genotypes differing in salt tolerance . J. Am. Soc. Hortic. Sci. 137 , 134 –143 . Hu Y. Yu D. (2014 ). BRASSINOSTEROID INSENSITIVE2 interacts with ABSCISIC ACID INSENSITIVE5 to mediate the antagonism of brassinosteroids to abscisic acid during seed germination in Arabidopsis . Plant Cell 26 , 4394 –4408 . 10.1105/tpc.114.130849 25415975 Hudson D. Guevara D. Yaish M. W. Hannam C. Long N. Clarke J. D. . (2011 ). GNC and CGA1 modulate chlorophyll biosynthesis and glutamate synthase (GLU1/Fd-GOGAT) expression in Arabidopsis . PLoS ONE 6 :e26765 . 10.1371/journal.pone.0026765 22102866 Jackson D. Culianez-Macia F. Prescott A. G. Roberts K. Martin C. (1991 ). Expression patterns of myb genes from Antirrhinum flowers . Plant Cell 3 , 115 –125 . 10.1105/tpc.3.2.115 1840903 Janská A. Marsík P. Zelenková S. Ovesná J. (2010 ). Cold stress and acclimation - what is important for metabolic adjustment? Plant Biol. (Stuttg). 12 , 395 –405 . 10.1111/j.1438-8677.2009.00299.x 20522175 Jansson S. (1994 ). The light-harvesting chlorophyll a/b-binding proteins . Biochim. Biophys. Acta 1184 , 1 –19 . 10.1016/0005-2728(94)90148-1 8305447 Jansson S. (1999 ). A guide to the Lhc genes and their relatives in Arabidopsis . Trends Plant Sci. 4 , 236 –240 . 10.1016/S1360-1385(99)01419-3 10366881 Journot-Catalino N. Somssich I. E. Roby D. Kroj T. (2006 ). The transcription factors WRKY11 and WRKY17 act as negative regulators of basal resistance in Arabidopsis thaliana . Plant Cell 18 , 3289 –3302 . 10.1105/tpc.106.044149 17114354 Kagale S. Divi U. K. Krochko J. E. Keller W. A. Krishna P. (2007 ). Brassinosteroid confers tolerance in Arabidopsis thaliana and Brassica napus to a range of abiotic stresses . Planta 225 , 353 –364 . 10.1007/s00425-006-0361-6 16906434 Kang D. J. Seo Y. J. Lee J. D. Ishii R. Kim K. U. Shin D. H. (2005 ). Jasmonic acid differentially affects growth, ion uptake and abscisic acid concentration in salt-tolerant and salt-sensitive rice cultivars . J. Agron. Crop Sci. 191 , 273 –282 . 10.1111/j.1439-037X.2005.00153.x Kim S. K. Park H. Y. Jang Y. H. Lee J. H. Kim J. K. (2013 ). The sequence variation responsible for the functional difference between the CONSTANS protein, and the CONSTANS-like (COL) 1 and COL2 proteins, resides mostly in the region encoded by their first exons . Plant Sci. 199–200 , 71 –78 . 10.1016/j.plantsci.2012.09.019 Li J. Yang P. Gan Y. Yu J. Xie J. (2015a ). Brassinosteroid alleviates chilling-induced oxidative stress in pepper by enhancing antioxidation systems and maintenance of photosystem II . Acta Physiol. Plant. 37 , 1 –11 . 10.1007/s11738-015-1966-9 Li J. Yang P. Xie J. Yu J. (2015b ). Effects of 24-epibrassinolide on growth and antioxidant enzymes system in pepper roots under chilling stress . J. Nucl. Agric. Sci. 29 , 1001 –1008 . 10.11869/j.issn.100-8551.2015.05.1001 Li Y. Baldauf S. Lim E. K. Bowles D. J. (2001 ). Phylogenetic analysis of the UDP-glycosyltransferase multigene family of Arabidopsis thaliana . J. Biol. Chem. 276 , 4338 –4343 . 10.1074/jbc.M007447200 11042215 Liscum E. Reed J. W. (2002 ). Genetics of Aux/IAA and ARF action in plant growth and development . Plant Mol. Biol. 49 , 387 –400 . 10.1023/A:1015255030047 12036262 Liu H. Zhang H. Yang Y. Li G. Yang Y. Wang X. . (2008 ). Functional analysis reveals pleiotropic effects of rice RING-H2 finger protein gene OsBIRF1 on regulation of growth and defense responses against abiotic and biotic stresses . Plant Mol. Biol. 68 , 17 –30 . 10.1007/s11103-008-9349-x 18496756 Liu J. Gao H. Wang X. Zheng Q. Wang C. Wang X. . (2014 ). Effects of 24-epibrassinolide on plant growth, osmotic regulation and ion homeostasis of salt-stressed canola . Plant Biol. (Stuttg). 16 , 440 –450 . 10.1111/plb.12052 24033882 Liu T. Zhu S. Tang Q. Tang S. (2015 ). Genome-wide transcriptomic profiling of ramie (Boehmeria nivea L. Gaud) in response to cadmium stress . Gene 558 , 131 –137 . 10.1016/j.gene.2014.12.057 25550046 Livak K. J. Schmittgen T. D. (2001 ). Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method . Methods 25 , 402 –408 . 10.1006/meth.2001.1262 11846609 Lu X. Kim H. Zhong S. Chen H. Hu Z. Zhou B. (2014 ). De novo transcriptome assembly for rudimentary leaves in Litchi chinesis Sonn. and identification of differentially expressed genes in response to reactive oxygen species . BMC Genomics 15 :805 . 10.1186/1471-2164-15-805 25239404 Marino D. Dunand C. Puppo A. Pauly N. (2012 ). A burst of plant NADPH oxidases . Trends Plant Sci. 17 , 9 –15 . 10.1016/j.tplants.2011.10.001 22037416 Mazorra L. M. Nunez M. Hechavarria M. Coll F. Sanchez-Blanco M. J. (2002 ). Influence of brassinosteroids on antioxidant enzymes activity in tomato under different temperatures . Biol. Plant. 45 , 593 –596 . 10.1023/A:1022390917656 Mortazavi A. Williams B. A. McCue K. Schaeffer L. Wold B. (2008 ). Mapping and quantifying mammalian transcriptomes by RNA-Seq . Nat. Methods 5 , 621 –628 . 10.1038/nmeth.1226 18516045 Pantin F. Monnet F. Jannaud D. Costa J. M. Renaud J. Muller B. . (2013 ). The dual effect of abscisic acid on stomata . New Phytol. 197 , 65 –72 . 10.1111/nph.12013 23106390 Peng Q. Zhou Q. (2009 ). The endogenous hormones in soybean seedlings under the joint actions of rare earth element La(III) and ultraviolet-B stress . Biol. Trace Elem. Res. 132 , 270 –277 . 10.1007/s12011-009-8404-z 19462161 Prasad T. K. Cline M. G. (1987 ). The role of gravity in apical dominance: effects of clinostating on shoot inversion-induced ethylene production, shoot elongation, and lateral bud growth . Plant Physiol. 83 , 505 –509 . 10.1104/pp.83.3.505 11539031 Qin C. Yu C. Shen Y. Fang X. Chen L. Min J. . (2014 ). Whole-genome sequencing of cultivated and wild peppers provides insights into Capsicum domestication and specialization . Proc. Natl. Acad. Sci. U.S.A. 111 , 5135 –5140 . 10.1073/pnas.1400975111 24591624 Ren L. Sun J. Chen S. Gao J. Dong B. Liu Y. . (2014 ). A transcriptomic analysis of Chrysanthemum nankingense provides insights into the basis of low temperature tolerance . BMC Genomics 15 :844 . 10.1186/1471-2164-15-844 25277256 Richter R. Behringer C. Muller I. K. Schwechheimer C. (2010 ). The GATA-type transcription factors GNC and GNL/CGA1 repress gibberellin signaling downstream from DELLA proteins and PHYTOCHROME-INTERACTING FACTORS . Genes Dev. 24 , 2093 –2104 . 10.1101/gad.594910 20844019 Sanghera G. S. Wani S. H. Hussain W. Singh N. B. (2011 ). Engineering cold stress tolerance in crop plants . Curr. Genomics 12 , 30 –43 . 10.2174/138920211794520178 21886453 Shi H. Chan Z. (2014 ). The cysteine2/histidine2-type transcription factor ZINC FINGER OF ARABIDOPSIS THALIANA 6-activated C-REPEAT-BINDING FACTOR pathway is essential for melatonin-mediated freezing stress resistance in Arabidopsis . J. Pineal Res. 57 , 185 –191 . 10.1111/jpi.12155 24962049 Shi H. Ye T. Chan Z. (2014 ). Comparative proteomic responses of two bermudagrass (Cynodon dactylon (L). Pers.) varieties contrasting in drought stress resistance . Plant Physiol. Biochem. 82 , 218 –228 . 10.1016/j.plaphy.2014.06.006 24992888 Sun Y. G. Wang B. Jin S. H. Qu X. X. Li Y. J. Hou B. K. (2013 ). Ectopic expression of Arabidopsis glycosyltransferase UGT85A5 enhances salt stress tolerance in tobacco . PLoS ONE 8 :e59924 . 10.1371/journal.pone.0059924 23533660 Takahashi Y. Ebisu Y. Kinoshita T. Doi M. Okuma E. Murata Y. . (2013 ). bHLH transcription factors that facilitate K(+) uptake during stomatal opening are repressed by abscisic acid through phosphorylation . Sci. Signal. 6, ra48. 10.1126/scisignal.2003760 23779086 Vogt T. Jones P. (2000 ). Glycosyltransferases in plant natural product synthesis: characterization of a supergene family . Trends Plant Sci. 5 , 380 –386 . 10.1016/S1360-1385(00)01720-9 10973093 Wang J. Li B. Meng Y. Ma X. Lai Y. Si E. . (2015 ). Transcriptomic profiling of the salt-stress response in the halophyte Halogeton glomeratus . BMC Genomics 16 :169 . 10.1186/s12864-015-1373-z 25880042 Wang L. Feng Z. Wang X. Wang X. Zhang X. (2010 ). DEGseq: an R package for identifying differentially expressed genes from RNA-seq data . Bioinformatics 26 , 136 –138 . 10.1093/bioinformatics/btp612 19855105 Wang X. Gao J. Zhu Z. Dong X. Wang X. Ren G. . (2015 ). TCP transcription factors are critical for the coordinated regulation of isochorismate synthase 1 expression in Arabidopsis thaliana . Plant J. 82 , 151 –162 . 10.1111/tpj.12803 25702611 Welch D. Hassan H. Blilou I. Immink R. Heidstra R. Scheres B. (2007 ). Arabidopsis JACKDAW and MAGPIE zinc finger proteins delimit asymmetric cell division and stabilize tissue boundaries by restricting SHORT-ROOT action . Genes Dev. 21 , 2196 –2204 . 10.1101/gad.440307 17785527 Werner T. Motyka V. Laucou V. Smets R. Van Onckelen H. Schmülling T. (2003 ). Cytokinin-deficient transgenic arabidopsis plants show multiple developmental alterations indicating opposite functions of cytokinins in the regulation of shoot and root meristem activity . Plant Cell 15 , 2532 –2550 . 10.1105/tpc.014928 14555694 Xia X. J. Wang Y. J. Zhou Y. H. Tao Y. Mao W. H. Shi K. . (2009 ). Reactive oxygen species are involved in brassinosteroid-induced stress tolerance in cucumber . Plant Physiol. 150 , 801 –814 . 10.1104/pp.109.138230 19386805 Xie C. Mao X. Huang J. Ding Y. Wu J. Dong S. . (2011 ). KOBAS 2.0: a web server for annotation and identification of enriched pathways and diseases . Nucleic Acids Res. 39 , W316 –W322 . 10.1093/nar/gkr483 21715386 Young M. D. Wakefield M. J. Smyth G. K. Oshlack A. (2010 ). Gene ontology analysis for RNA-seq: accounting for selection bias . Genome Biol. 11 :R14 . 10.1186/gb-2010-11-2-r14 20132535 Yu X. Li L. Zola J. Aluru M. Ye H. Foudree A. . (2011 ). A brassinosteroid transcriptional network revealed by genome-wide identification of BESI target genes in Arabidopsis thaliana . Plant J. 65 , 634 –646 . 10.1111/j.1365-313X.2010.04449.x 21214652 Yuan G. F. Jia C. G. Li Z. Sun B. Zhang L. P. Liu N. (2010 ). Effect of brassinosteroids on drought resistance and abscisic acid concentration in tomato under water stress . Sci. Hortic. 126 , 103 –108 . 10.1016/j.scienta.2010.06.014 Yuan X. Deng K. Q. Zhao X. Y. Wu X. J. Qin Y. Z. Tang D. Y. . (2007 ). A calcium-dependent protein kinase is involved in plant hormone signal transduction in Arabidopsis . J. Plant Physiol. Mol. Biol. 33 , 227 –234 . 17556810 Zhang L. Li Y. Xing D. Gao C. (2009 ). Characterization of mitochondrial dynamics and subcellular localization of ROS reveal that HsfA2 alleviates oxidative damage caused by heat stress in Arabidopsis . J. Exp. Bot. 60 , 2073 –2091 . 10.1093/jxb/erp078 19342427

PMC005xxxxxx/PMC5002409.txt

==== Front Front PsycholFront PsycholFront. Psychol.Frontiers in Psychology1664-1078Frontiers Media S.A. 10.3389/fpsyg.2016.01301PsychologyOriginal ResearchThe Environment Makes a Difference: The Impact of Explicit and Implicit Attitudes as Precursors in Different Food Choice Tasks König Laura M. *Giese Helge Schupp Harald T. Renner Britta Department of Psychology, University of KonstanzKonstanz, GermanyEdited by: Jens Blechert, University of Salzburg, Austria Reviewed by: Benjamin Missbach, University of Vienna, Austria; Laura Nynke Van Der Laan, Utrecht University, Netherlands *Correspondence: Laura M. König, laura.koenig@uni-konstanz.deThis article was submitted to Eating Behavior, a section of the journal Frontiers in Psychology 29 8 2016 2016 7 130107 3 2016 15 8 2016 Copyright © 2016 König, Giese, Schupp and Renner.2016König, Giese, Schupp and RennerThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.Studies show that implicit and explicit attitudes influence food choice. However, precursors of food choice often are investigated using tasks offering a very limited number of options despite the comparably complex environment surrounding real life food choice. In the present study, we investigated how the assortment impacts the relationship between implicit and explicit attitudes and food choice (confectionery and fruit), assuming that a more complex choice architecture is more taxing on cognitive resources. Specifically, a binary and a multiple option choice task based on the same stimulus set (fake food items) were presented to ninety-seven participants. Path modeling revealed that both explicit and implicit attitudes were associated with relative food choice (confectionery vs. fruit) in both tasks. In the binary option choice task, both explicit and implicit attitudes were significant precursors of food choice, with explicit attitudes having a greater impact. Conversely, in the multiple option choice task, the additive impact of explicit and implicit attitudes was qualified by an interaction indicating that, even if explicit and implicit attitudes toward confectionery were inconsistent, more confectionery was chosen than fruit if either was positive. This compensatory ‘one is sufficient’-effect indicates that the structure of the choice environment modulates the relationship between attitudes and choice. The study highlights that environmental constraints, such as the number of choice options, are an important boundary condition that need to be included when investigating the relationship between psychological precursors and behavior. explicit attitudesimplicit attitudesfood choicechoice architecturedual process models ==== Body Introduction Coke and potato sticks is probably not a common breakfast food choice, but this eat “like a 6-year-old” diet has received a lot of public attention since Warren Buffett, an 84-year-old, top world investor, stated that this is his preferred breakfast choice1. Food choices and eating behavior show great diversity across people, and, with people making more than 200 food decisions a day, they represent a core aspect of our everyday life (Wansink and Sobal, 2007; see also Hofmann et al., 2012). Hence, actual eating behavior is likely to be driven by an interplay of explicit, or reflective, precursors (e.g., food attitudes, risk perceptions, self-efficacy, and behavioral intentions) and implicit, or automatic, reactions to stimuli in the environment which occur without conscious reflection (cf. Marteau et al., 2012). That explicit and implicit attitudes play a significant role in explaining human behavior such as food choice is commonly accepted; this is reflected in a number of dual-process theories (e.g., Epstein, 1991; Sloman, 1996; Strack and Deutsch, 2004 and see Shiv and Fedorikhin, 2002; Ackermann and Mathieu, 2015 for consumer choices) and supported by results from various meta-analyses (e.g., Greenwald et al., 2003; Hofmann et al., 2005; Rooke et al., 2008; Reich et al., 2010). However, the question of how implicit and explicit processes work in conjunction to control an overt behavioral response is still under research. In their reflective-impulsive model (RIM), Strack and Deutsch (2004) suggest that these processes operate in parallel and compete for control of an overt behavioral response. Since the impulsive system requires little cognitive capacity, it is commonly assumed that its impact on behavior increases as conditions become more resource demanding (see also MODE Model; Fazio and Olson, 2003). Along these lines, it has been repeatedly shown that restrained cognitive resources (e.g., due to cognitive load, ego depletion) facilitate the influence of implicit attitudes on food choice (e.g., Shiv and Fedorikhin, 1999, 2002; Cervellon et al., 2007; Hofmann et al., 2007; Friese et al., 2008; see also Hofmann et al., 2009b; Sheeran et al., 2013). For example, implicit attitudes were more strongly related to the food choice observed (number of chocolates chosen) when participants were asked to remember an eight-digit number (high cognitive load) instead of a one-digit number (low cognitive load) (Friese et al., 2008; see also Shiv and Fedorikhin, 1999, 2002; Cervellon et al., 2007). Similarly, implicit attitudes were more strongly related to chocolate consumption when participants were asked to suppress their emotion while watching a video clip (ego depletion condition) as compared to when no emotion regulation instruction was given (control condition) (Hofmann et al., 2007). However, past research has predominantly focused on boundary conditions related to either the person (e.g., habitualness, Conner et al., 2007) or additional behavioral demands (e.g., cognitive load, ego-depletion, Hofmann et al., 2009a), ignoring situational conditions such as the structure of the target behavior. Most studies assessing the impact of implicit and explicit attitudes on food choices realized a choice task where participants were asked to choose a specific food item. Richetin et al. (2007), for instance, offered participants a selection of fruits and snacks and asked them to choose one free snack or a piece of fruit as compensation for their participation (see also Perugini, 2005; Conner et al., 2007; Ayres et al., 2012; Holland et al., 2012). Using a similar behavioral choice set, Friese et al. (2008) asked the participants to select five items from a selection of fruits and chocolates. Thus, the structure of the target behavior was typically unvaried and its impact has yet to be assessed systematically (see also Perugini, 2005). The structure of the target behavior, that is which types of behavioral choices are possible, is dependent on the food availability or assortment size (number of food options), which represents a core part of the ‘choice architecture’ (Thaler and Sunstein, 2008; see also Marteau et al., 2011) or ‘obesogenic food environment’ (Swinburn et al., 1999). We propose that the context of the behavioral choices, in particular the assortment size (number of food options), has a significant impact on the relative weight of implicit and explicit attitudes for food choices. In the real world, the context of food choices is rather complex. Today, German supermarkets typically offer more than 25,000 choices (EHI Retail Institute, 2014). Moreover, several decisions have to be made in parallel when grocery shopping or selecting dishes for a multiple course meal as not only the items but also the quantity has to be considered (Wansink et al., 2009). Considering that the complexity of behavioral options impacts the quality of and satisfaction with behavioral choices (Lee and Lee, 2004; Scheibehenne et al., 2010), it appears plausible to assume that the context of the target behavior is an important boundary condition that can shift the weight between implicit and explicit attitudes as precursors. Building on previous work by Conner et al. (2007) investigating personal moderators of the relationship between explicit and implicit attitudes and snack choice, the present study aimed at examining the impact of implicit and explicit attitudes in a binary and a multiple option food choice context to broaden the knowledge about situational moderators. Specifically, assuming that resource demanding conditions increase the impact of implicit attitudes, we hypothesized that implicit measures of attitudes are better predictors of food choices in a multiple compared to a binary option choice task. However, realizing a standardized target stimuli set for spontaneous food choices is challenging since it often conflicts with practical issues, for example, high costs, high preparation effort and waste. To overcome these problems, we applied a method recently developed by Bucher et al. (2012) using replica food items (Figure 1) (see also Sproesser et al., 2015). The Fake Food method has been shown to be reliable and a valid assessment of real food choices (Bucher et al., 2012). Using fake food, a multiple option context, such as eating in buffet style restaurants, can be simulated under well-controlled conditions. Moreover, the same food items can be used in multiple and binary options contexts, reducing variability due to different target stimuli sets. FIGURE 1 Two course fake food buffet including 15 different food items for lunch. Materials and Methods Participants A power analysis using G∗Power 3.1 (Faul et al., 2009) to detect a medium effect (f2 = 0.15), assuming three predictors in a linear multiple regression, yielded an N of 77 for 80% power. Ninety-seven participants (75% female) were recruited by e-mail using student mailing lists from the University of Konstanz, short notices distributed around the university and postings on Facebook groups. Potential participants were informed that omnivores (i.e., non-vegans and non-vegetarians) and people without health related dietary restrictions were eligible for the study. In addition, participants were asked on-site whether they were vegetarian, vegan or had any health related dietary restrictions. As none of the participants reported these limitations, all participants were included in the analyses. Participants were aged 19 to 46 (M = 23.22, SD = 4.49) and 97.9% were students, representing a wide range of subjects. As compensation, participants received either 12€ or 1.5 h of course credit. Procedure The participants were invited to the laboratory for individual sessions. Prior to participation, all participants gave written informed consent in accordance with the Declaration of Helsinki. They were seated in front of a 19 inch computer screen and were asked to fill in a questionnaire assessing explicit attitudes. Afterward, a single category implicit attitude test (SC-IAT) was conducted followed by the binary food choice task. Participants were then asked to step in front of the fake food buffet that was covered by a table cloth. The experimenter asked the participants to serve themselves a meal that they would typically have for lunch, provided them with tableware and removed the table cloth. Participants were free to choose the contents and size of their meal (multiple food choice task). When the participants were finished, they were asked to place the dishes on the serving tray. In a final step, they filled in a second computer-based questionnaire assessing demographic and anthropomorphic variables, as well as eating motives that will not be discussed in the present publication. Finally, participants were thanked and paid. During the study, participants wore SMI eye tracking glasses to monitor their eye movements; however, eye movements and selective attention are not the focus of the current paper and are reported only in the interest of full disclosure. Materials Food Stimuli: Fake Food Items In total, 15 different food items from five food categories (vegetables, meat, side dishes, confectionery, and fruit) were used in the study (Figure 1). The same 15 stimuli were used both as realistic food replicas on a buffet and as pictures of the replicas in the questionnaire and the two computerized tasks (SC-IAT and binary option choice task). The food pictures had a size of 400 × 300 pixels and depicted one standard serving of the respective food item according to German dietary guidelines (Koelsch and Brüggemann, 2012). All food items were pre-tested to ensure that each of the replica food items in the pictures were clearly identifiable and realistic. Twenty-two participants (68% female; age: M = 24.86, SD = 3.64) rated a total of 27 pictures of food replicas from the five food categories. Participants were asked to indicate what food item they saw in the picture and indicate to what extent the food item was identifiable and realistic on a four point Likert scale, with higher values representing a more positive evaluation. For each food category, three items that had been correctly identified by all participants and rated the most identifiable and realistic were selected (Supplementary Table 1). In order to increase comparability with previous studies on food choice, we focused on confectioneries and fruit in the main analyses, specifically an assortment of chocolates, chocolate or strawberry cream cake, and apples, bananas and oranges (Figure 1; Supplementary Table 1). Measures Explicit Attitudes Explicit attitudes were assessed using a computerized questionnaire. For each food item used in the study, a picture was presented at the top of the page followed by questions concerning the depicted food item. Participants were asked to indicate whether the depicted food item was tasty vs. not tasty, good vs. bad and appealing vs. unappealing on a six point semantic differential, with higher values representing a more positive evaluation. In addition, participants evaluated the healthiness (unhealthy vs. healthy) of the foods, but this item was not included in the scale because it reduced internal consistency (confectioneries: α = 0.83 with healthiness item versus α = 0.91 without healthiness item; fruit: α = 0.85 with healthiness item versus α = 0.93 without healthiness item). This concurs with previous research, which consistently finds a difference between palatability and health attitudes (Verplanken et al., 1998; Ayres et al., 2012). The three respective items, chocolates, chocolate, or strawberry cream cake as well as apple, banana, and orange, were aggregated to yield attitude scores for confectionery and fruit, respectively. The final attitude score was taken as the difference between the attitude toward confectionery and the attitude toward fruit (i.e., positive scores indicate a preference for confectionery over fruit). Implicit Attitudes A SC-IAT was implemented in Presentation software (Version 17.02) and conducted following the procedure described by Karpinski and Steinman (2006). The stimulus material consisted of 15 fake food pictures and six positive and six negative pictures taken from the IAPS database (Lang et al., 1999)3 (see also Hofmann and Friese, 2008 for a similar procedure). The IAPS pictures were selected on the basis of the ratings provided by Lang et al. (1999). The positive and negative picture sets were comparable regarding arousal (MPositive = 5.46, SDPositive = 0.53; MNegative = 5.6, SDNegative = 0.92). They differed significantly regarding valence ratings (MPositive = 7.7, SDPositive = 0.50; MNegative = 2.43, SDNegative = 0.48; t(10) = 18.61, p < 0.001), but the valence ratings were equally extreme, as shown by the absolute z-standardized values (MPositive = 1.5, SDPositive = 0.28; MNegative = 1.47, SDNegative = 0.27). The SC-IAT consisted of eleven blocks in total. For each block, participants were instructed to categorize the images as accurately as possible by pressing one of two keys. The category labels were presented as reminders at the top of the screen during the task. Whenever participants made an error in categorizing, they were presented with a 50 ms black screen before they were informed about the error by a red cross and asked to correct the error by pressing the other key. Reaction times were recorded. If no reaction occurred within 10 s after trial onset, the trial was aborted and the next trial started. Between trials, there was an intertrial interval of 300 ms. Raw data was scored according to the recommendations of Greenwald et al. (2003), including the calculation of D-scores with a positive score representing a stronger association with the positive category. The first block consisted of 24 trials and served as a training phase where the participants were familiarized with the positive and negative IAPS pictures to ensure that they were able to correctly categorize images as ‘positive’ or ‘negative.’ Afterward, the food pictures and categories were added in the following 10 blocks, with two blocks per food category. The present publication focuses on the four blocks for the categories confectionery and fruit. The two target categories were coupled with the category ‘negative’ in the first block and ‘positive’ in the second block. Each block consisted of 42 trials with each picture appearing at least twice. Similar to the explicit attitudes measure, a final score for implicit attitudes was obtained by subtracting the D-score for fruit (α = 0.67) from the D-score for confectionery (α = 0.66), with a positive value reflecting a preference for confectionery over fruit. Binary Option Choice Task In order to assess food choices in a binary option context, participants were presented with two fake food pictures per trial using Presentation software (Version 17.02). The two presented food items were centered vertically on a gray background on the computer screen with a 6 cm distance between the inner edges of the two images. The presentation procedure followed Graham et al. (2011). First, a 1.000 to 1.600 ms jittered interval with a black fixation cross was displayed in the center of the screen, followed by a 100 ms blank screen. Then, the two food pictures were presented simultaneously. Participants watched freely for three seconds before being asked to choose which of the two food items they would prefer to eat by pressing a key (Figure 2). FIGURE 2 Flow chart of the binary option choice task. In total, all 15 fake food items were compared with one another over 105 trials, leading to a choice score per item ranging from 0, where all other food items were preferred to this item, to 14, where no other food item was preferred to this item. To obtain sum scores for the categories, all comparisons between an item and items from each of the other categories were summed up, thus reducing the possible maximum number of choices per item to 12. The scores per item then were summed up to yield a score per category, leading to choice scores per category from 0 to 36. A food choice difference score was created by subtracting the score for fruit from the score for confectionery. Positive scores indicate that confectionery was chosen more often than fruit. Multiple Option Choice Task To create a realistic, multiple option context to investigate food choice, a fake food buffet was composed according to Bucher et al. (2013). In total, the lunch buffet included 15 different foods which were placed in serving bowls and arranged on a table to resemble an actual buffet (Figure 1). Participants were provided with a serving tray (55 cm × 35 cm) containing a large (27 cm diameter) and a small plate (21 cm diameter) as well as a small bowl (12 cm diameter) and were asked to serve themselves a meal which they would usually eat for lunch. Pictures showing examples for the meals self-served can be found in Supplementary Data Sheet 1. After the participants left the laboratory, the components of the meals were weighted (continuous items, e.g., pasta) or counted (e.g., chocolate candies) by the experimenter. Before analyzing the data, the weight of each food replica was converted into the respective weight of the real food by multiplying the amount of each replica with a predetermined factor based on a comparison of the replica item and the respective real item. The energy estimates and nutrient content of each self-served food item was calculated using the German food database Bundeslebensmittelschlüssel (Hartmann et al., 2005). To yield the total amount of self-served confectionery in grams, the values for the food replica items chocolate cake, chocolates, and strawberry cream cake were summed up. Similarly, the amount of fruit was computed by summing up the amount of self-served apples, bananas, and oranges. To obtain a food choice score, the converted amount of self-served fruit was subtracted from the converted amount of self-served confectionery (see also Conner et al., 2007 for a similar procedure). Thus, a positive choice score indicates that more confectionery than fruit was chosen in grams. Body Mass Index Body Mass Index (BMI) was calculated based on self-reported weight and height. Hunger Participants were asked to indicate their hunger on a six point Likert scale ranging from (1) very hungry to (6) very full. Statistical Analysis To analyze the relationships between explicit and implicit attitudes and food choices in the binary and multiple option conditions, a path model analysis was conducted in Mplus 7.1 (Muthén and Muthén, 2012). For explicit attitudes, there were missing values for two participants (2.1%). To estimate missing values, full information maximum likelihood (FIML) estimates were applied (Muthén and Muthén, 2012) that estimate missing values based on the observed model variables. This method has been shown to be robust and more efficient than other missing value estimation methods (Enders and Bandalos, 2001). Results Descriptive statistics as well as zero-order correlations between study variables are depicted in Table 1. Table 1 Descriptive statistics and zero-order correlations between study variables. M SD 2 3 4 5 6 7 8 1 Explicit attitudes -0.67 1.15 0.04 0.63∗∗ 0.31∗∗ -0.21∗ -0.06 0.13 0.03 2 Implicit attitudes -0.20 0.42 0.22∗ 0.29∗∗ -0.10 -0.21∗ 0.07 -0.03 3 Binary option context -1.55 4.28 0.50∗∗ -0.03 -0.11 0.09 -0.25∗ 4 Multiple option context 107.00 216.83 -0.05 -0.19 0.08 -0.22∗ 5 Age 23.22 4.49 0.27∗ -0.03 -0.22∗ 6 BMI 22.21 2.87 0.03 0.15 7 Gender1 (25 %) -0.23∗ 8 Hunger 3.58 1.15 Multiple option context: weight of food replicas converted to weight of real food (in grams), reported as a difference score for confectionery versus fruit. Gender: 0 = female, 1 = male. 1For Gender, the proportion of males in the sample is reported instead of the mean. ∗p < 0.05, ∗∗p < 0.01.On average, participants held a more positive explicit attitude toward fruit than confectionery (M = -0.67, SD = 1.15). The SC-IAT also indicated a more positive implicit attitude toward fruit (M = -0.20, SD = 0.42). Explicit and implicit attitudes were not significantly correlated [r(95) = 0.04, p = 0.720]. In the binary option context, participants on average chose fruit more often than confectionery (M = -1.55, SD = 4.28). While confectionery was chosen in 7.41 (SD = 3.33) out of 36 comparisons, fruit was chosen in 8.96 (SD = 2.38) out of 36 comparisons. In the multiple option context, participants served themselves on average 153.86 (SD = 93.93) grams of fruit and 53.32 grams (SD = 56.05) of confectionery. On average, they selected 107.0 grams more confectionery than fruit (SD = 216.83). Of the total meal, confectionery accounted for 8.43% (SD = 8.51) of the total real weight (grams) and 21.98% (SD = 19.99) of the energy (kcal), while fruit accounted for 25.20% (SD = 14.32) of the total real weight and 13.81% (SD = 10.60) of the energy. Together, they accounted for 33.64% (SD = 14.7) of the total real weight and 35.79% (SD = 18.94) of the energy of the meals self-served. Food Choice To test the relationship between implicit and explicit attitudes and food choice (confectionery versus fruit) in the two different choice tasks, a path analysis was performed (Figure 3). Hunger was included as a control variable.4 FIGURE 3 Path diagram of the associations between explicit and implicit attitudes and confectionery choice in the two choice tasks option contexts. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. In the binary option choice task, the amount of confectionery versus fruit chosen significantly increased as a function of explicit attitudes (β = 0.63, p < 0.001, 95% CI [0.46; 0.77]). Also, the positive relation between implicit attitudes and confectionery choice reached statistical significance (β = 0.20, p = 0.003, 95% CI [0.07; 0.33]). The interaction between explicit and implicit attitudes was not significant (β = -0.11, p = 0.319, 95% CI [-0.33; 0.09]). Furthermore, hunger was negatively associated with the amount of confectionery versus fruit chosen (β = -0.26, p = 0.003, 95% CI [-0.44; -0.10]). Together these variables explained 51% of the variance in food choice in the binary option choice task. Turning to the multiple option choice task shows that the amount of confectionery versus fruit chosen from the buffet increased as a function of explicit (β = 0.29, p = 0.002, 95% CI [0.11; 0.49]) as well as implicit attitudes (β = 0.29, p = 0.001, 95% CI [0.14; 0.47]). The two main effects were qualified by a significant interaction between explicit and implicit attitudes on food choice (β = -0.28, p = 0.026, 95% CI [-0.52; -0.03]). As Figure 4 depicts, a greater amount of confectionery was chosen on average when both attitudes showed a consistent preference for confectionery. However, a greater amount of confectionery was also chosen when explicit and implicit attitudes toward confectionery were inconsistent (i.e., one attitude type indicated a stronger preference for confectionery while the other indicated a stronger preference for fruit), indicating a compensatory ‘one is sufficient’-effect. Conversely, participants only chose more fruit on average when both attitude types consistently indicated a greater preference for fruit. Accordingly, simple effects revealed that food choice differed significantly between participants with low (-1 SD) and high (+1 SD) implicit attitudes (β = 0.29, p = 0.003) when explicit attitudes were 1 SD below the mean. However, food choice did not differ between participants with low or high implicit attitudes (β = 0.01, p = 0.925) when explicit attitudes were 1 SD above the mean. Moreover, hunger was negatively associated with the amount of confectionery versus fruit chosen (β = -0.22, p = 0.022, 95% CI [-0.41; -0.02]). Together, the included variables explained 26% of the variance in food choice in the multiple option context5. FIGURE 4 Simple effects of attitudes on food choice in the multiple option choice task. Simple slope effects for | 1 SD| for both explicit and implicit attitudes. Food choice is reported as the food choice difference score. To further test the relative impact of implicit and explicit attitudes across the two choice tasks, we additionally compared the path coefficients. Comparing the two attitude-food choice path coefficients within the respective choice task shows that the impact of implicit attitudes was significantly weaker than the impact of explicit attitudes in the binary option choice task [χ2 (1, N = 97) = 15.74, p < 0.001], while the impact of implicit attitudes was comparable to the impact of explicit attitudes on food choice in the multiple option choice task [χ2 (1, N = 97) = 0.00, p = 0.996]. Explicit attitudes were more strongly associated with confectionery choices in the binary option choice task [χ2 (1, N = 97) = 11.70, p < 0.001]. Regarding implicit attitudes, the relationships in both choice tasks did not differ significantly [χ2 (1, N = 97) = 0.94, p = 0.332]. Discussion In the present study, the relationships between explicit and implicit attitudes and food choice (confectionery vs. fruit) were investigated in two different choice tasks. In the binary option choice task, both types of attitudes were significant precursors of food choice. However, explicit attitudes clearly had a greater impact. Conversely, in the multiple option choice task, the additive impact of explicit and implicit attitudes was qualified by an interaction indicating that more confectionery was chosen even if explicit and implicit attitudes toward confectionaries were inconsistent and one of them was positive. Therefore, the results support the notion that the context of the target behavior is a significant boundary condition which shifts the balance between implicit and explicit attitudes as precursors within choice environments. Shift between Implicit and Explicit Attitudes as Precursors Inherent in dual process models (e.g., Fazio, 1990; Strack and Deutsch, 2004) is the assumption that the degree of taxes on cognitive resources shifts the weight between a reflective and an impulsive system, which may lead to the activation of different behavioral schemata (e.g., Strack and Deutsch, 2004; Hofmann et al., 2008). In the literature (Perugini, 2005; see also Cervellon et al., 2007) three different models describing the interplay of explicit and implicit determinants are discussed: The additive model assumes that implicit and explicit attitudes independently predict the target behavior, the double dissociation model proposes that implicit attitudes predict spontaneous whereas explicit attitudes predict deliberative behavior and the multiplicative model assumes that implicit and explicit attitudes interact in predicting the target behavior. According to the dissociative pattern, explicit precursors shape behavior in some situations, while implicit precursors shape behavior in others (Perugini, 2005). As, in the present study, explicit and implicit attitudes co-varied with food choice in both choice option contexts, the results support the additive model, suggesting that both explicit and implicit attitudes independently contribute to snack choice and explain differing variance in the target behavior. This pattern of results is in line with previous studies showing that both explicit and implicit attitudes were related to sweet versus fruit choice (e.g., Conner et al., 2007). However, in the multiple option choice task, the main effects of implicit and explicit attitudes on food choice were qualified by an interaction between the attitude measures, also supporting the multiplicative pattern. As predicted, the results showed that participants served themselves a greater amount of confectionery compared to fruit when implicit and explicit attitudes indicated a preference for confectionery over fruit. Interestingly, more confectionery than fruit was also chosen when explicit and implicit attitudes toward confectionery were inconsistent and one was positive. Only when both attitudes indicated a preference for fruit over confectionery did participants serve themselves more fruit than confectionery. Hence, the overall pattern of results in the multiple option choice task seems to indicate a compensatory ‘one is sufficient’-effect. While most studies that investigated the relationship between implicit and explicit attitudes and food choice did not include the interaction in their analysis, the studies that tested the multiplicative model have yet to report significant results (Perugini, 2005; Richetin et al., 2007; Spence and Townsend, 2007; Friese et al., 2008). However, the measures these studies applied to assessing food choice are more comparable to the binary option choice task in the present study, where the interaction also did not reach significance. For instance, Perugini (2005) and Richetin et al. (2007) assessed snack choice by offering participants two alternatives to choose from, while Spence and Townsend (2007) offered genetically modified cookies and assessed whether participants accepted them. The most similar measure was used by Friese et al. (2008, Study 2), where participants were allowed to select five items out of a box that contained a selection of two chocolate bars and two kinds of fruit. As the amount if items was fixed and the number of available choice alternatives was lower (four overall compared to six deserts in addition to nine main meal foods in the present study), we argue that the choice environment provided by Friese et al. (2008) is still more limited than the multiple option choice task realized in the present study. From the literature and the present study, a clear threshold between a binary and a multiple option setting causing the two observed patterns cannot be derived, thus it is open to future studies to examine this question in greater detail. The Environment Makes a Difference To investigate shifting between implicit and explicit attitudes as precursors, one behavior is examined under different circumstances by taking dispositional or situational moderators into account. Thus, to investigate the relationship between explicit and implicit attitudes and food choice, one also needs to ask when they are associated in addition to whether they are associated in general (c.f. Perugini, 2005). Previous studies have focused on examining whether explicit and implicit attitudes are related to food choice in general, thus assessing food choice with only one measure, e.g., daily snacking recorded in a diary (Ayres et al., 2012, study 1; Conner et al., 2007, study 1) or choice of sweets versus fruits in a behavioral choice task (Richetin et al., 2007; Prestwich et al., 2011). The present study adds to this line of research by comparing two choice tasks that differ in structure and showing a shift from a strong explicit and weaker implicit precursor of food choice to two interacting precursors in a multiple and thus more complex choice environment. Similar to studies inducing cognitive load by an additional task, e.g., memory tasks (Friese et al., 2008, Study 1) or emotion suppression (Hofmann et al., 2007), one might argue that cognitive demands varied between the two choice tasks. In the binary option choice task, decisions were sequential and choices between trials unrelated while the multiple option choice task required numerous parallel decisions: whether to have or abstain from having a dessert, which combination of the six different available cakes, candies, and fruits to select and how much confectionery to choose. Although multiple comparisons were made in the binary option choice task and the task required more time to complete than the multiple option choice task, tiredness, or fatigue might be unlikely to have influenced the results in the present study as a recent analysis of respondent fatigue in choice experiments has reported inconclusive results across several data sets (Hess et al., 2012). Although fatigue induced by the choice tasks was not assessed in the present study, the observed pattern of results might support this notion: if cognitive resources had been depleted by the task, a stronger relationship between food choice and implicit attitudes would have been expected based on the literature (e.g., Friese et al., 2008, Study 1). Thus, it is likely that cognitive resources were more taxed in the multiple option choice task than the binary option choice task due to the need to make several decisions simultaneously (Scheibehenne et al., 2010). However, since the fake food buffet might have been more realistic than the binary and picture based choice task, the food items might have provided a greater challenge for self-control resources due to temptation. From the choice tasks compared in the present study, a parallel can be drawn to food choice contexts in daily life that may also differ in cognitive and self-control demands. In situations where only limited information is available or the number of available foods to choose from is limited, more cognitive resources may be available for deliberate decision making. In a recent point-of-purchase intervention study, Allan et al. (2015) showed that decreasing cognitive demands during choice by providing easily accessible healthiness information decreased the purchase of high calorie snacks. On the other hand, in situations with a large number of choice alternatives, e.g., when selecting foods from a cafeteria buffet or restaurant menu, the increased number of alternatives may increase the cognitive load, leading to a possible shift in the weight between reflective and impulsive precursors and subsequent overeating due to reduced cognitive control (see also Swinburn et al., 1999, 2011; Hardman et al., 2015). This notion is supported by observational studies investigating the link between eating out and BMI, which suggest that in less controlled environments that potentially provide more opportunities to eat and a greater selection of foods like restaurants increase obesity (Ma et al., 2003; Bezerra et al., 2012). The same relationship has been observed for supermarkets: the more supermarkets are available within a neighborhood, the higher the residents’ BMI (Wang et al., 2007), and people who consistently shop at the same supermarket have been shown to have a more similar BMI (Chaix et al., 2012). Environmental influences on eating behavior have recently gained a lot of attention. These studies suggest an impact of the eating environment on the evaluation of food items or foods chosen (Hollands et al., 2013; for reviews, see Wansink, 2004; Cohen and Babey, 2012). For example, changes in the physical environment, e.g., in the accessibility of foods (Wansink et al., 2006; Rozin et al., 2011; Kroese et al., 2015) facilitate healthier food choices. These studies, however, are silent regarding the processes underlying the different choices and thus have not investigated the moderating role of the environment on the relationship between precursor and behavior. The present study adds to this line of research by showing that changes in the choice architecture also modulate the relationship between precursor and behavior. The results highlight the need for a more systematic investigation of the relationship between a precursor and a behavior in different contexts. It can thus be suggested that health behavior interventions targeting motivation or other socio-cognitive variables might also be context-dependent. Therefore, they underline the need to tailor interventions not only to the characteristics of the target population (e.g., Schwarzer, 2008) but also to the characteristics of the context in which they are applied. Limitations The present study used two tasks to assess food choice. The two tasks differed in sensory stimulation and the number of options presented simultaneously. Whereas 2D pictures were displayed on a computer screen in the binary option choice task, the multiple option choice task presented 3D objects that had to be touched by the participants. The multiple option choice task thus increased cognitive load through the combination of a greater stimulus array. To disentangle the effects of choice assortment and sensory stimulation, fakes foods could be used in both choice tasks in future studies. However, realizing a sufficient number of trials for the binary option choice task might be challenging since an experimental set up needs to be realized where choice tasks are presented in a timely manner without participants glimpsing previous or upcoming tasks. Another limitation might be that the weight of food replica and real food diverges. However, a study which invited participants on two separate occasions showed a high agreement between self-served meals from a real food buffet and a fake food buffet (r = 0.77 to r = 0.93; Bucher et al., 2012) suggesting a high validity of the fake food buffet method. Regarding the study design, further limitations need to be acknowledged. The two choice tasks differed slightly in the instructions given. While participants were asked to choose which food they prefer in the binary option choice task, they were asked to choose what they typically eat for lunch in the multiple option choice task. Although the differences in instructions were subtler in the present study than in previous research examining task instruction differences on reaction times and eye movements during food choice (e.g., van der Laan et al., 2015), we cannot preclude the possibility that the differences influenced the relationship between explicit attitudes and food choice. Furthermore, the two choice tasks were not counterbalanced, thus choices in the binary option choice task may have influenced choice in the multiple option choice task due to consistency strivings. Furthermore, the present study was conducted in a laboratory setting, raising the question whether the found effects generalize to food choice in daily life. Two previous studies investigating the relationship between explicit and implicit attitudes and food choice in real life found associations between snack choice and explicit attitudes but not implicit attitudes (Conner et al., 2007; Ayres et al., 2012). This might indicate that, in real life, explicit attitudes are more strongly related to snacking than implicit attitudes. Alternatively, one might argue that the different pattern of results is due to the different methodological approaches used to assess the target behavior. For example, Ayres et al. (2012) used a relative consumption score (daily ratio of chocolate to fruit, with scores over 0.5 indicating that more than half the choices were for chocolate, see also Conner et al., 2007, Studies 1 and 2; Perugini, 2005, Study 2 for a similar approach). Averaged, relative consumption scores over many different types of eating occasions and environments might conceal the relationship between the variables. While situational characteristics can be controlled and manipulated in the laboratory, diary studies have yet to assess situational characteristics like choice assortment. To overcome this problem, future studies could use Ecological Momentary Assessment (Shiffman et al., 2008; Schüz et al., 2015), allowing a food diary to be combined with context measures. The sample of the present study mainly consisted of female undergraduate students without any dietary restrictions, thus it has to be acknowledge that the generalizability of the findings is limited. Although female students are a common population in previous studies on the relationship between reflective and impulsive influences and food choice (e.g., Perugini, 2005; Conner et al., 2007; Richetin et al., 2007; Friese et al., 2008; Ayres et al., 2012) and the use of the same study population enhances comparability of the findings, it would be desirable to replicate the findings with more diverse samples to enhance generalizability. Conclusion The results of the present study demonstrate that different food choice tasks might lead to different conclusions about how food choices are made. Thus, choice environments and their ‘choice architecture’ offering different kinds of choice assortments might only be comparable to a limited extent. Both explicit and implicit attitudes were related to food choice in both tasks, with explicit attitudes being more strongly related to food choice in the ‘less demanding’ binary option choice task than implicit attitudes. Importantly, in the ‘more complex,’ multiple option choice task, an interaction between implicit and explicit attitudes emerged, indicating a compensatory ‘one-is-sufficient’ effect that triggers an ‘unhealthy’ choice pattern. Generally, the results suggest that both explicit and implicit attitudes independently contribute to shaping behavior in contexts providing a limited number of choice options, while explicit and implicit attitudes interact in more complex and demanding contexts. These findings underline that the environment in which food choice takes place is an important moderator that could be included as a boundary condition in existing dual process models (c.f. Strack and Deutsch, 2004; Hofmann et al., 2008) and further explored in future (intervention) studies. Author Contributions LK, HG, BR, and HS designed the study. LK analyzed the data. LK and BR drafted the manuscript with critical comments from HG and HS. All authors read and approved the final manuscript. Conflict of Interest Statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Funding. The study was funded by the Faculty of Sciences, University of Konstanz. 1 http://fortune.com/2015/02/25/warren-buffett-diet-coke/ 2 www.neurobs.com 3 Positive: 1710, 2070, 5470, 5480, 7502, and 8531; negative: 1525, 2095, 9220, 9600, 9911, and 9925. 4 The pattern of results did not change when all four control variables (hunger, gender, age, and BMI) were included in the analysis. 5 In an additional control analysis, perceived healthiness was included. While a significant positive correlation between perceived healthiness and explicit attitudes was found (r = 0.24, p = 0.012), no significant relationship between perceived healthiness and food choice emerged in either the binary or the multiple option choice task. Hence, the pattern of results does not change when perceived healthiness is included in the model. Supplementary Material The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpsyg.2016.01301 Click here for additional data file. Click here for additional data file. ==== Refs References Ackermann C.-L. Mathieu J.-P. (2015 ). Implicit attitudes and their measurement: Theoretical foundations and use in consumer behavior research. Rech. Appl. Mark. 30 55 –77 . 10.1177/2051570715579144 Allan J. L. Johnston M. Campbell N. (2015 ). Snack purchasing is healthier when the cognitive demands of choice are reduced: a randomized controlled trial. Health Psychol. 34 750 –755 . 10.1037/hea0000173 25420063 Ayres K. Conner M. T. Prestwich A. Smith P. (2012 ). Do implicit measures of attitudes incrementally predict snacking behaviour over explicit affect-related measures? Appetite 58 835 –841 . 10.1016/j.appet.2012.01.019 22326883 Bezerra I. N. Curioni C. Sichieri R. (2012 ). Association between eating out of home and body weight. Nutr. Rev. 70 65 –79 . 10.1111/j.1753-4887.2011.00459.x 22300594 Bucher T. van der Horst K. Siegrist M. (2012 ). The fake food buffet–a new method in nutrition behaviour research. Br. J. Nutr. 107 1553 –1560 . 10.1017/S000711451100465X 21920063 Bucher T. van der Horst K. Siegrist M. (2013 ). Fruit for dessert. How people compose healthier meals. Appetite 60 74 –80 . 10.1016/j.appet.2012.10.003 23085475 Cervellon M.-C. Dube L. Knäuper B. (2007 ). Implicit and explicit influences on spontaneous and deliberate food choices. Adv. Consum. Res. 34 104 –109 . Chaix B. Bean K. Daniel M. Zenk S. N. Kestens Y. Charreire H. (2012 ). Associations of supermarket characteristics with weight status and body fat: a multilevel analysis of individuals within supermarkets (RECORD study). PLoS ONE 7 :e32908 10.1371/journal.pone.0032908 Cohen D. A. Babey S. H. (2012 ). Contextual influences on eating behaviours: heuristic processing and dietary choices. Obes. Rev. 13 766 –779 . 10.1111/j.1467-789X.2012.01001.x 22551473 Conner M. T. Perugini M. O’Gorman R. Ayres K. Prestwich A. (2007 ). Relations between implicit and explicit measures of attitudes and measures of behavior: Evidence of moderation by individual difference variables. Pers. Soc. Psychol. Bull. 33 1727 –1740 . 10.1177/0146167207309194 18000106 EHI Retail Institute (2014 ). Durchschnittliche Artikelzahl der großen Supermärkte in den Jahren 2000 2008 und 2011 [Online]. Available: http://www.handelsdaten.de/statistik/daten/studie/249902/umfrage/artikelzahl-dergrossen-supermaerkte-jahresvergleich/ [accessed September 6 2014 ]. Enders C. K. Bandalos D. L. (2001 ). The relative performance of full information maximum likelihood estimation for missing data in structural equation models. Struct. Equ. Modeling 8 430 –457 . 10.1207/S15328007SEM0803_5 Epstein S. (1991 ). “Cognitive-experiential self-theory: An integrative theory of personality,” in The Relational Self: Theoretical Convergences in Psychoanalysis and Social Psychology ed. Curtis R. C. (New York, NY : Guilfort ) 111 –137 . Faul F. Erdfelder E. Buchner A. Lang A.-G. (2009 ). Statistical power analyses using G∗ Power 3.1: tests for correlation and regression analyses. Behav. Res. Methods 41 1149 –1160 . 10.3758/BRM.41.4.1149 19897823 Fazio R. H. (1990 ). Multiple processes by which attitudes guide behavior: the MODE model as an integrative framework. Adv. Exp. Soc. Psychol. 23 75 –109 . 10.1016/S0065-2601(08)60318-4 Fazio R. H. Olson M. A. (2003 ). Implicit measures in social cognition research: their meaning and use. Annu. Rev. Psychol. 54 297 –327 . 10.1146/annurev.psych.54.101601.145225 12172003 Friese M. Hofmann W. Wänke M. (2008 ). When impulses take over: moderated predictive validity of explicit and implicit attitude measures in predicting food choice and consumption behaviour. Br. J. Soc. Psychol. 47 397 –419 . 10.1348/014466607X241540 17880753 Graham R. Hoover A. Ceballos N. A. Komogortsev O. (2011 ). Body mass index moderates gaze orienting biases and pupil diameter to high and low calorie food images. Appetite 56 577 –586 . 10.1016/j.appet.2011.01.029 21291928 Greenwald A. G. Nosek B. A. Banaji M. R. (2003 ). Understanding and using the implicit association test: I. An improved scoring algorithm. J. Pers. Soc. Psychol. 85 197 –216 . 10.1037/0022-3514.85.2.197 12916565 Hardman C. A. Ferriday D. Kyle L. Rogers P. J. Brunstrom J. M. (2015 ). So many brands and varieties to choose from: does this compromise the control of food intake in humans? PLoS ONE 10 :e0125869 10.1371/journal.pone.0125869 Hartmann B. M. Bell S. Vásquez-Caicedo A. L. Götz A. Erhardt J. Brombach C. (2005 ). Der Bundeslebensmittelschlüssel. Karlsruhe : German Nutrient Data Base . Hess S. Hensher D. A. Daly A. (2012 ). Not bored yet–revisiting respondent fatigue in stated choice experiments. Transp. Res. Part A 46 626 –644 . 10.1016/j.tra.2011.11.008 Hofmann W. Friese M. (2008 ). Impulses got the better of me: alcohol moderates the influence of implicit attitudes toward food cues on eating behavior. J. Abnorm. Psychol. 117 420 –427 . 10.1037/0021-843X.117.2.420 18489218 Hofmann W. Friese M. Roefs A. (2009a ). Three ways to resist temptation: the independent contributions of executive attention, inhibitory control, and affect regulation to the impulse control of eating behavior. J. Exp. Soc. Psychol. 45 431 –435 . 10.1016/j.jesp.2008.09.013 Hofmann W. Friese M. Strack F. (2009b ). Impulse and self-control from a dual-systems perspective. Perspect. Psychol. Sci. 4 162 –176 . 10.1111/j.1745-6924.2009.01116.x 26158943 Hofmann W. Friese M. Wiers R. W. (2008 ). Impulsive versus reflective influences on health behavior: a theoretical framework and empirical review. Health Psychol. Rev. 2 111 –137 . Hofmann W. Gawronski B. Gschwendner T. Le H. Schmitt M. (2005 ). A meta-analysis on the correlation between the implicit association test and explicit self-report measures. Pers. Soc. Psychol. Bull. 31 1369 –1385 . 10.1177/0146167205275613 16143669 Hofmann W. Rauch W. Gawronski B. (2007 ). And deplete us not into temptation: automatic attitudes, dietary restraint, and self-regulatory resources as determinants of eating behavior. J. Exp. Soc. Psychol. 43 497 –504 . 10.1016/j.jesp.2006.05.004 Hofmann W. Vohs K. D. Baumeister R. F. (2012 ). What people desire, feel conflicted about, and try to resist in everyday life. Psychol. Sci. 23 582 –588 . 10.1177/0956797612437426 22547657 Holland R. W. de Vries M. Hermsen B. van Knippenberg A. (2012 ). Mood and the attitude–behavior link the happy act on impulse, the sad think twice. Soc. Psychol. Pers. Sci. 3 356 –364 . 10.1177/1948550611421635 Hollands G. J. Shemilt I. Marteau T. M. Jebb S. A. Kelly M. P. Nakamura R. (2013 ). Altering micro-environments to change population health behaviour: towards an evidence base for choice architecture interventions. BMC Public Health 13 :1218 10.1186/1471-2458-13-1218 Karpinski A. Steinman R. B. (2006 ). The single category implicit association test as a measure of implicit social cognition. J. Pers. Soc. Psychol. 91 16 –32 . 10.1037/0022-3514.91.1.16 16834477 Koelsch C. Brüggemann I. (2012 ). Die Aid-Ernährungspyramide – Richtig essen lehren und lernen. Bonn : aid Infodienst . Kroese F. M. Marchiori D. R. de Ridder D. T. (2015 ). Nudging healthy food choices: a field experiment at the train station. J. Public Health (Oxf). 38 e133 –e137 . 10.1093/pubmed/fdv096 26186924 Lang P. J. Bradley M. M. Cuthbert B. N. (1999 ). International Affective Picture System (IAPS): Technical Manual and Affective Ratings. Gainesville, FL : The Center for Research in Psychophysiology, University of Florida . Lee B.-K. Lee W.-N. (2004 ). The effect of information overload on consumer choice quality in an on-line environment. Psychol. Mark. 21 159 –183 . 10.1002/mar.20000 Ma Y. Bertone E. R. Stanek E. J. Reed G. W. Hebert J. R. Cohen N. L. (2003 ). Association between eating patterns and obesity in a free-living US adult population. Am. J. Epidemiol. 158 85 –92 . 10.1093/aje/kwg117 12835290 Marteau T. M. Hollands G. J. Fletcher P. C. (2012 ). Changing human behavior to prevent disease: the importance of targeting automatic processes. Science 337 1492 –1495 . 10.1126/science.1226918 22997327 Marteau T. M. Ogilvie D. Roland M. Suhrcke M. Kelly M. P. (2011 ). Judging nudging: can nudging improve population health? BMJ 342 263 –265 . 10.1136/bmj.d228 Muthén L. K. Muthén B. O. (2012 ). Mplus User’s Guide 7th Edn. Los Angeles, CA : Muthén & Muthén . Perugini M. (2005 ). Predictive models of implicit and explicit attitudes. Br. J. Soc. Psychol. 44 29 –45 . 10.1348/014466604X23491 15901390 Prestwich A. Hurling R. Baker S. (2011 ). Implicit shopping: Attitudinal determinants of the purchasing of healthy and unhealthy foods. Psychol. Health 26 875 –885 . 10.1080/08870446.2010.509797 21432730 Reich R. R. Below M. C. Goldman M. S. (2010 ). Explicit and implicit measures of expectancy and related alcohol cognitions: a meta-analytic comparison. Psychol. Addict. Behav. 24 13 –25 . 10.1037/a0016556 20307108 Richetin J. Perugini M. Prestwich A. O’Gorman R. (2007 ). The IAT as a predictor of food choice: the case of fruits versus snacks. Int. J. Psychol. 42 166 –173 . 10.1080/00207590601067078 Rooke S. E. Hine D. W. Thorsteinsson E. B. (2008 ). Implicit cognition and substance use: a meta-analysis. Addict. Behav. 33 1314 –1328 . 10.1016/j.addbeh.2008.06.009 18640788 Rozin P. Scott S. Dingley M. Urbanek J. K. Jiang H. Kaltenbach M. (2011 ). Nudge to nobesity I: minor changes in accessibility decrease food intake. Judgm. Decis. Mak. 6 323 –332 . Scheibehenne B. Greifeneder R. Todd P. M. (2010 ). Can there ever be too many options? A meta-analytic review of choice overload. J. Consum. Res. 37 409 –425 . 10.1086/651235 Schüz B. Schüz N. Ferguson S. G. (2015 ). It’s the power of food: individual differences in food cue responsiveness and snacking in everyday life. Int. J. Behav. Nutr. Phys. Act. 12 :149 10.1186/s12966-015-0312-3 Schwarzer R. (2008 ). Modeling health behavior change: how to predict and modify the adoption and maintenance of health behaviors. Appl. Psychol. 57 1 –29 . 10.1111/j.1464-0597.2007.00325.x Sheeran P. Gollwitzer P. M. Bargh J. A. (2013 ). Nonconscious processes and health. Health Psychol. 32 460 –473 . 10.1037/a0029203 22888816 Shiffman S. Stone A. A. Hufford M. R. (2008 ). Ecological momentary assessment. Annu. Rev. Clin. Psychol. 4 1 –32 . 10.1146/annurev.clinpsy.3.022806.091415 18509902 Shiv B. Fedorikhin A. (1999 ). Heart and mind in conflict: the interplay of affect and cognition in consumer decision making. J. Consum. Res. 26 278 –292 . 10.1086/209563 Shiv B. Fedorikhin A. (2002 ). Spontaneous versus controlled influences of stimulus-based affect on choice behavior. Organ. Behav. Hum. Decis. Process. 87 342 –370 . 10.1006/obhd.2001.2977 Sloman S. A. (1996 ). The empirical case for two systems of reasoning. Psychol. Bull. 119 3 –22 . 10.1037/0033-2909.119.1.3 Spence A. Townsend E. (2007 ). Predicting behaviour towards genetically modified food using implicit and explicit attitudes. Br. J. Soc. Psychol. 46 437 –457 . 10.1348/014466606X152261 17565791 Sproesser G. Kohlbrenner V. Schupp H. Renner B. (2015 ). I eat healthier than you: differences in healthy and unhealthy food choices for oneself and for others. Nutrients 7 4638 –4660 . 10.3390/nu7064638 26066013 Strack F. Deutsch R. (2004 ). Reflective and impulsive determinants of social behavior. Pers. Soc. Psychol. Rev. 8 220 –247 . 10.1207/s15327957pspr0803_1 15454347 Swinburn B. A. Egger G. Raza F. (1999 ). Dissecting obesogenic environments: the development and application of a framework for identifying and prioritizing environmental interventions for obesity. Prev. Med. 29 563 –570 . 10.1006/pmed.1999.0585 10600438 Swinburn B. A. Sacks G. Hall K. D. McPherson K. Finegood D. T. Moodie M. L. (2011 ). The global obesity pandemic: shaped by global drivers and local environments. Lancet 378 804 –814 . 10.1016/S0140-6736(11)60813-1 21872749 Thaler R. H. Sunstein C. R. (2008 ). Nudge: Improving Decisions About Health, Wealth, and Happiness. New York, NY : Penguin . van der Laan L. N. Hooge I. T. De Ridder D. T. Viergever M. A. Smeets P. A. (2015 ). Do you like what you see? The role of first fixation and total fixation duration in consumer choice. Food Q. Prefer. 39 46 –55 . Verplanken B. Hofstee G. Janssen H. J. (1998 ). Accessibility of affective versus cognitive components of attitudes. Eur. J. Soc. Psychol. 28 23 –35 . 10.1002/(SICI)1099-0992(199801/02)28:1<23::AID-EJSP843>3.3.CO;2-Q Wang M. C. Kim S. Gonzalez A. A. MacLeod K. E. Winkleby M. A. (2007 ). Socioeconomic and food-related physical characteristics of the neighbourhood environment are associated with body mass index. J. Epidemiol. Community Health 61 491 –498 . 10.1136/jech.2006.051680 17496257 Wansink B. (2004 ). Environmental factors that increase the food intake and consumption volume of unknowing consumers∗. Annu. Rev. Nutr. 24 455 –479 . 10.1146/annurev.nutr.24.012003.132140 15189128 Wansink B. Just D. R. Payne C. R. (2009 ). Mindless eating and healthy heuristics for the irrational. Am. Econ. Rev. 99 165 –169 . 10.1257/aer.99.2.165 Wansink B. Painter J. E. Lee Y.-K. (2006 ). The office candy dish: proximity’s influence on estimated and actual consumption. Int. J. Obes. (Lond). 30 871 –875 . 10.1038/sj.ijo.0803217 16418755 Wansink B. Sobal J. (2007 ). Mindless eating the 200 daily food decisions we overlook. Environ. Behav. 39 106 –123 . 10.1177/0013916506295573

PMC005xxxxxx/PMC5002410.txt

==== Front Front PharmacolFront PharmacolFront. Pharmacol.Frontiers in Pharmacology1663-9812Frontiers Media S.A. 10.3389/fphar.2016.00277PharmacologyOriginal ResearchPhotosensitivity to Triflusal: Formation of a Photoadduct with Ubiquitin Demonstrated by Photophysical and Proteomic Techniques Nuin Edurne 1Pérez-Sala Dolores 2Lhiaubet-Vallet Virginie 1Andreu Inmaculada 3*Miranda Miguel A. 1*1Instituto de Tecnología Química, Universitat Politècnica de València-Consejo Superior de Investigaciones CientíficasValencia, Spain2Departamento de Biología Físico-Química, Centro de Investigaciones Biológicas, Consejo Superior de Investigaciones CientíficasMadrid, Spain3Unidad Mixta de Investigación IIS La Fe-UPV, Hospital Universitari i Politècnic La FeValencia, SpainEdited by: Emanuela Corsini, University of Milan, Italy Reviewed by: Ignacio Vaya, University of East Anglia, UK; Miolo Giorgia, University of Padova, Italy *Correspondence: Miguel A. Miranda, mmiranda@qim.upv.es Inmaculada Andreu iandreur@qim.upv.esThis article was submitted to Pharmacogenetics and Pharmacogenomics, a section of the journal Frontiers in Pharmacology 29 8 2016 2016 7 27706 5 2016 12 8 2016 Copyright © 2016 Nuin, Pérez-Sala, Lhiaubet-Vallet, Andreu and Miranda.2016Nuin, Pérez-Sala, Lhiaubet-Vallet, Andreu and MirandaThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.Triflusal is a platelet aggregation inhibitor chemically related to acetylsalicylic acid, which is used for the prevention and/or treatment of vascular thromboembolisms, which acts as a prodrug. Actually, after oral administration it is absorbed primarily in the small intestine, binds to plasma proteins (99%) and is rapidly biotransformed in the liver into its deacetylated active metabolite 2-hydroxy-4-trifluoromethylbenzoic acid (HTB). In healthy humans, the half-life of triflusal is ca. 0.5 h, whereas for HTB it is ca. 35 h. From a pharmacological point of view, it is interesting to note that HTB is itself highly active as a platelet anti-aggregant agent. Indeed, studies on the clinical profile of both drug and metabolite have shown no significant differences between them. It has been evidenced that HTB displays ability to induce photoallergy in humans. This phenomenon involves a cell-mediated immune response, which is initiated by covalent binding of a light-activated photosensitizer (or a species derived therefrom) to a protein. In this context, small proteins like ubiquitin could be appropriate models for investigating covalent binding by means of MS/MS and peptide fingerprint analysis. In previous work, it was shown that HTB forms covalent photoadducts with isolated lysine. Interestingly, ubiquitin contains seven lysine residues that could be modified by a similar reaction. With this background, the aim of the present work is to explore adduct formation between the triflusal metabolite and ubiquitin as model protein upon sunlight irradiation, combining proteomic and photophysical (fluorescence and laser flash photolysis) techniques. Photophysical and proteomic analysis demonstrates monoadduct formation as the major outcome of the reaction. Interestingly, addition can take place at any of the ε-amino groups of the lysine residues of the protein and involves replacement of the trifluoromethyl moiety with a new amide function. This process can in principle occur with other trifluoroaromatic compounds and may be responsible for the appearance of undesired photoallergic side effects. covalent binding to proteinfluorescencelaser flash photolysislysinemass spectrometrymetabolitephotoallergyGeneralitat Valenciana10.13039/501100003359Prometeo II/2013/005Ministerio de Economía y Competitividad10.13039/501100003329CTQ2015-70164_P, SAF2012-36519Instituto de Salud Carlos III10.13039/501100004587RD12/0013/0009, RD12/0013/0008, CP11/00154 ==== Body Introduction Triflusal (Figure 1) is a platelet aggregation inhibitor chemically related to acetylsalicylic acid, which is used for the prevention and/or treatment of vascular thromboembolisms (McNeely and Goa, 1998; Gonzalez-Correa and De La Cruz, 2006). Additionally, triflusal increases nitric oxide synthesis in neutrophils, leading to an increased vasodilator potential (Matías-Guiu et al., 2003). FIGURE 1 Chemical structures of triflusal and its metabolite HTB. From a clinical standpoint, triflusal shows similar efficacy to aspirin in preventing stroke, but the former is associated with a reduced risk of hemorrhagic complications. Structurally, it differs from the latter in having a trifluoromethyl moiety at position 4 (Matías-Guiu et al., 2003; Gonzalez-Correa and De La Cruz, 2006). Actually, triflusal acts as a prodrug since after oral administration, it is absorbed primarily in the small intestine, binds to plasma proteins (99%) and is rapidly biotransformed in the liver into its deacetylated active metabolite 2-hydroxy-4-trifluoromethylbenzoic acid (HTB, Figure 1) (Ramis et al., 1991; McNeely and Goa, 1998; Cho et al., 2003) In healthy humans, the half-life of triflusal is ca. 0.5 h, whereas for HTB it is ca. 35 h (Gonzalez-Correa and De La Cruz, 2006). From a pharmacological point of view, it is interesting to note that HTB is itself highly active as a platelet anti-aggregant agent. Indeed, studies on the clinical profile of both drug and metabolite have shown no significant differences between them (Ramis et al., 1991). Although their side effects are mainly gastrointestinal, it has been evidenced that both triflusal and HTB display ability to induce photoallergy in humans (Serrano et al., 1987; Lee et al., 1999, 2001; Nagore et al., 2000). This phenomenon involves a cell-mediated immune response, which is initiated by covalent binding of a light-activated photosensitizer (or a species derived therefrom) to a protein, a process known as haptenation (Ariza et al., 2011). Interestingly, although aspirin presents a similar chemical structure, it is not associated with photosensitivity disorders. Therefore, as mentioned above, the trifluoromethyl group at position 4 present in triflusal and HTB should be responsible for the photosensitizing properties of the molecule (Bosca et al., 2001; Caffieri et al., 2007; Montanaro et al., 2009). Ubiquitin is a small (8.5 KDa) regulatory protein, with 76 amino acids, present in all eukaryotic cells, which was discovered in the early 1970s (Pickart and Eddins, 2004; Herrmann et al., 2007; Hochstrasser, 2009). Covalent ubiquitination is a major regulatory post-translational process, involving attachment of the ubiquitin to lysine residue/s on a substrate protein or on another ubiquitin molecule, leading to mono or polyubiquitination (Dikic et al., 2009; Suryadinata et al., 2014). Thus, monobiquitination can modify protein activity and localization by endocytosis, cell-cycle control or lysosomal targeting (Dikic et al., 2009), whereas polyubiquitination is implicated in DNA repair and immune signaling (Dikic et al., 2009; McIntyre and Woodgate, 2015). Besides, the discovery that ubiquitin chains target proteins to the proteasome, which degrades and recycles proteins, was honored with the Nobel Prize in chemistry in 2004. Moreover, ubiquitin does not have a well-defined active site although it binds to the so-called ubiquitin binding sites, which are modular protein domains that non-covalently bind ubiquitin (Hicke et al., 2005; Hurley et al., 2006; Dikic et al., 2009) and reveal information about the functionality and the mechanism of intermolecular regulation. In this context, small proteins like ubiquitin could be appropriate models for investigating covalent binding by means of MS/MS and peptidic fingerprint analysis (Jeram et al., 2009; Hong et al., 2015; Ramirez et al., 2015). In previous work, it was shown that HTB forms covalent photoadducts with isolated lysine and polylysine (Montanaro et al., 2009). Interestingly, ubiquitin contains seven lysine residues (Lys6, Lys11, Lys27, Lys29, Lys33, Lys48, and Lys 63) that could be modified by a similar reaction. With this background, the goal of the present work is to identify possible adduct formed between the triflusal metabolite and ubiquitin as model protein, upon sunlight irradiation, combining proteomic and photophysical (fluorescence and laser flash photolysis) techniques. Materials and Methods Materials and Solvents HTB was purchased from WAKO (Osaka, Japan) and was used without further purification; n-butylamine and bovine ubiquitin, whose sequence is identical to the human protein, were provided from Sigma–Aldrich (Steinheim, Germany). Phosphate buffered saline solution (PBS, pH = 7.4, 0.01 M) was prepared by dissolving Sigma tablets in the appropriate amount of deionized water. Sephadex G-25 columns (PD-10) were acquired from Amersham Pharmacia Biotech (UK). Dichloromethane and ethyl acetate were from Scharlab (Sentmenat, Spain). Photoaddition of HTB to Ubiquitin Solutions containing HTB (5 × 10-5 M) and ubiquitin (5 × 10-5 M) in PBS were incubated 1 h in the dark. Samples were then irradiated for 1 h (18.9 J/cm2) under the sunlight. For all photophysical studies, the photoadduct was separated from the protein using guanidine chloride and disposable Sephadex G-25 columns (PD-10) (Amersham Pharmacia Biotech, UK) equilibrated with PBS and the final product was further lyophilized at -55°C for 16 h. Control included drug–protein mixture kept in the dark and ubiquitin without irradiation. Synthesis of HTB-n-Butylamine Adduct A mixture of HTB (150 mg, 0.73 mmol) and n-butylamine (720 μL, 7.3 mmol) dissolved in deaerated PBS solution (20 mL) was placed in quartz tubes and irradiated overnight by means of a multilamp photoreactor equipped with six lamps (Hitachi, F15T8/BL) with a maximal output at ca. 300 nm (Gaussian distribution). The crude product was dissolved in ethyl acetate, washed with 1 M HCl and brine, dried over MgSO4 and concentrated under vacuum. The residue was washed with dichloromethane several times to get HTB-butylNH2 as a white solid (94 mg, 63%). 1H NMR (300 MHz, CD3OD): δ 0.98 (t, J = 7.3 Hz, 3H), 1.38–1.45 (m, 2H), 1.56–1.66 (m, 2H), 3.32–3.40 (m, 2H), 7.29 (dd, J = 8.2 Hz and 1.6 Hz, 1H), 7.34 (d, J = 1.6 Hz, 1H), 7.93 (d, J = 8.2 Hz, 1H); 13C NMR (75 MHz, CD3OD): δ 14.2, 21.2, 32.5, 40.9, 116.4, 117.0, 118.5, 131.9, 142.4, 163.0, 169.1, 173.0. Exact mass: m/z found, 236.0922 calculated for C12H14NO4 (M-H)+ 236.0928. Laser Flash Photolysis Experiments Laser flash photolysis (LFP) experiments were carried out with a pulsed XeCl excimer laser (αexc = 308 nm, ca. 17 ns pulse width, <100 mJ per pulse). In general, samples received between 1 and 3 pulses for all the kinetic experiments. A pulsed Lo255 Oriel Xenon lamp was used as detecting light source. The observation wavelength was selected with a 77200 Oriel monochromator, and the signal amplified by an Oriel photomultiplier tube (PMT) system made up of a 77348 side-on tube, 70680 housing and a 70705 power supply. The signal was registered with a TDS-640A Tektronix oscilloscope and subsequently transferred to a personal computer. The absorbance of the solutions was adjusted at ca. 0.21 at the excitation wavelength. All transient spectra were recorded at room temperature using 10 mm × 10 mm quartz cells with 4 mL capacity and were bubbled for 15 min with N2 before acquisition. Fluorescence Measurements Steady-state fluorescence experiments were carried out using a Photon Technology International (PTI, Germany) LPS-220B spectrofluorometer, equipped with a monochromator in the range of 200–700 nm. Time-resolved fluorescence measurements were performed with a Time Master fluorescence lifetime spectrometer TM 2/2003 from PTI, using a hydrogen/nitrogen flash lamp as the excitation source. The kinetic traces were fitted by monoexponential decay functions, using a deconvolution procedure to separate them from the lamp pulse profile. Emission measurements were performed in the region of 330–600 nm. The absorbance of the solutions was adjusted at ca. 0.08 at 308 nm. All measurements were performed at room temperature using 10 mm × 10 mm quartz cells of 4 mL capacity, under aerobic conditions. Mass Spectrometry Analysis of HTB-Modified Ubiquitin by MALDI-TOF Solutions containing HTB (5 × 10-5 M) and ubiquitin (5 × 10-6 M) in PBS were incubated 1 h in the dark, after which they were exposed to sunlight (doses of UVA light of 18.9 J/cm2). Incubation mixtures were directly mixed with the matrix and applied to the MALDI-TOF analysis on an Autoflex III MALDI-TOF-TOF mass spectrometer (Bruker), operated in the positive mode as previously described (Renedo et al., 2007; Oeste et al., 2011). Steady-state fluorescence (αexc = 308 nm) of phosphate buffer solution of HTB (black line), sunlight irradiated HTB-ubiquitin (1:1) after sephadex filtration (green line) or non-irradiated HTB-ubiquitin (1:1) after sephadex filtration (blue line), and HTB-butylNH2 adduct (red line). (Bottom) Corresponding decays. Protein Digestion and LC-ESI-MS/MS Analysis Solution containing HTB (5 × 10-5 M) and ubiquitin (5 × 10-6 M) in PBS was incubated 1 h in the dark and irradiated by sunlight. The samples were enzymatically digested into smaller peptides using trypsin. Subsequently, these peptides were analyzed using nanoscale liquid chromatography coupled to tandem mass spectrometry (nano LC-MS/MS). Briefly, 20 μg of sample were taken (according to Qubit quantitation) and the volume was set to 20 μL. Digestion was achieved with sequencing grade trypsin (Promega, trypsin: protein ratio 1:20 w/w) V = 64 μL, overnight 37°C. Digestion was stopped with 7 μL 10% TFA (trifluoroacetic acid). The final peptide mixture was at a concentration ca 0.5 μg/μL. Next, 5 μL of sample were loaded onto a trap column (NanoLC Column, 3 μ C18 CL, 100 umx 15 cm; Nikkyo) and desalted with 0.1% TFA at 2 μL/min during 10 min. The peptides were then loaded onto an analytical column (LC Column, 3 μ C18 CL, 75 umx12 cm, Nikkyo) equilibrated in 5% acetonitrile 0.1% formic acid. Elution was carried out with a linear gradient of 5–40% B in A for 30 min (A: 0.1% formic acid; B: acetonitrile, 0.1% formic acid) at a flow rate of 300 nL/min. Peptides were analyzed in a mass spectrometer nanoESI qQTOF (5600 TripleTOF, ABSCIEX). The tripleTOF was operated in information-dependent acquisition mode, in which a 0.25-s TOF MS scan from 350 to 1250 m/z was performed, followed by 0.05-s product ion scans from 100 to 1500 m/z on the 10 most intense 2–5 charged ions. ProteinPilot v4.5. (ABSciex) search engine default parameters were used to generate peak list directly from 5600 TripleTOF wiff files. The obtained mgf was used for identification with MASCOT (v 4.0, Matrix- Science). Database search was performed on Home Made (includes sequence of interest and the contaminants described in Expasy). Searches were done with tryptic specificity allowing one missed cleavage and a tolerance on the mass measurement of 75 ppm in MS mode and 0.6 Da in MS/MS mode. Oxidation of Met and deamidation of Asn and Gln as variable modifications. HTB modification was set as variable for K, S, T. Results and Discussion Sunlight irradiation of buffered aqueous solutions of HTB (5 × 10-5 M, Figure 1) and ubiquitin (1:1 molar ratio) was performed at noon in Valencia (Spain, July). Then, the protein material was separated from the free metabolite by gel-filtration chromatography (Sephadex). The high-molecular weight fraction was examined spectroscopically to reveal the presence of a covalently linked chromophore. As a control a 1:1 solution of HTB:ubiquitin was kept in the dark and filtered by Sephadex. This way the comparison between irradiated and non-irradiated samples would inform about the formation of a covalent adduct between the metabolite and the protein. A first approach was based on UV-Vis spectrophotometry, so the absorption spectrum of the obtained proteinaceous fraction was registered together with those of ubiquitin and HTB for comparison. The protein alone in solution showed a band with a maximum at 270 nm and no significant absorption at α>300 nm (Figure 2, violet line); by contrast HTB displayed a UVB-band centered at 300 nm (Figure 2, black line). Interestingly, the irradiated and filtered HTB:ubiquitin solution clearly exhibited an absorption peaking at 310 nm (Figure 2, green line), whereas no UVB-band was observed for the non-irradiated and filtered sample (see Supplementary Material). This result points toward the formation of a covalent adduct between the protein and the metabolite. FIGURE 2 UV-Vis spectra of phosphate buffer solutions of HTB (black), ubiquitin (violet), HTB-ubiquitin adduct (green) and HTB-ButylNH2 adduct (red). To get further insight, steady-state and time-resolved fluorescence experiments were conducted. At the excitation wavelength of 308 nm, HTB emission was detected at ca. 405 nm in agreement with previous reports, (Bosca et al., 2001; Montanaro et al., 2009) whereas a red shifted fluorescence spectrum, with αem at ca. 420 nm, was obtained for the irradiated and filtered HTB-ubiquitin sample (Figure 3 top, green line). No emission was detected for the control solution kept in the dark (Figure 3 top, blue line). Moreover, time-resolved experiments revealed different lifetimes for the HTB and HTB-ubiquitin samples, τ being of ca. 9.5 and 13.5 ns, respectively (Figure 3 bottom, black and green lines). FIGURE 3 (Top) Steady-state fluorescence (αexc = 308 nm) of phosphate buffer solution of HTB (black line), sunlight irradiated HTB-ubiquitin (1:1) after sephadex filtration (green line) or non-irradiated HTB-ubiquitin (1:1) after sephadex filtration (blue line), and HTB-butylNH2 adduct (red line). (Bottom) Corresponding decays. Additional spectroscopic studies of the covalent adduct between HTB and ubiquitin were performed by nanosecond LFP using a 308 nm XeCl excimer laser for excitation. The transient spectra registered for a nitrogen-flushed solution of HTB alone or irradiated and filtered HTB-ubiquitin in phosphate buffer solutions are shown in Figure 4. The HTB-ubiquitin adduct exhibited a transient absorption centered at 520 nm, a wavelength close to that observed for the HTB triplet excited state at αTT of ca. 500 nm (Bosca et al., 2001; Montanaro et al., 2009). FIGURE 4 Normalized transient absorption spectra of HTB (black line), HTB/ubiquitin 1:1 (blue line), HTB-ubiquitin (green line) and HTB-butylNH2 (red line) adducts in phosphate buffer N2-flushed solutions, 0.5 μs after the laser pulse. Inset: normalized decay traces of HTB/ubiquitin 1:1 (blue) and HTB-ubiquitin adduct (green). Thus, detection of a spectroscopic response for the high molecular weight fraction of the irradiated HTB:ubiquitin solution evidenced the formation of a covalent photoadduct. Furthermore, the spectral similarities between the free metabolite and this proteinaceous fraction indicated the presence of a HTB-like chromophore in the modified ubiquitin biomolecule. Thus, based on the previous studies that have established photoaddition of HTB at the ε-amino group of lysine, (Montanaro et al., 2009) the model compound (HTB-butylNH2, Figure 5) was synthesized by UVB-irradiation of a phosphate buffer solution of n-butylamine and HTB. Photoadduct HTB-butylNH2 was obtained as main product in 63% yield and fully characterized by NMR and HRMS (see Supplementary Material). The spectroscopic characterization of this model compound in phosphate buffer revealed a strong analogy between its photobehavior and that of HTB-ubiquitin photoadduct. Indeed, it displayed the same UVB absorption maximum (Figure 2, red line), a coincident fluorescence emission and lifetime (Figure 3), as well as an identical transient absorption spectrum (Figure 4). FIGURE 5 Photoadduct HTB-butylNH2 formation. Altogether, these results support formation of a HTB-ubiquitin adduct through photoaddition at the ε-amino group of lysine (Scheme 1). To obtain more precise structural information, the photoreactivity of HTB with the whole ubiquitin biomolecule was addressed by MALDI-TOF and HPLC-nanoESI analysis. Comparison between the MALDI-TOF spectra of sunlight irradiated solutions of ubiquitin alone (5 × 10-5 M, m/z 8599) and HTB-ubiquitin (10:1) mixture revealed the appearance of a new peak at m/z 8763 that corresponds to an increment of 164 amu (Figure 6), compatible with L(-H). Similar results were obtained for 1:1 ratio. Next, incubation mixtures were filtered to remove excess HTB and trypsin digestion followed by HPLC-MS/MS was performed in order to investigate the modified peptide sequence and to undertake a detailed characterization of the HTB-ubiquitin adduct. Full scan, as well as fragmentation, data files were analyzed by means of the Mascot® database search engine (Matrix Science, Boston, MA, USA) and by entering variable modifications that take into account the main nucleophilic sites able to react with the trifluoromethyl group of HTB, i.e., Lys, Thr and Ser, (Caffieri et al., 2007; Montanaro et al., 2009). Results confirmed identification of six HTB-ubiquitin derived peptide adducts: 1MQIFVKTLTGK11, 7TLTGKTITLEVEPSDTIENVK27, 12TITLEVEPSDTIENVKAK29, 30IQDKEGIPPDQQR42, 43LIFAGKQLEDGR54, 55TLSDYNIQKESTLHLVLR72 (all peptides have one missed cleavage at Lys), their related data are summarized in Table 1. SCHEME 1 Proposed mechanism of HTB-ubiquitin adduct formation. FIGURE 6 MALDI-TOF spectra of HTB-ubiquitin (10:1) kept in the dark (Top) or solution after 30 min of sunlight irradiation (Bottom). Table 1 Identification of HTB-modified peptide by MS/MS spectrometry. Peptide adduct ID Observed precursor ion (m/z) Charge (z) Mr (exp) Mr (calcd) Sequence Adduct site 1 715.7316 2 1428.7316 1428.7323 1MQIFVK#TLTGK11 K6 2 818.0825 3 2451.2257 2451.2268 7TLTGK#TITLEVEPSDTIENVK27 K11 3 717.6962 3 2150.0668 2150.0630 12TITLEVEPSDTIENVK#AK29 K27 4 844.3987 2 1686.7828 1686.7849 30IQDK#EGIPPDQQR42 K33 5 755.8790 2 1509.7434 1509.7463 43LIFAGK#QLEDGR54 K48 6 765.3953 3 2293.1641 2293.1590 55TLSDYNIQK#ESTLHLVLR72 K63 Thus, the modification site of each peptide was assessed by tandem mass experiments on the trypsin digests. The MS/MS fragmentation was achieved by selecting the precursor ions given in Table 1. The peptide sequence well agreed with the y and b ion series (see Supplementary Material). Here, the case of fragment 4, namely 30IQDKEGIPPDQQR42, is discussed as an example (Figure 7); further information on the others fragments is given in the Supplementary Material. The MS/MS fragment ions showed an unmodified y ion series from y1 to y9, whereas an increment of m/z 292 L(-H)-Lys(-H2O) was detected between y9 to y10. Accordingly, the b ion series suffered the same increase from b3 to b4. Thus, the modified amino acid is the Lys 33. Examination of the other five tryptic peptides confirms that Lys is the main site for the adduct formation with modifications detected at K6, K11, K27, K48, and K63. It should be mentioned that no diagnostic y and b fragment ions were found for Lys 27 and 29; however, detection of peptide 3 points toward formation of an adduct at site 27, which represents a missed cleavage due to the bulky L substituent that likely hinders trypsin ability to cleave at this site. FIGURE 7 ESI-MS/MS spectra and fragmentation pathway of the HTB-modified peptide 4. Conclusion Irradiation of HTB, the active metabolite of triflusal, in the presence of ubiquitin under sunlight gives rise to covalent photobinding. Photophysical and proteomic analysis demonstrate that although the main product found is a monoadduct, the reaction take places at all the ε-amino groups of the lysine residues of the protein and involves replacement of the trifluoromethyl moiety with a new amide function. Concentrations of both drug and protein used in the present work are compatible with those present in blood plasma. Therefore, it is expected that immunologic reactions occur in patients under triflusal therapy in combination with sunlight exposure. This process can in principle occur with other trifluoroaromatic compounds and may be responsible for the appearance of undesired photoallergic side effects. Author Contributions All authors listed, have made substantial, direct and intellectual contribution to the work, and approved it for publication. Conflict of Interest Statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Funding. Financial support from the Generalitat Valenciana (Prometeo Program), the Spanish Government (MINECO CTQ2015-70164-P to VL-V and SAF2012-36519 to DP-S) and the Carlos III Institute of Health (Grant RIRAAF, RETICS program, RD12/0013/0009 to MM and RD12/0013/0008 to DP-S, and Miguel Servet Contract CP11/00154 for IA) is gratefully acknowledged. The proteomic analysis was performed in the proteomics facility of SCSIE University of Valencia, which belongs to ProteoRed, PRB2-ISCIII, supported by grant PT13/0001, of the PE I+performingD+i 2013–2016, funded by ISCIII and FEDERPT13/0001. We would also like to thank Dr. José Aguilera for performing the determination of UVA irradiation doses. Supplementary Material The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fphar.2016.00277 Click here for additional data file. ==== Refs References Ariza A. Montañez M. I. Pérez-Sala D. (2011 ). Proteomics in immunological reactions to drugs. Curr. Opin. Allergy Clin. Immunol. 11 305 –312 . 10.1097/ACI.0b013e3283489ae5 21659862 Bosca F. Cuquerella M. C. Marin M. L. Miranda M. A. (2001 ). Photochemistry of 2-hydroxy-4-trifluoromethylbenzoic acid, major metabolite of the photosensitizing platelet antiaggregant drug triflusal. Photochem. Photobiol. 73 463 –468 . 10.1562/0031-8655(2001)0730463POHTAM2.0.CO2 11367565 Caffieri S. Miolo G. Seraglia R. Dalzoppo D. Toma F. M. van Henegouwen G. M. (2007 ). Photoaddition of fluphenazine to nucleophiles in peptides and proteins. Possible cause of immune side effects. Chem. Res. Toxicol. 20 1470 –1476 . 10.1021/tx700123u 17883260 Cho H.-Y. Jeong T.-J. Lee Y.-B. (2003 ). Simultaneous determination of triflusal and its major active metabolite, 2-hydroxy-4-trifluoromethyl benzoic acid, in rat and human plasma by high-performance liquid chromatography. J. Chromatogr. B 798 257 –264 . 10.1016/j.jchromb.2003.09.049 Dikic I. Wakatsuki S. Walters K. J. (2009 ). Ubiquitin-binding domains - from structures to functions. Nat. Rev. Mol. Cell Biol. 10 659 –671 . 10.1038/nrm2767 19773779 Gonzalez-Correa J. A. De La Cruz J. P. (2006 ). Triflusal: an antiplatelet drug with a neuroprotective effect? Cardiovasc. Drug Rev. 24 11 –24 . 10.1111/j.1527-3466.2006.00011.x 16939630 Herrmann J. Lerman L. O. Lerman A. (2007 ). Ubiquitin and ubiquitin-like proteins in protein regulation. Circ. Res. 100 1276 –1291 . 10.1161/01.RES.0000264500.11888.f0 17495234 Hicke L. Schubert H. L. Hill C. P. (2005 ). Ubiquitin-binding domains. Nat. Rev. Mol. Cell Biol. 6 610 –621 . 10.1038/nrm1701 16064137 Hochstrasser M. (2009 ). Origin and function of ubiquitin-like proteins. Nature 458 422 –429 . 10.1038/nature07958 19325621 Hong J. H. Ng D. Srikumar T. Raught B. (2015 ). The use of ubiquitin lysine mutants to characterize E2–E3 linkage specificity: Mass spectrometry offers a cautionary “tail”. Proteomics 15 2910 –2915 . 10.1002/pmic.201500058 26036340 Hurley J. H. Lee S. Prag G. (2006 ). Ubiquitin-binding domains. Biochem. J. 399 361 –372 . 10.1042/bj20061138 17034365 Jeram S. M. Srikumar T. Pedrioli P. G. A. Raught B. (2009 ). Using mass spectrometry to identify ubiquitin and ubiquitin-like protein conjugation sites. Proteomics 9 922 –934 . 10.1002/pmic.200800666 19180541 Lee A. Y. Joo H. J. Chey W. Y. Kim Y. G. (2001 ). Photopatch testing in seven cases of photosensitive drug eruptions. Ann. Pharmacother. 35 1584 –1587 . 10.1345/aph.1A007 11793626 Lee A. Y. Yoo S. H. Lee K. H. (1999 ). A case of photoallergic drug eruption caused by triflusal (Disgren). Photodermatol. Photoimmunol. Photomed. 15 85 –86 . 10.1111/j.1600-0781.1999.tb00062.x 10321521 Matías-Guiu J. Ferro J. M. Alvarez-Sabín J. Torres F. Jiménez M. D. Lago A. (2003 ). Comparison of triflusal and aspirin for prevention of vascular events in patients after cerebral infarction: the tacip study: a randomized, double-blind, multicenter trial. Stroke 34 840 –848 . 10.1161/01.str.0000063141.24491.50 12649515 McIntyre J. Woodgate R. (2015 ). Regulation of translesion DNA synthesis: posttranslational modification of lysine residues in key proteins. DNA Repair. 29 166 –179 . 10.1016/j.dnarep.2015.02.011 25743599 McNeely W. Goa K. L. (1998 ). Triflusal. Drugs 55 823 –833 . 10.2165/00003495-199855060-00011 9617597 Montanaro S. Lhiaubet-Vallet V. Jimenez M. C. Blanca M. Miranda M. A. (2009 ). Photonucleophilic addition of the ε-amino group of lysine to a triflusal metabolite as a mechanistic key to photoallergy mediated by the parent drug. ChemMedChem 4 1196 –1202 . 10.1002/cmdc.200900066 19462393 Nagore E. Perez-Ferriols A. Sanchez-Motilla J. M. Serrano G. Aliaga A. (2000 ). Photosensitivity associated with treatment with triflusal. J. Eur. Acad. Dermatol. Venereol. 14 219 –221 . 10.1046/j.1468-3083.2000.00074.x 11032071 Oeste C. L. Díez-Dacal B. Bray F. García de Lacoba M. de la Torre B. G. Andreu D. (2011 ). The C-terminus of H-Ras as a target for the covalent binding of reactive compounds modulating Ras-dependent pathways. PLoS ONE 6 :e15866 10.1371/journal.pone.0015866 Pickart C. M. Eddins M. J. (2004 ). Ubiquitin: structures, functions, mechanisms. BBA Mol. Cell Res. 1695 55 –72 . 10.1016/j.bbamcr.2004.09.019 Ramirez J. Martinez A. Lectez B. Lee S. Y. Franco M. Barrio R. (2015 ). Proteomic analysis of the ubiquitin landscape in the Drosophila embryonic nervous system and the adult photoreceptor cells. PLoS ONE 10 :e0139083 10.1371/journal.pone.0139083 Ramis J. Mis R. Forn J. Torrent J. Gorina E. Jané F. (1991 ). Pharmacokinetics of triflusal and its main metabolite HTB in healthy subjects following a single oral dose. Eur. J. Drug Metab. Pharmacokinet. 16 269 –273 . 10.1007/bf03189971 1823870 Renedo M. Gayarre J. García-Domínguez C. A. Pérez-Rodríguez A. Prieto A. Cañada F. J. (2007 ). Modification and activation of Ras proteins by electrophilic prostanoids with different structure are site-selective. Biochemistry 46 6607 –6616 . 10.1021/bi602389p 17489560 Serrano G. Aliaga A. Planells I. (1987 ). Photosensitivity associated with triflusal (Disgren). Photodermatol. 4 103 –105 .3658819 Suryadinata R. Roesley S. N. A. Yang G. Sarcevic B. (2014 ). Mechanisms of generating polyubiquitin chains of different topology. Cells 3 674 –689 . 10.3390/cells3030674 24987835

PMC005xxxxxx/PMC5002411.txt

==== Front Front Plant SciFront Plant SciFront. Plant Sci.Frontiers in Plant Science1664-462XFrontiers Media S.A. 10.3389/fpls.2016.01296Plant ScienceReviewCalcium-Mediated Abiotic Stress Signaling in Roots Wilkins Katie A. Matthus Elsa Swarbreck Stéphanie M. Davies Julia M. *Department of Plant Sciences, University of CambridgeCambridge, UKEdited by: Janin Riedelsberger, University of Talca, Chile Reviewed by: Girdhar Kumar Pandey, University of Delhi, India; Joachim Krebs, Max Planck Institute for Biophysical Chemistry, Germany *Correspondence: Julia M. Davies, jmd32@cam.ac.ukThis article was submitted to Plant Physiology, a section of the journal Frontiers in Plant Science 29 8 2016 2016 7 129624 5 2016 12 8 2016 Copyright © 2016 Wilkins, Matthus, Swarbreck and Davies.2016Wilkins, Matthus, Swarbreck and DaviesThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.Roots are subjected to a range of abiotic stresses as they forage for water and nutrients. Cytosolic free calcium is a common second messenger in the signaling of abiotic stress. In addition, roots take up calcium both as a nutrient and to stimulate exocytosis in growth. For calcium to fulfill its multiple roles must require strict spatio-temporal regulation of its uptake and efflux across the plasma membrane, its buffering in the cytosol and its sequestration or release from internal stores. This prompts the question of how specificity of signaling output can be achieved against the background of calcium’s other uses. Threats to agriculture such as salinity, water availability and hypoxia are signaled through calcium. Nutrient deficiency is also emerging as a stress that is signaled through cytosolic free calcium, with progress in potassium, nitrate and boron deficiency signaling now being made. Heavy metals have the capacity to trigger or modulate root calcium signaling depending on their dose and their capacity to catalyze production of hydroxyl radicals. Mechanical stress and cold stress can both trigger an increase in root cytosolic free calcium, with the possibility of membrane deformation playing a part in initiating the calcium signal. This review addresses progress in identifying the calcium transporting proteins (particularly channels such as annexins and cyclic nucleotide-gated channels) that effect stress-induced calcium increases in roots and explores links to reactive oxygen species, lipid signaling, and the unfolded protein response. abiotic stresscalciumheavy metalhypoxianutritionsalinitysignalingBiotechnology and Biological Sciences Research Council10.13039/501100000268BB/K009869/1 ==== Body Introduction Plant roots are exposed to a variety of abiotic stresses as they navigate the soil, foraging for nutrients and water. Cytosolic free calcium ([Ca2+]cyt) is central to the response to these stresses, acting as a second messenger but also driving exocytosis (Carroll et al., 1998). Specificity of [Ca2+]cyt signaling is determined by the amplitude and duration (and possible oscillation) of the [Ca2+]cyt increase, often referred to as the “signature”’ (McAinsh and Pittman, 2009), that is elicited by the stimulus. This signature would be driven by the opening of plasma membrane (PM) and endomembrane Ca2+-permeable channels and terminated by the activity of Ca2+ efflux transporters in those membranes, plus Ca2+-binding proteins, to restore the resting [Ca2+]cyt of 100–200 nM. Use of organelle-targeted Ca2+ reporting proteins has shown that the Ca2+ content of the endoplasmic reticulum (ER) and Golgi increases after stress-induced transient increases in [Ca2+]cyt, strongly suggesting that Ca2+ is sequestered there to terminate the [Ca2+]cyt signal (Ordenes et al., 2012; Bonza et al., 2013). Transport of Ca2+ into organelles is catalyzed by Ca2+-ATPases. There are two distinct families: The Auto-inhibited Ca2+-ATPases, ACA (that also operate at the PM) and the ER Ca2+-ATPases, ECA; reviewed by Bonza et al., 2016). The lower affinity CAX (Cation/H+ Exchangers) appear to be restricted to endomembranes but also facilitate Ca2+ sequestration (Connorton et al., 2012). Changes in organelle free Ca2+ in roots could also play a part in signaling, most notably in the formation of symbioses and cell death (Stael et al., 2012; Zhao et al., 2013; Wagner et al., 2015). Decoding the [Ca2+]cyt signature will be effected by specific Ca2+-binding proteins. Calmodulins (CaMs) and Calmodulin-like proteins (CMLs) are encoded by multi-gene families in plants. They lack kinase domains, suggesting these proteins must target others with enzymatic activity. CaMs modulate transcription by binding to Calmodulin-binding Transcription Activators (CAMTAs) (Virdi et al., 2015). Other multi-gene families are also evident for Ca2+-Dependent Protein Kinases (CPKs) and Calcineurin-B Like proteins (CBLs). The latter target CBL-Interacting Protein Kinases (CIPKs) to effect cellular responses (Thoday-Kennedy et al., 2015). Changes in [Ca2+]cyt also have the potential to activate lipid signaling pathways. A somewhat forgotten aspect of Ca2+ signaling is the Ca2+ activation of members of the Phospholipase C and Phospholipase D families (Qin et al., 1997; Hunt et al., 2004; Dressler et al., 2014; Ruelland et al., 2015; Hou et al., 2016). Phospholipase C catalyses production of diacylglycerol and inositol trisphosphate (InsP3) while Phospholipase D catalyses production of phosphatidic acid, thus [Ca2+]cyt would have the capacity to trigger distinct lipid signals depending on the location and Ca2+-sensitivity of the phospholipases. Targets of lipid signals have been reviewed recently by Hou et al. (2016). The vast majority of [Ca2+]cyt measurements are from Arabidopsis thaliana seedlings and guard cells, achieved using the luminescent Ca2+-interacting aequorin protein. Far fewer studies have focused specifically on roots or utilized the greater sensitivity and spatial resolution of ratiometric fluorescent dyes. The genetically encoded YC3.6 Ca2+ reporter is now being used for both Arabidopsis and rice roots (Behera et al., 2015), holding much promise for the future. It is now clear that an identical stimulus can elicit markedly different root [Ca2+]cyt signatures depending on genus. So far, rice root [Ca2+]cyt signals have been found to be lower in amplitude but of longer duration than those of Arabidopsis (Behera et al., 2015). Electrophysiological studies of root cell plasma membrane (PM) have advanced our understanding of the Ca2+ influx routes that could generate [Ca2+]cyt signatures. There is a central role for PM voltage in [Ca2+]cyt signaling, as individual stresses can hyperpolarize (render it more negative) or depolarize (render it less negative). Manipulating PM voltage elicits distinct [Ca2+]cyt signatures and resultant transcriptional responses (Whalley et al., 2011; Whalley and Knight, 2013). Studies on root epidermal and root hair PM have shown that this membrane harbors channels that are activated by hyperpolarized voltage (Hyperpolarization-Activated Ca2+ Channels (HACCs); Véry and Davies, 2000; Demidchik et al., 2002, 2009; Ma et al., 2012), Depolarization-Activated Ca2+ Channels (DACCS); Demidchik et al., 2002; Miedema et al., 2008) and Voltage-Independent Ca2+ Channels (VICCs) (Demidchik et al., 2002). Thus changes in voltage would activate specific suites of channels to generate a signature. An additional tier of regulation of the PM Ca2+ influx routes is afforded by reactive oxygen species (ROS) that are produced during development and stress responses (Figure 1). This regulation depends on the specific ROS, its position, the cell type and the cell’s developmental state. In Arabidopsis roots, sensitivity of PM Ca2+ channel activation by extracellular H2O2 decreases as epidermal cells mature but is still greater than that of the cortex (Demidchik et al., 2007). A similar picture emerges for extracellular hydroxyl radicals, which elicit greater PM Ca2+ influx currents in the epidermis and root hairs than the pericycle (Demidchik et al., 2003; Foreman et al., 2003). In epidermal PM of the elongation zone, extracellular hydroxyl radicals elicit different Ca2+ channel activity to extracellular H2O2 (Demidchik et al., 2003, 2007). Thus, ROS will play a significant part of generating cell-specific [Ca2+]cyt signatures in response to stress. FIGURE 1 Reactive oxygen species (ROS) regulation of PM Ca2+ channels in the Arabidopsis root elongation zone epidermis. NADPH oxidases (RBOH) generate extracellular superoxide anion that can undergo conversion to H2O2 and hydroxyl radicals (OH⋅) (Richards et al., 2015). H2O2 could activate HACC (orange) at the extracellular PM face or enter the cytosol through aquaporins (blue) to activate at the cytosolic face (directly or indirectly) (Demidchik et al., 2007). Extracellular hydroxyl radicals activate Annexin 1 (pink) (Demidchik et al., 2003; Foreman et al., 2003; Laohavisit et al., 2012). Ca2+ influx would depolarise the PM and if unopposed this could activate DACCs (yellow) (Demidchik et al., 2002). Increased [Ca2+]cyt could further activate RBOH. Stress-induced [Ca2+]cyt elevation in roots remains poorly understood in terms of the genes encoding the PM or endo membrane Ca2+ channels involved. Plants have multi-gene families of Glutamate Receptor-Like channels (GLR; activated by a range of extracellular nitrogenous ligands) and Cyclic Nucleotide-Gated channels (CNGC; activated by intracellular cyclic nucleotides), with each gene encoding a potential subunit of a potentially tetrameric channel. Some members have been characterized as having Ca2+ channel forming ability (reviewed by Swarbreck et al., 2013 and Weiland et al., 2016). Membrane residency has yet to be determined for all proteins and while the majority tested are in the PM, in Arabidopsis GLR3.5 has been localized to both mitochondria and chloroplast, depending on its splicing variant (Teardo et al., 2015), CNGC19 to the vacuole (Yuen and Christopher, 2013) and CNGC20 potentially to both PM and vacuole (Fischer et al., 2013; Yuen and Christopher, 2013). Mechanosensitive Ca2+ channels of the PM have also been identified (Nakagawa et al., 2007; Hou et al., 2014; Yuan et al., 2014; Kamano et al., 2015), as has a vacuolar Ca2+ efflux channel TPC1 (Two Pore Channel1; Peiter et al., 2005). Root cells will express specific complements of these genes and their transcription can change under abiotic stress (Dinneny et al., 2008; Roy et al., 2008), with the implication that stress resets the [Ca2+]cyt signaling system. The threat of abiotic stress is global. Drought threatens plant productivity across continents, with water shortage not only imposing an osmotic challenge but also leading to soil hardness that roots must overcome. Changing weather patterns are bringing greater rainfalls to some areas (particularly Northern Europe) thus leading to the threat of hypoxic challenge from waterlogged soil (Shabala et al., 2014). Salinity stress arising from sodic soils is made worse by irrigation and counteracting nutritional deprivation by fertilizer application comes with an increasing economic and environmental cost. In this review, the effects of salinity, water availability (including soil hardness), nutritional deprivation, heavy metals and cold on root [Ca2+]cyt will be addressed. The candidate channels for elevating [Ca2+]cyt in roots will be introduced and the downstream consequences of the signal will be reviewed. Salinity Stress from Channel to Transcription The transporters for Na+ influx into the root are not fully known but include the PM cyclic nucleotide-gated channels CNGC3 (Gobert et al., 2006) and CNGC10 (Guo et al., 2008; Jin et al., 2015) in Arabidopsis. Na+ ingress is opposed by the Annexin1 protein and the AGB1 heterotrimeric G protein subunit in Arabidopsis roots (Laohavisit et al., 2013; Yu and Assmann, 2015). The mechanisms for sensing the increase in cytosolic Na+ remain obscure (Maathuis, 2014; Shabala et al., 2015) however what is clear is that Na+ entry depolarizes the root epidermal PM voltage (Maathuis, 2014). This is significant in that it implicates depolarization-activated and voltage-independent PM Ca2+-permeable channels in generating the transient [Ca2+]cyt increases observed in roots in response to NaCl (Figure 2). Both channels types are present in Arabidopsis root epidermal PM (Demidchik et al., 2002). However, involvement of hyperpolarization-activated PM Ca2+ channels should not be dismissed because increasing [Ca2+]cyt shifts their activation voltage to more depolarized values (Véry and Davies, 2000; Demidchik et al., 2002) and they are implicated in the Arabidopsis root response to NaCl (Ma et al., 2012; Laohavisit et al., 2013). Critically, hyperosmotic stress hyperpolarizes root epidermal PM voltage (Maathuis, 2014) and this would potentially be a basis for generating a component of the [Ca2+]cyt signal specific to the hyperosmotic component of NaCl stress. FIGURE 2 Ca2+ transporters in salinity-stress signaling. Na+ enters root epidermis and depolarises the PM, possibly activating DACCs (yellow) with the [Ca2+]cyt increase also possibly activating HACCs (orange). Na+ sensing results in cGMP/cAMP production that could activate CNGC HACCs (Shabala et al., 2015). InsP3 and cADPR production causes Ca2+ release from stores by unknown channels (DeWald et al., 2001; Tracy et al., 2008; Zhang Y. et al., 2015; Zhang X. et al., 2015). ARP2/3 restricts mitochondrial Ca2+ efflux (Zhao et al., 2013). [Ca2+]cyt activation of RBOH NADPH oxidase activity would promote extracellular hydroxyl radical formation and activation of Annexin1-mediated Ca2+ influx (pink) (Laohavisit et al., 2013). We hypothesize that these are held in a microdomain (red). This leads to SOS1 transcription (Laohavisit et al., 2013). CaM activation would restrict CNGC-mediated Ca2+ influx and promote efflux by ACA Ca2+-ATPase activation (Boursiac and Harper, 2007). [Ca2+]cyt elevation activates CBL4 and CIPK24 (purple) leading to SOS1-mediated Na+ extrusion (Liu et al., 2000). CIPK24 activates vacuolar V-ATPase to provide the driving force for NHX1 and CAX1 to sequester Na+ and Ca2+ respectively, also activated by CIPK24 (Cheng et al., 2004; Batelli et al., 2007; Quintero et al., 2011). Ca2+ Influx across the PM Application of NaCl can cause a heterogeneous increase in Arabidopsis and rice root [Ca2+]cyt that depends on cell type, external [Ca2+], and the bathing medium’s effect on PM voltage (Kiegle et al., 2000; Tracy et al., 2008; Laohavisit et al., 2013; Choi et al., 2014; Zhang Y. et al., 2015). Pericycle cells of Arabidopsis have a lower amplitude of [Ca2+]cyt increase and a more variable recovery phase than the surrounding tissues (Kiegle et al., 2000), suggesting a cell-specific transporter complement. Block of NaCl-induced [Ca2+]cyt increase by lanthanides implicates PM Ca2+ influx channels in the response of both Arabidopsis and rice roots (Tracy et al., 2008; Zhang Y. et al., 2015). The anti-apoptotic protein Bcl-2 mimics lanthanides in that its overexpression impairs the NaCl-induced [Ca2+]cyt increase in rice roots (Kim et al., 2014). Clearly, the [Ca2+]cyt increase could lead to cell death if it were great enough and Bcl-2 may interact directly with PM channels. The genetic identities of the PM Ca2+ influx channels initiating the NaCl-induced [Ca2+]cyt increase remain unknown. It has been suggested that they may be Ca2+-permeable CNGCs (Shabala et al., 2015) as salinity stress increased cytosolic cGMP in Arabidopsis seedlings within seconds (Donaldson et al., 2004). The source of cGMP may prove to be critical as only the [Ca2+]cyt response to low NaCl (50 mM) was sensitive to inhibition of soluble guanylyl cyclase activity and the signal evoked by an equivalent osmotic stress was insensitive (Donaldson et al., 2004). It remains feasible that neither the ionic nor osmotic components of the [Ca2+]cyt increase in response to high [NaCl] require cGMP. The osmotic component of NaCl stress could increase [Ca2+]cyt in Arabidopsis roots through the PM mechanosensitive Ca2+-permeable channel OSCA1 (Reduced Hyperosmolality-Induced [Ca2+]I increase1). This channel mediates the [Ca2+]cyt response to hyperosmotic stress (Yuan et al., 2014). Its discovery through a screen of aequorin-expressing mutants (Yuan et al., 2014) highlights the potential of this approach for identification of channels (and their families) involved in stress-induced [Ca2+]cyt elevation. The initial increase in [Ca2+]cyt could be amplified by the production of ROS sourced ultimately by PM NADPH oxidases (encoded by Respiratory Burst Oxidase Homolog genes), with [Ca2+]cyt activating these enzymes through their EF hands (Figure 2). In accordance with this, Arabidopsis root cortical cells lacking RBOHD and F have much lower PM hyperpolarization-activated Ca2+ activity in response to NaCl challenge than wild type (Ma et al., 2012). Mutant seedlings have an impaired [Ca2+]cyt response. Arabidopsis RBOHD can be activated by CIPK26/CBL1/9 (Drerup et al., 2013) but a direct link to stress has not been shown. Extracellular hydroxyl radicals are likely to be the ROS involved in [Ca2+]cyt elevation (Chung et al., 2008; Demidchik et al., 2010; Laohavisit et al., 2013; Richards et al., 2015). The extracellular superoxide anions produced by NADPH oxidases are readily converted to H2O2 and wall Fe/Cu act as Fenton catalysts to generate hydroxyl radicals (Richards et al., 2015). Their production by Arabidopsis roots is significantly increased under NaCl stress (Demidchik et al., 2010) and RBOHC is implicated as a driver for production (Chung et al., 2008). Extracellular hydroxyl radicals activate a PM Ca2+ influx in Arabidopsis root epidermis (Demidchik et al., 2003; Foreman et al., 2003) that has now been shown to be mediated by Annexin1 (Laohavisit et al., 2012). Annexins are Ca2+-binding proteins that can bind to or insert into membranes and are implicated in stress reactions (Laohavisit and Davies, 2011; Davies, 2014). Critically, when root epidermal protoplasts are challenged with NaCl, the resultant PM hyperpolarization-activated Ca2+ influx is lost in the annexin1 loss of function mutant and the [Ca2+]cyt signal is impaired (Laohavisit et al., 2013). Hydroxyl radicals are potent but short-lived so their effects must be close to the site of production (Richards et al., 2015). As NADPH oxidases can be held in lipid rafts it may be that hydroxyl radicals target co-resident channels. Finally, extracellular ATP levels of Arabidopsis roots increase in response to NaCl (Dark et al., 2011). Extracellular ATP activates root epidermal PM hyperpolarization-activated Ca2+ influx channels via RBOHC (Demidchik et al., 2009), suggesting the involvement of extracellular ROS in channel activation. Whether these channels involve Annexin1 remains to be determined but both the annexin1 and rbohc loss of function mutants are impaired in [Ca2+]cyt-dependent transcriptional responses under NaCl stress (Chung et al., 2008; Laohavisit et al., 2013). Calcium Release from Stores Although not demonstrated in roots, the Arabidopsis Actin-Related Protein2/3 (ARP2/3) acts to limit NaCl-induced [Ca2+]cyt increase, partly by limiting Ca2+ release from mitochondria. In the arp2/3 mutant, the [Ca2+]cyt increase is greater than wild type and so is the extent of mitochondrial-driven cell death (Zhao et al., 2013). Release of vacuolar Ca2+ to the cytosol in Arabidopsis roots may be by a Na+/Ca2+ exchanger encoded by AtNCL (Na+/Ca2+ Exchanger-Like; Wang et al., 2012; Li P.H. et al., 2016). This tonoplast protein is thought to sequester Na+ into the vacuole, coupled to the release of vacuolar Ca2+. The concomitant increase in [Ca2+]cyt could provide a negative feedback mechanism to limit further transport as Ca2+ binding to the exchanger’s EF hands has been shown to be inhibitory in vitro (Li P.H. et al., 2016). Pharmacological approaches have also implicated internal stores in the NaCl-induced [Ca2+]cyt increase in roots of both Arabidopsis and rice. Inhibitors of store release of Ca2+ by cADPR (cyclic ADP ribose) and inositol trisphosphate (InsP3) suggested involvement of the InsP3 pathway in Arabidopsis roots (Tracy et al., 2008). Moreover, salt and hyperosmotic stress in Arabidopsis roots caused an InsP3 accumulation that correlated well with [Ca2+]cyt increase (DeWald et al., 2001). The rice [Ca2+]cyt signal was also sensitive to disruption of putative InsP3-gated store release whilst impairment by thapsigargin implicated the ER as a participating store (Zhang Y. et al., 2015). However, using ER-targeted YC3.6, Bonza et al. (2013) detected an increase in Arabidopsis root ER [Ca2+] in response to salt stress. This followed the salt-induced [Ca2+]cyt increase and critically, a drop in ER [Ca2+] prior to the [Ca2+]cyt increase was never observed. Therefore, with the level of resolution available, it appears that in this system the ER does not contribute to the [Ca2+]cyt signal through store release but acts to return [Ca2+]cyt to resting levels. Studies on Populus euphratica cultured cells identified the vacuole as the site of InsP3 and cADPR action (Zhang X. et al., 2015). Certainly, Phospholipase C isoforms (as the source of InsP3) are firmly implicated in salt stress responses (reviewed by Ruelland et al., 2015). The genetic identities of the endomembrane Ca2+-permeable channels implicated by pharmacological studies remain elusive. GLRs have recently been postulated to be involved in ER Ca2+ release (Weiland et al., 2016). The TPC1 vacuolar channel of Arabidopsis would be capable of releasing Ca2+ to the cytosol and recent analyses of its crystal structure has thrown greater light on its regulation by voltage and EF hands (Guo et al., 2016; Kintzer and Stroud, 2016). At current resolution afforded by YC3.6, its loss does not appear to have a significant impact on the magnitude of the Arabidopsis root [Ca2+]cyt increase to salt, rather it slightly delays the response (Choi et al., 2014). However, TPC1 has a significant part to play in the propagation of a [Ca2+]cyt wave that travels from the root apex through the cortical and endodermal tissue to signal the NaCl challenge to the shoots and elicit a transcriptional response. ROS also relay a salinity stress signal (reliant on RBOHD) in Arabidopsis leaves (Miller et al., 2009) but whether this occurs in roots and is involved in propagating the [Ca2+]cyt wave remains to be tested. Decoding, Na+ Clearance, Transcription and the Unfolded Protein Response Salinity-induced [Ca2+]cyt elevation in roots drives a transcriptional response (Laohavisit et al., 2013; Zhang Y. et al., 2015) and post-translational modifications. The proteins sensing the NaCl-induced [Ca2+]cyt increase are now being elucidated. For example, the Arabidopsis vacuolar Two Pore K+ channel 1 (TPK1) would bind Ca2+, and open to release K+ to the cytosol to maintain a favorable Na+/K+ ratio (Latz et al., 2013). This could be further enhanced by phosphorylation by CPK3, which requires micromolar [Ca2+] for activity. CPK3 is present at both the PM and vacuole. It does not appear to contribute to a transcriptional response under salt stress but has a discrete set of protein targets to phosphorylate (Mehlmer et al., 2010). On prolonged salt stress, CPK29 expression is induced. This protein can phosphorylate TPK1 at sub-micromolar [Ca2+] and is envisaged to be part of longer-term K+ homeostasis in adapted roots (Latz et al., 2013). Also in Arabidopsis, CPK27 (present at the root PM) acts to promote Na+ efflux (Zhao et al., 2015). Intriguingly, CPK7 acts to limit water transport in Arabidopsis roots through lowering PIP1 aquaporin abundance (Li et al., 2015) but whether this is relevant to salinity or osmotic stress is not yet known. It can be readily envisaged that calmodulins will bind Ca2+ and as these are negative regulators of CNGC channels (Hua et al., 2003), would act to limit further Na+ or Ca2+ influx at the PM. Further, CaM activation of ACAs (Autoinhibited Ca2+ ATPases) would restore [Ca2+]cyt to pre-stimulus levels (Boursiac and Harper, 2007). Expression of ACAs varies with salt stress and, as shown by rice roots, can relate to salt tolerance (Yamada et al., 2014). The Salt Overly Sensitive (SOS) pathway lies downstream of the root [Ca2+]cyt increase. Delineated in Arabidopsis and now acknowledged as operating in crops and trees (Thoday-Kennedy et al., 2015), the SOS pathway leads to Na+ efflux from the cytosol. Efflux across the PM is mediated by the SOS1 Na/H+ antiporter. Salt stress induction of SOS1 transcription lies downstream of Annexin1 in Arabidopsis roots and as SOS1 is required for adaptive adventitious root formation, the annexin1 loss of function mutant accordingly produces fewer of these than wild type (Laohavisit et al., 2013). Additionally, the stability of salt stress-induced SOS1 transcript requires RBOHC (Chung et al., 2008), further suggesting that this NADPH oxidase and Annexin1 may be in the same pathway. In Arabidopsis roots, increased [Ca2+]cyt is sensed at the PM by CBL4 (SOS3) which can then react with the serine/threonine protein kinase CIPK24 (SOS2) (Liu et al., 2000). The resultant CBL/CIPK complex phosphorylates the PM Na/H+ antiporter SOS1 at its auto-inhibitory C-terminus to achieve activation of Na+ efflux (Quintero et al., 2011). Activation of SOS1 can also be achieved in Arabidopsis by Mitogen Activated Protein Kinase6 (MPK6) and loss of MPK6 function impairs root growth under salt stress (Yu et al., 2010). MPK6 is activated by a phosphatidic acid produced under salt stress by Phospholipase Dα1 (PLDα1) (Yu et al., 2010). As this PLD isoform contains a Ca2+-binding C2 domain for its activation it is feasible that salt-induced [Ca2+]cyt increase could activate SOS1 through this lipid-mediated pathway via MPK6 (Yu et al., 2015). This may help explain the importance of activation of PLDs in salt-stressed crop roots such as barley (Meringer et al., 2016). SOS2 may also promote activity of the vacuolar V-type H+-ATPase to provide the driving force for Na+ sequestration (Batelli et al., 2007). The NHX1 Na+-H+ exchanger may also be regulated by SOS2 to aid vacuolar Na+ sequestration (Qiu et al., 2004) and SOS2 activation of the vacuolar CAX1 Ca2+/H+ antiporter would help terminate a [Ca2+]cyt signature through Ca2+ sequestration (Cheng et al., 2004). Other CAX are involved in root tolerance of salt stress and the pathways to their induction and regulation now need to be identified (Yamada et al., 2014). Salt exposure puts the plant’s ER under stress, leading to an accumulation of unfolded or misfolded proteins that could lead to cell death (Liu et al., 2007, 2011). Such ER stress triggers upregulation of a suite of responses termed the “Unfolded Protein Response” (UPR), in which folding capacity is upregulated (including by Ca2+-regulated chaperones), translation is curtailed and the ER-associated degradation pathway acts to lower the aberrant protein load (Deng et al., 2013; Ruberti et al., 2015; Hossain et al., 2016; Wan and Jiang, 2016). Expression of the ER Ca2+-binding chaperones Calnexin and Calreticulin has been shown to upregulated in the Arabidopsis UPR (Christensen et al., 2008; Liu et al., 2011) and it would now be interesting to test whether these are involved in regulating levels of Ca2+ in the ER under stress. However, expression of rice’s only calnexin gene is decreased under salt stress (Sarwat and Naqvi, 2013). It is not yet clear whether the salt-induced [Ca2+]cyt signal in roots (or other parts of the plant) helps initiate or regulate the UPR. It is feasible that the salt-induced [Ca2+]cyt increase activates the root PM Phospholipase C2 (Hunt et al., 2004) because the Arabidopsis loss of function mutant is hypersensitive to tunicamycin, which can induce the UPR (Kanehara et al., 2015). With IP3 as the product of Phospholipase C activity, this would implicate IP3-mediated release of Ca2+ from stores. Indeed, for Arabidopsis seedlings the application of 2-aminoethoxydiphenyl borate as an inhibitor of IP3-mediated Ca2+ release prevented salt stress-induced transcription of BIP1/2 (encoding an ER chaperone) as a diagnostic of a UPR response (Liu et al., 2011). Intriguingly, application of La3+ as a blocker of PM Ca2+ influx prevented the upregulation of UPR gene expression which was produced by application of spermine to Arabidopsis seedlings (Sagor et al., 2015). Exogenous spermine (as with salt) can depolarise the root epidermal PM (Pottosin et al., 2014) and can also attenuate hydroxyl radical-induced cation fluxes at root epidermal PM (Zepeda-Jazo et al., 2011). Whether spermine and salt stress share a common pathway to the UPR response merits further investigation. Water Availability is Signaled Through [Ca2+]cyt The hyperosmotic challenge in [Ca2+]cyt determinations is acute and does not mimic the chronic, progressive drought conditions that roots may face. Nevertheless, such studies have proved fruitful. As described above, the Arabidopsis OSCA1 PM Ca2+ influx channel drives the root’s initial hyperosmotic stress [Ca2+]cyt signal (Yuan et al., 2014). In an elegant study, heterologous expression of Arabidopsis genes in Chinese Hamster Ovary (CHO) cells containing the Fura-2 Ca2+-reporting dye lead to the identification of Calcium-permeable Stress-gated cation Channel1 (CSC1; Hou et al., 2014), a close relative of OSCA1. CSC1 has been characterized in CHO cells as a PM Ca2+-permeable channel that is activated by hyperosmotic stress. It is resident in plant PM (Hou et al., 2014) and is expressed in roots but to date an in planta role remains unreported. Targeting aequorin to the cytosolic face of the Arabidopsis vacuolar membrane has revealed the capacity of the vacuole to release Ca2+ in response to acute hyperosmotic stress, with pharmacological intervention suggesting an involvement of InsP3 (Knight et al., 1997). As with salt stress, hyperosmotic stress-induced [Ca2+]cyt increase could activate Phospholipases and initiate lipid signaling. Osmotic stress activates PLD in barley roots (Meringer et al., 2016). Although not tested directly, phosphatidic acid downstream of PLD could be involved in the activation of two sucrose non-fermenting-1 related protein kinase 2 proteins (SnRK2.4 and 2.10) in Arabidopsis roots under hyperosmotic stress. These SnRK2s are also activated under salt stress and relocate from the root epidermal cytosol to the PM; loss of function impairs root growth (McLoughlin et al., 2012). Work using Golgi-targeted aequorin in Arabidopsis seedlings has shown that an increase in Golgi [Ca2+] follows the [Ca2+]cyt increase induced by hyperosmotic stress, suggesting that this organelle helps terminate the [Ca2+]cyt signal (Ordenes et al., 2012). The ABA produced under drought stress inhibits primary root growth and [Ca2+]cyt is likely to play a role in the signaling pathway as exogenous ABA elevates root [Ca2+]cyt, which in Arabidopsis roots is controlled by the PM Proline-rich Extensin-like Receptor Kinase4 (PERK4) (Bai et al., 2009). PERK4’s extracellular domain is wall associated and its intracellular domain has kinase activity. PM HACCs lie downstream of ABA and PERK4. The perk4 mutant is not only impaired in ABA-induced HACC activity and [Ca2+]cyt activation but also in ABA inhibition of primary root elongation (Bai et al., 2009). This implies that [Ca2+]cyt elevation acts to arrest growth. RBOHD/F may lie downstream of PERK4 and upstream of HACCs in the root ABA pathway if the HACCs were ROS-activated. The rbohd/f mutant is impaired in both ABA-induced HACC activation and [Ca2+]cyt elevation (Jiao et al., 2013). Another possibility is that CNGCs are involved as ABA can increase cGMP levels (Isner et al., 2012.). GLRs are firmly implicated in the drought response; overexpression of rice GLR1 and GLR2 enhances drought tolerance of both rice and Arabidopsis (Lu et al., 2014). The [Ca2+]cyt increase is likely to lead to a transcriptional response as Ca2+-sensing proteins are activated. ABA increases abundance and activity of Arabidopsis CPK4 and CPK11 leading to phosphorylation of the ABA-responsive transcription factors ABF1 and ABF4 and induction of drought stress genes (Zhu et al., 2007). Drought leads to CAMTA1 activity and the regulation of discrete gene sets including those for drought recovery (Pandey et al., 2013). In maize roots, drought stress triggers activity of CIPK8 (which interacts with CBL1, 4, and 9) and likely leads to adaptive transcription (Tai et al., 2016). Hypoosmotic stress also elevates [Ca2+]cyt and is relevant to waterlogged soils. In Arabidopsis, root PM harbors two mechanosensitive Ca2+-permeable channels, MCA1 and MCA2 (Mid1-Complementing Activity; Kamano et al., 2015). MCA1 responds to hypoosmotic stress to elevate [Ca2+]cyt (Nakagawa et al., 2007). Rice only harbors an MCA1 in the PM. It is present in the root and mediates hypoosmotic shock-induced [Ca2+]cyt expression in cultured cells, probably lying upstream of NADPH oxidase activity (Kurusu et al., 2012). Abundant water not only exposes roots to potential hypoosmotic stress but also risks limiting their oxygen supply, the consequences of which are reviewed in the following section. Mechanistic Basis of [Ca2+]cyt Response to O2 Deficiency Remains Poorly Understood Oxygen deficiency (hypoxia) or absence (anoxia) causes transient increases in root [Ca2+]cyt but the signature is organ- and species-dependent (reviewed by Shabala et al., 2014). For example, when challenged by anoxia, root protoplasts from hypoxia-tolerant rice display a greater [Ca2+]cyt signature than hypoxia-intolerant wheat root protoplasts (Yemelyanov et al., 2011). The location of the Ca2+ influx also varies; use of pharmacological blockers showed that the rice signature was generated by both PM influx and store release whilst wheat appeared to rely solely on stores (Yemelyanov et al., 2011). The types of Ca2+ channels mediating the [Ca2+]cyt increase have yet to be identified in any species. However, those activated by PM voltage depolarization are implicated. As O2 deficiency lessens ATP production, activity of the PM H+-ATPase can be compromised thus resulting in a less negative (depolarized) PM voltage as H+ efflux is curtailed. The extent and duration of membrane depolarization varies with sensitivity to O2 deprivation and cell type. Values of -70 to -80 mV have been reported for O2-deprived barley root cells (Zeng et al., 2014) and theoretically these would be sufficient to activate PM depolarization-activated Ca2+ influx channels (Miedema et al., 2008) to contribute to an hypoxia/anoxia [Ca2+]cyt signature. Downstream of the [Ca2+]cyt signature, it is likely that CaM and ROPs (Rho GTPase) are activated (reviewed by Shabala et al., 2014). In Arabidopsis roots, hypoxia causes rapid upregulation of CML38 expression and this protein appears to require Ca2+ to associate with the cytosolic stress granules that form and store messenger RNA ribonucleoproteins (Lokdarshi et al., 2016). RBOH activity is firmly implicated in the response to low O2. In Arabidopsis, RBOHD expression is induced by hypoxia and is required for transcription of hypoxia-induced genes (Yang and Hong, 2015). Hypoxia also induces ethylene production and in wheat roots this causes RBOH induction (Yamauchi et al., 2014). A further level of regulation has been found in Arabidopsis; the Hypoxia Responsive Universal Stress Protein 1 (HRU1) interacts with GTP-bound ROP2 and RBOHD (Gonzali et al., 2015). HRU1 is one of 44 putative universal stress proteins in Arabidopsis. It exists as a cytosolic dimer but anoxia promotes monomer formation and increased association with the PM. There it is thought to be part of a mobile complex with ROP2 and AtRBOHD leading to activation of that NADPH oxidase (Gonzali et al., 2015). Although RBOH activity has been linked to PM Ca2+ channel activation in several abiotic stress scenarios, hypoxia and anoxia have yet to be tested. Ethylene has been shown to activate PM Ca2+-permeable channels (with a weak voltage dependence) in tobacco suspension cells (Zhao et al., 2007) and it remains a possibility that these may play a part in the root hypoxia [Ca2+]cyt signal with RBOH as an intermediary. Oxygen deprivation (and also sulfur and phosphate, Pi, deprivation) triggers programmed cell death (PCD) in mid-cortical cells for aerenchyma formation (Fagerstedt, 2010). This PCD is stimulated by ethylene and ROS and involves Ca2+ (Xu et al., 2013; Petrov et al., 2015). In wheat roots, anoxia causes mitochondria to release their Ca2+ and high [Ca2+]cyt causes cytochrome c release (Virolainen et al., 2002). The abnormal mitochondrial ultrastructure in Arabidopsis caused by hypoxia is partially phenocopied by loss of GLR3.5 from the inner mitochondrial membrane, suggesting that this channel must be deactivated during PCD (Teardo et al., 2015). Further channels must now be identified to understand the hypoxic/anoxic response. Recent work on Arabidopsis roots has shown distinct, cell-specific levels of [Ca2+]cyt after 24 h of hypoxia and highlighted the importance of CAX11 in controlling meristem [Ca2+]cyt (Wang et al., 2016). Loss of CAX4 function resulted in lower tolerance of hypoxia thus further demonstrating the importance of Ca2+ sequestration in this stress response (Wang et al., 2016). Ca2+ Signaling in Nutrient Deprivation is an Emerging Area Investigating [Ca2+]cyt elevation in response to nutrient deprivation or resupply is technically challenging, particularly if using aequorin. Nevertheless it is now clear that nutrient levels can induce [Ca2+]cyt changes and that downstream Ca2+ sensors regulate appropriate responses. The nutritional status of the root will have a part to play in determining the [Ca2+]cyt signatures, particularly of the endodermis as the extent of suberization is set by nutrition (Barberon et al., 2016) and suberin lamellae determine endodermal [Ca2+]cyt responses (Moore et al., 2002). To date, potassium, nitrate and boron have been studied. Potassium Plants must maintain cytosolic K+ at around 80 mM for optimal growth even though soil concentration may be sub-millimolar and they deploy a multigene family of K+ transporters in homeostatic control (Shabala and Pottosin, 2014). As extracellular K+ decreases, the root epidermal PM hyperpolarizes and [Ca2+]cyt increases (Demidchik et al., 2002). This increase can be abolished by gadolinium, implicating PM Ca2+ influx. The genetic identities of the PM Ca2+-permeable channels that would effect the [Ca2+]cyt increase remain unknown. The hyperpolarised PM correlates with induction of HAK5 expression in tomato root (Nieves-Cordones et al., 2008). HAK5 encodes the PM High Affinity K+ transporter that facilitates K+ uptake from low external concentrations that are thermodynamically unfavorable for channel-mediated influx. In Arabidopsis roots, low K+-induction of HAK5 expression (and other K+-deprivation genes) has been shown to depend on RBOHC activity and ROS (Shin and Schachtman, 2004). At present it is unclear whether the hyperpolarization of the PM directly activates RBOHC (which as an electron exporter may be voltage-dependent) or whether ROS-activated PM Ca2+ channels are involved in the K+-deprivation [Ca2+]cyt signal. HAK5 activity in Arabidopsis roots is regulated by CIPK23-mediated phosphorylation, downstream of CBL1, CBL8, CBL9, and CBL10 (Ragel et al., 2015). The extent to which the transcriptional response to low K+ availability is governed by [Ca2+]cyt is unclear but deficiency does result in upregulation of transcripts of Ca2+ signaling proteins (CaM, CBL, CIPK) in Arabidopsis seedlings (Armengaud et al., 2004) and sugarcane roots (Zeng et al., 2015). At higher external [K+], the AKT1 channel (Arabidopsis K+ Transporter1) facilitates uptake and its activity is promoted by CIPK23-mediated phosphorylation, downstream of CBL1 and CBL9 (Li et al., 2006; Cheong et al., 2007). This regulation is recapitulated in rice roots where CIPK23/CBL1 activate AKT1 (Li et al., 2014). Nitrate Nitrate is the most important form of nitrogen for agriculture and deprivation triggers significant transcriptional and developmental responses. The effect of nitrate withdrawal on [Ca2+]cyt has yet to be reported but recently it was shown that nitrate-starved Arabidopsis roots responded to nitrate resupply with a rapid, monophasic transient increase in [Ca2+]cyt that was sensitive to lanthanides and phospholipase C (PLC) inhibition (Riveras et al., 2015). Lanthanum also blocked nitrate-induced InsP3 production, suggesting that Ca2+ influx across the PM activated a PLC. The [Ca2+]cyt and InsP3 increases were entirely dependent on the PM nitrate influx transporter NRT1.1 (Nitrate Transporter1.1; Riveras et al., 2015). By using the nrt1.1 mutant and pharmacological blockers, nitrate-induced gene transcription was also found to lie downstream of NRT1.1, and [Ca2+]cyt elevation from PM influx and InsP3-gated store release. Calcium is also key to the regulation of nitrate uptake capacity as CIPK23, which is activated by CBL9 and CBL1, and dephosphorylated by ABI2 (a member of the PP2C protein phosphatase family; Léran et al., 2015,), phosphorylates NRT1.1 under low nitrate condition, thus converting it from a low to high affinity transporter (Ho et al., 2009). By contrast CIPK8 positively regulates the low-affinity phase of the nitrate primary response which includes transcriptional regulation, but its regulation of primary root elongation is concentration independent in Arabidopsis (Hu et al., 2009). CBL7, which is upregulated under nitrate deprivation, positively regulates the nitrate-dependent induction of NTR2.4 and NTR2.5 gene expression (Ma Q. et al., 2015). Given the lack of a [Ca2+]cyt reporter line available in crops up until recently for rice, little is known about nitrate deficiency-induced [Ca2+]cyt signaling but CaM protein abundance of wheat roots declines under nitrate deficiency, suggesting a remodeling of signaling systems (Jiang et al., 2015). Boron Boron deficiency is widespread worldwide and particularly prevalent in China (Shorrocks, 1997). As B plays a dominant role in co-ordinating cell wall structure (Kobayashi et al., 1996), changes in cell wall stability are likely to influence the signal relayed into the cell upon B deprivation and indeed a rapid change in cell wall modulus has been observed (Goldbach et al., 2001). Early work in Vicia faba showed a release of membrane-bound Ca2+ into the apoplast (Mühling et al., 1998), raising the possibility of Ca2+ signaling during the early stages of B deficiency. Increased levels of both [Ca2+]cyt and ROS have been suggested to lead to the increased root hair growth known to occur under B deprivation (González-Fontes et al., 2016). In cultured tobacco cells, transcriptional changes in response to short-term B deprivation (1 h) were abolished when withdrawing Ca2+ from the growth medium or upon treatment with the Ca2+ channel blocker lanthanum, thus implicating PM Ca2+ influx channels in generating a [Ca2+]cyt signal (Koshiba et al., 2010). However, a transient [Ca2+]cyt signal in direct response to B withdrawal has yet to be reported. Rather, work has focused on the possible remodeling of Ca2+ transport and signaling as a consequence of B deprivation. Challenging cultured tobacco cells with Ca2+ resulted in a higher amplitude of [Ca2+]cyt transient in B-deprived cells (1 h deprivation) than those grown under replete conditions (Koshiba et al., 2010). This suggests that B deprivation rapidly “resets” the PM’s Ca2+ transport systems to generate altered [Ca2+]cyt responses. The [Ca2+]cyt response of B-deprived cells was sensitive to lanthanum and diphenyleneiodonium, pointing to the involvement of PM Ca2+ channels and NADPH oxidases respectively (Koshiba et al., 2010). Arabidopsis roots expressing the YC3.6 [Ca2+]cyt reporter exhibited higher levels of [Ca2+]cyt at the apex than controls after 6 and 24 h of B deprivation (Quiles-Pando et al., 2013). This time course of B deprivation also resulted in significant upregulation of CNGC19 (encoding a vacuolar channel), four genes of the ACA family of PIIB-type Ca2+-ATPases (ACA1,10,12,13) and CAX3 encoding a vacuolar cation-H+ antiporter (Quiles-Pando et al., 2013). This suite of transporters could effect Ca2+ efflux from the vacuole (CNGC19) to increase [Ca2+]cyt with clearance to the apoplast by ACA10-13 and sequestration to the vacuole by CAX3. How they are regulated remains to be determined, as does the involvement of the structurally compromised wall and the consequence of this higher level of apical [Ca2+]cyt. The area of higher [Ca2+]cyt reported appears to correspond with the zone of inhibition of primary root elongation and the induction of cell death (Oiwa et al., 2013; Camacho-Cristobal et al., 2015). Once again, ethylene production appears to be upstream of NADPH oxidase-driven ROS production in growth arrest and death (Oiwa et al., 2013; Camacho-Cristobal et al., 2015). Eight CML genes were also significantly upregulated after a day’s B deprivation of Arabidopsis roots (CML11,12,23,24, 30.37,45.47), as were three CPK genes (CPK1,28,29) all suggesting a distinct change in intracellular Ca2+ signaling (Quiles-Pando et al., 2013). This B deprivation also caused upregulation of WRKY transcription factors (TF) (WRKY38,40,46), two three MYB family TF (MYB14,15,78) and downregulation of two BZIP family TFs (bZIP34,61) (González-Fontes et al., 2016). B deprivation has also been found to promote the senescence-associated WRKY6 TF in the root tip (Kasajima et al., 2010). It has been suggested that the CMLs and CPKs upregulated by B deprivation are part of the chain of events leading to TF activation in the nucleus (González-Fontes et al., 2016) and this now requires direct testing. CIPK8 also positively regulates nitrate induced up-regulation of BOR1, a gene encoding a boron transporter responsible for xylem loading (Hu et al., 2009), suggesting that root signaling results in preservation of shoot B supply. Heavy Metal Stress has the Capacity to Distort Ca2+ Signaling At the opposite end of nutritional deprivation is heavy metal stress. Industrial activity, mining and modern agricultural practices can lead to soil contamination by heavy metals (defined here as 7 g/cm3 and above). Although some of these metals (such as Zn, Cu) are required as micronutrients they can be damaging in excess whilst others (such as Cd) have no physiological role and can be deleterious even at low concentrations, often impairing mineral nutrition. The consequences of heavy metal exposure have been reviewed recently by Singh et al. (2016) and these authors explore signaling pathways (although not explicitly addressing Ca2+), intersects with hormonal responses and detoxification. Cadmium Cadmium is a particular threat to Ca2+-based processes because of its similar size. A recent review by Chmielowska-Bak et al. (2014) summarized the effects of Cd on ROS accumulation, NO accumulation, MAP kinase activation and downstream responses in a wide range of plant systems and, importantly, did this as a function of Cd exposure. With regards to Ca2+ a key question is whether Cd (as a Ca2+ “substitute”) generates a signal in its own right or whether its impairment of Ca2+ homeostasis will be, in effect, the signal. Certainly Cd depolarises the rice root epidermal PM, which would impair Ca2+ influx and results in inhibition of root elongation (Li et al., 2012). Cd can permeate guard cell PM Ca2+ channels (Perfus-Barbeoch et al., 2002) and may compete with Ca2+ for entry into rice root hairs through PM HACC, thus limiting Ca2+ influx (Li et al., 2012). Competitive effects appear likely given the ability of exogenous Ca2+ addition to alleviate Cd inhibition of root growth of both terrestrial and aquatic plants, implying competition for uptake (Zhang et al., 2012; Rodriguez-Hernandez et al., 2015). Additionally, competitive inhibition by Cd of Ca2+ uptake through the wheat LCT1 transporter expressed in yeast has been observed (Clemens et al., 1998). Root hair growth in Arabidopsis is inhibited by Cd; it inhibits Ca2+ influx and so dissipates the apical [Ca2+]cyt gradient needed for growth (Fan et al., 2011). This could reflect block of root hair Ca2+ influx channels or Cd outcompeting Ca2+ for that influx pathway. However, Yeh et al. (2007) observed an increase in rice root [Ca2+]cyt following 15 min of Cd exposure, with a more tolerant variety sustaining a greater increase than a sensitive variety. In roots of the aquatic plant Typha latifolia, Cd exposure increases transcription of TPC1 which suggests that vacuoles might release Ca2+ to the cytosol (Rodriguez-Hernandez et al., 2015). Inhibitor treatments indicated involvement of NADPH oxidases, hydroxyl radicals and CIPK upstream of a MAP kinase induction that linked to tolerance (Yeh et al., 2007). Cd stimulates expression and activity of NADPH oxidases in cucumber roots but it is not known if these responses are Ca2+-dependent (Jakubowska et al., 2015). It is worth noting that it is not only an increase in [Ca2+]cyt that could stimulate NADPH oxidase activity but that also the restriction of Ca2+ entry to the cytosol could cause stimulation (Mortimer et al., 2008). It seems likely that in Arabidopsis roots it would be CAXs predominating in terminating a Cd-induced increase in [Ca2+]cyt. CAX4 mediates Cd and Ca2+ transport into Arabidopsis thaliana vacuoles and is needed for correct growth under Cd stress (Mei et al., 2009). CAX1 expression is higher in roots of the Cd-tolerant A. halleri compared to the sensitive A. thaliana (Baliardini et al., 2015). A key point for future studies is the intersect between Cd and hormones in relation to [Ca2+]cyt. Exogenous Ca2+ can ameliorate Cd’s inhibition of Arabidopsis root growth by counteracting effects of NO on auxin homeostasis (Hu et al., 2013; Li P. et al., 2016; Yuan and Huang, 2016). Cd also interferes with auxin homeostasis in barley roots (Zelinova et al., 2015). Auxin itself can increase Arabidopsis root [Ca2+]cyt and this increase is mediated by CNGC14 at the PM, downstream of an unidentified auxin receptor (Shih et al., 2015). This begs the question of whether CNGC14 is an entry route for Cd and the pathway to disrupted auxin homeostasis. Additionally, Cd has been described as a “metallohormone” in that it triggers expression of brassinosteroid-regulated genes in Arabidopsis roots (Villiers et al., 2012). Brassinosteroids are themselves capable of transiently elevating [Ca2+]cyt in Arabidopsis roots (through PM Ca2+ influx) and activate a possible DACC in wheat root PM (Straltsova et al., 2015). Again this raises the question of channel identity to help elucidate the relationship between Cd and brassinosteroid signaling and combat the effects of this potent soil contaminant. Copper, Gadolinium and Lead In contrast to Cd, transition heavy metals could be capable of generating ROS directly and so perturb [Ca2+]cyt signaling. Transition metals can catalyze production of hydroxyl radicals from superoxide anion and hydrogen peroxide through the Haber-Weiss reaction, with Cu+ and Fe2+ catalyzing hydroxyl radical production from hydrogen peroxide through the Fenton reaction (Richards et al., 2015). Unless levels of these catalytic metals are tightly controlled, production of hydroxyl radicals (the most potent of the ROS) could inflict significant oxidative damage. Taking Cu as the exemplar, it plays a positive role in hydroxyl radical-activated cell wall loosening for root elongation (Fry et al., 2002) and this could be coupled in Arabidopsis roots to hydroxyl radical-activation of PM Annexin1-mediated Ca2+ influx to stimulate exocytosis and growth (Foreman et al., 2003; Laohavisit et al., 2012). This model proposes regulation by extracellular hydroxyl radicals while Rodrigo-Moreno et al. (2013) have proposed that Cu+ binding at the intracellular face of the PM in Arabidopsis root tips catalyzes hydroxyl radical production to regulate ion flux. This would allow coupling of Cu transport into the root to regulation of PM Ca2+ influx channels. Certainly, low levels of Cu can promote elongation of Arabidopsis primary root and this effect is diminished by blocking PM Ca2+ influx channels (Demidchik et al., 2003). The effects of extracellular Cu on PM Ca2+ channels and therefore on [Ca2+]cyt signaling are likely to be complex. Electrophysiological analysis of Arabidopsis root epidermal PM has shown that Cu not only activates Ca2+ channels through ROS production (Annexin 1; Laohavisit et al., 2012) but also blocks channels, the molecular identity of which remains unknown (Demidchik et al., 2003). Catalytic production of ROS by Cu is not the only route to modulating [Ca2+]cyt. Longer-term exposure to Cu can stimulate exocytosis-mediated ROS production. Lin et al. (2013) found that inhibiting vesicle traffic with brefeldin also inhibited Cu-stimulated ROS production in rice roots. Whether this involved NADPH oxidase or other ROS generators remains to be determined. Rice root [Ca2+]cyt increases in response to Cu addition and could link to NADPH oxidases and CIPK activity leading to MAP kinase activation (Yeh et al., 2007). It triggers oxidative stress in Populus roots that leads to regulation of CaM genes, implicating perturbation of [Ca2+]cyt (Guerra et al., 2009). In common with Cd, excess Cu also alters auxin homeostasis in roots and interferes with NO signaling (Lequeux et al., 2010; Kolbert et al., 2012). In excess, Cu stunts Arabidopsis root and root hair elongation and can inhibit lateral root outgrowth (Kolbert et al., 2012). Accumulation of lignin has been observed in both Arabidopsis and rice roots (Lequeux et al., 2010; Liu et al., 2015) and in the former accumulation is associated with the endodermis. It would now be interesting to ascertain whether such wall modification changes [Ca2+]cyt signals in central cells or affects [Ca2+]cyt propagative signaling to the shoot. Root tip cell death has also been observed (Huang et al., 2007; Lequeux et al., 2010; Rodrigo-Moreno et al., 2013). Cu-induced cell death in rice roots was attenuated by chelating extracellular Ca2+ thus implicating Ca2+ influx across the PM (Huang et al., 2007) while in Arabidopsis roots cell death was attenuated by addition of Gd3+ or verapamil as PM Ca2+ channel blockers (Rodrigo-Moreno et al., 2013). Gd3+ is routinely used as a PM Ca2+ channel blocker, for example with proven efficacy against Arabidopsis root epidermal and root hair HACC, DACC and VICC (Véry and Davies, 2000; Demidchik et al., 2002; Miedema et al., 2008). It is also effective against the root PM hydroxyl radical-activated and H2O2-activated Ca2+ influx channels (Demidchik et al., 2003, 2007) and the Annexin1-mediated pathway (Laohavisit et al., 2012). As a consequence Gd3+ lowers Ca2+ influx (measured with 45Ca; Demidchik et al., 2002). From this it can be anticipated that Gd3+ would have the capacity to dampen stress-induced [Ca2+]cyt signaling in roots, including that mediated by oxidative stress (Demidchik et al., 2003, 2007) and elicited by NaCl (Laohavisit et al., 2013). Pb is less well studied in terms of [Ca2+]cyt. Exogenous Ca2+ can protect against root Pb accumulation, suggesting competitive uptake (Rodriguez-Hernandez et al., 2015). However, Pb was found to evoke Ca2+ accumulation and diphenyleneiodinium-sensitive ROS increase in rice roots, the latter leading to MAPK activation (Huang and Huang, 2008). Entry of Pb into Arabidopsis roots involves CNGC1 as its deletion confers greater tolerance but whether Pb permeates this channel is unknown (Sunkar et al., 2000). Pb has also been shown to increase expression of TPC1 in Typha roots (Rodriguez-Hernandez et al., 2015) but whether this results in increased [Ca2+]cyt in this or other roots is not known yet. Mechanical Stress Roots experience a range of mechanical stimuli, induced as they encounter soil particles or neighboring roots. An increase in [Ca2+]cyt together with an apoplastic alkanisation are early effects induced by mechanical stimuli. One of the known downstream events from the increase in [Ca2+]cyt is the upregulation of touch-sensitive genes such as CML12 and CML24 (Braam, 1992). Mechanically triggered [Ca2+]cyt changes are dependent on the type of stimulus and the responding tissue (Legue et al., 1997; Monshausen et al., 2009). For example, manually bending an Arabidopsis root induces a rapid, biphasic increase in [Ca2+]cyt on the convex side of the roots where cells are stretched (Monshausen et al., 2009) while previously compressed cells on the concave side of the roots show a monophasic and less intense increase in [Ca2+]cyt upon return to their resting position. The source of Ca2+ is likely to be extracellular (Monshausen et al., 2009; Richter et al., 2009; Kurusu et al., 2012); with internal stores being mobilized to amplify the response (Chehab et al., 2009; Toyota and Gilroy, 2013). Over-expressing MCA1 in Arabidopsis lead to an enhanced [Ca2+]cyt transient post mechanical stimulation by the addition of a membrane crenator, however, the mca1 mutant showed no difference from the wild type (Nakagawa et al., 2007). Accordingly, Shih et al. (2014) showed no change in mca1 apoplastic alkalinisation following root bending, which is closely related to the [Ca2+]cyt signature. Thus some levels of compensation occur in mutants deficient in mechanosensitive Ca2+ channels. In contrast to the MCAs, MSLs (MscS-Like) were identified in Arabidopsis due to their sequence similarity to the bacterial Mechanosensitive channels of small conductance (MscS) (Haswell et al., 2008). These are encoded by a multiple gene family of 10 members, with MSL9 and MSL10 found in the PM of root cells and required for a mechanosensitive channel activity. They are predominantly anion channels and their relationship to mechano-stimulated [Ca2+]cyt increase has yet to be shown but may be through PM voltage regulation. At this point, it is unclear whether these channels represent mechano-sensors or act downstream of yet unknown mechano-sensors. Recently, the PM receptor-like kinase FERONIA has been implicated in regulating the Arabidopsis mechano-stimulated [Ca2+]cyt increase and downstream transcriptional regulation (Shih et al., 2014). feronia mutants lack the second peak of the biphasic increase in [Ca2+]cyt elicited in stretched cells by manual bending. However, the mechanism by which FERONIA can regulate [Ca2+]cyt remains unclear as its kinase activity is not essential and targets are unknown. As far as we know, this study was the first in which a mutant with an aberrant mechano-stimulated [Ca2+]cyt increase also showed a root skewing phenotype. Root skewing (deviation from the vertical when grown on vertical or inclined agar) may be influenced by mechano-sensing. At this point, it is unclear whether FERONIA also plays a role in the mechanical induction of lateral root formation (Richter et al., 2009). Cold Stress Cold plays a key role in the regulation of physiology and development; the signaling processes relaying non-stressful temperatures (12°C and above) have been reviewed by Wigge (2013). The signaling cascades activated by cold stress (typically 4°C experimentally) and their relations with hormonal signaling have been reviewed by Knight and Knight (2012), Jeon and Kim (2013), Shi et al. (2015) and Eremina et al. (2016). Most studies on the effect of cold stress on [Ca2+]cyt report on seedlings or leaves. From these it is well established that [Ca2+]cyt elevation involves both Ca2+ influx across the PM and release from predominantly vacuolar stores (e.g., Knight et al., 1996; Mazars et al., 1997; Knight and Knight, 2000: Zhu et al., 2013). By using an extracellular Ca2+-reporting microelectrode, Sulaiman et al. (2012) confirmed that Ca2+ influx from the extracellular space is a component of the Arabidopsis root response to cold stress. For Arabidopsis roots, the [Ca2+]cyt response to cold stress measured using aequorin is lower in amplitude and of much shorter duration than leaves (Zhu et al., 2013). Using aequorin Plieth et al. (1999) determined that the faster the rate of cooling Arabidopsis roots, the greater the [Ca2+]cyt response. Sensitivity and magnitude of the [Ca2+]cyt cooling response were enhanced by low temperature but repeated exposure to cold lead to desensitizing of the response. These results imply the existence of Ca2+ transport systems that could be regulated at the post-translational and possibly transcriptional levels. In Arabidopsis, cold-induced [Ca2+]cyt elevation would activate a root PM calcium/calmodulin-regulated receptor-like kinase (CRLK1; Yang et al., 2010). This would then activate a specific mitogen-activated protein kinase kinase kinase (MEKK1) that then targets protein kinase kinase2 (MKK2; Furuya et al., 2013). This pathway leads to gene activation and freezing tolerance (Furuya et al., 2014). [Ca2+]cyt elevation would also activate CIPK3, that acts an intersect with ABA signaling (Kim et al., 2003). Part of the transcriptional response to cold in Arabidopsis could be activated by increased [Ca2+]cyt through CAMTA3 regulation of the transcriptional regulator CBF2 (C-Repeat/Dehydration Responsive Element Binding Factor2) (Doherty et al., 2009). Cold stress has been found to depolarise the PM of root cells from cucumber and Triana bogotensis (Lyalin and Ktitorova, 1969; Minorsky and Spanswick, 1989), consistent with the effect of Ca2+ influx across the PM. A modeling exercise revealed the possible importance of a PM DACC in cold stress-induced [Ca2+]cyt elevation of roots (White, 2009). Cold has been shown to activate a PM DACC in leaf protoplasts (Carpaneto et al., 2007) but this has not yet been shown for roots. As microtubule destabilization activates the Arabidopsis root PM DACC (Thion et al., 1998), this could link to the increased cold-induced [Ca2+]cyt signal observed when microtubules are disrupted in Nicotiana plumbaginifolia leaf protoplasts (Mazars et al., 1997). Actin depolymerization is implicated in cold-induced Ca2+ influx across the PM into Medicago sativa suspension cells (Orvar et al., 2000). This in turn could relate to the cold-induced activation of pear pollen PM HACC that involves actin depolymerization (Wu et al., 2012). It would be timely therefore to investigate root PM for cold effects on HACC and DACC, exploring also the involvement of the cytoskeleton. In common with mechanical stress, cold stress could perturb the PM bilayer sufficiently to change Ca2+ channel activity. Work on M. sativa suspension cells showed that increasing membrane fluidity at 4°C prevented the Ca2+ influx across the PM that is necessary to trigger gene expression for freezing tolerance (Orvar et al., 2000). Conversely, imposing membrane rigidity at normal temperature triggers activation of the Ca2+-dependent MEKK1-MKK2-MPK4 pathway in Arabidopsis seedlings (Furuya et al., 2014). The possibility that mechanosensitive channels are involved in cold-induced [Ca2+]cyt signaling merits investigation. However, recent work on rice roots indicates another route to Ca2+ influx. COLD1 has been identified as a PM activator of the heterotrimeric G protein subunit α (Ma Y. et al., 2015). It is required for cold-activated Ca2+ influx to roots (measured using extracellular Ca2+-reporting microelectrode) and elevation of root [Ca2+]cyt (aequorin and cameleon determinations). This indicates that heterotrimeric G proteins lie upstream of PM Ca2+ channels in the cold response. In common with other stress responses, NADPH oxidase activity also appears in cold stress and needs to be placed in relation to Ca2+ channels. Arabidopsis roots exposed to cold stress use AtSRC2 (Soybean gene Regulated by Cold2) to enhance Ca2+ activation of the AtRBOHF NADPH Oxidase (Kawarazaki et al., 2013). This suggests that NADPH oxidase activity can lie downstream of Ca2+ channel activity. Curtailing cold-induced [Ca2+]cyt elevation could involve vacuolar CAX1 activity in Arabidopsis (Catala et al., 2003). Conclusion and Future Prospects A repeated message from this review is how incomplete our knowledge is of the channels mediating stress-induced [Ca2+]cyt increases, and by extension those of organelles. Many members of channel gene families still await characterization. The identification of new families of channels is challenging and will require different approaches linking forward and reverse genetics to electrophysiology. Targets of Ca2+-binding and interacting proteins also require further study. Components common to different abiotic stresses are emerging such as Arabidopsis CIPK23 in K+ and nitrate deprivation. These common regulatory components are likely to represent critical steps where complex stress signals encountered in the soil are integrated in unified responses. Receptor like kinases such as FERONIA or PERK4 have emerged as new components in [Ca2+]cyt signaling and perhaps other related proteins will be found to have a role in abiotic stress signaling. Remodeling of calcium signaling machinery after stress is also apparent with the possibility of components common to different stresses. For example, Arabidopsis CNGC19 is upregulated under B limitation and salinity stress (Kugler et al., 2009). Finally, [Ca2+]cyt-dependent transcriptional responses can be delineated and future work could include the impact of stress-induced calcium signaling on epigenetic inheritance (Sani et al., 2013; Probst and Scheid, 2015). Author Contributions All authors listed, have made substantial, direct and intellectual contribution to the work, and approved it for publication. KW, EM, SS, and JD wrote the review. KW produced the figures. Conflict of Interest Statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Funding. Funding for this work was from the BBSRC (BB/K009869/1 and Doctoral Training Programme) and the University of Cambridge Broodbank Trust. ==== Refs References Armengaud P. Beitling R. Amtmann A. (2004 ). The potassium-dependent transcriptome of Arabidopsis reveals a prominent role of jasmonic acid in nutrient signaling. Plant Physiol. 136 2556 –2576 . 10.1104/pp.104.046482 15347784 Bai L. Zhang G. Zhou Y. Zhang Z. Wang W. Du Y. (2009 ). Plasma membrane-associated proline-rich extension-like receptor kinase 4, a novel regulator of Ca2+ signaling, is required for abscisic acid responses in Arabidopsis thaliana. Plant J. 60 314 –327 . 10.1111/j.1365-313X.2009.03956.x 19566594 Baliardini C. Meyer C.-L. Salis P. Saumitou-Laprade P. Verbruggen N. (2015 ). CATION EXCHANGER1 cosegregates with cadmium tolerance in the metal hyperaccumulator Arabidopsis halleri and plays a role in limiting oxidative stress in Arabidopsis Spp. Plant Physiol. 169 549 –559 . 10.1104/pp.15.01037 26162428 Barberon M. Vermeer J. E. M. De Bellis D. Wang P. Naseer S. Anderson T. G. (2016 ). Adaptation of root function by nutrient-induced plasticity of endodermal differentiation. Cell 164 447 –459 . 10.1016/j.cell.2015.12.021 26777403 Batelli G. Verslues P. R. Agius F. Qiu Q. Fujii H. Pan S. (2007 ). SOS2 promotes salt tolerance in part by interacting with the vacuolar H+-ATPase and upregulating its transport activity. Mol. Cell Biol. 27 7781 –7790 . 10.1128/MCB.00430-07 17875927 Behera S. Wang N. Zhang C. Schmitz-Thom I. Strohkamp S. Schueltke S. (2015 ). Analyses of Ca2+ dynamics using a ubiquitin-10 promoter-driven Yellow Cameleon 3.6 indicator reveal reliable transgene expression and differences in cytoplasmic Ca2+ responses in Arabidopsis and rice (Oryza sativa) roots. New Phytol. 206 751 –760 . 10.1111/nph.13250 25641067 Bonza M. C. Loro G. Behera S. Wong A. Kudla J. Costa A. (2013 ). Analyses of Ca2+ accumulation and dynamics in the endoplasmic reticulum of Arabidopsis root cells using a genetically encoded cameleon sensor. Plant Physiol. 163 1230 –1241 . 10.1104/pp.113.226050 24082028 Bonza M. C. Luoni L. Olivari C. DeMichelis M. I. (2016 ). Plant type 2B Ca2+-ATPases: the diversity of isoforms of the model plant Arabidopsis thaliana. Adv. Biochem. Health Dis. 14 227 –241 . Boursiac Y. Harper J. F. (2007 ). The origin and function of calmodulin regulated Ca2+ pumps in plants. J. Bioen. Biomem. 39 409 –414 . 10.1007/s10863-007-9104-z Braam J. (1992 ). Regulated expression of the calmodulin-related TCH genes in cultured Arabidopsis cells – induction by calcium and heat shock. Proc. Natl. Acad. Sci. U.S.A. 89 3213 –3216 . 10.1073/pnas.89.8.3213 1373491 Camacho-Cristobal J. J. Martin-Rejano E. M. Begona Herrera-Rodriguez M. Teresa Navarro-Gochicoa M. Rexach J. Gonzalez-Fontes A. (2015 ). Boron deficiency inhibits root cell elongation via an ethylene/auxin/ROS-dependent pathway in Arabidopsis seedlings. J. Expt. Bot. 66 3831 –3840 . 10.1093/jxb/erv186 Carpaneto A. Ivashikina N. Levchenko V. Krol E. Jeworutzki E. Zhu J.-K. (2007 ). Cold transiently activates calcium-permeable channels in Arabidopsis mesophyll cells. Plant Physiol. 143 487 –494 . 10.1104/pp.106.090928 17114272 Carroll A. D. Moyen C. Van Kesteren P. Tooke F. Battey N. H. Brownlee C. (1998 ). Ca2+, annexins, and GTP modulate exocytosis from maize root cap protoplasts. Plant Cell 10 1267 –1276 . 10.1105/tpc.10.8.1267 9707528 Catala R. Santos E. Alonso J. M. Ecker J. R. Martinez-Zapater J. M. Salinas J. (2003 ). Mutations in the Ca2+/H+ transporter CAX1 increase CBF/DREB1 expression and the cold-acclimation response in Arabidopsis. Plant Cell 15 2940 –2951 . 10.1105/tpc.015248 14630965 Chehab E. W. Eich E. Braam J. (2009 ). Thigmomorphogenesis: a complex plant response to mechanostimulation. J. Expt. Bot. 60 43 –56 . 10.1093/jxb/ern315 Cheng N. H. Pittman J. K. Zhu J. K. Hirschi K. D. (2004 ). The protein kinase SOS2 activates the Arabidopsis H+/Ca2+ antiporter CAX1 to integrate calcium transport and salt tolerance. J. Biol. Chem. 279 2922 –2926 . 10.1074/jbc.M309084200 14583601 Cheong Y. H. Pandey G. K. Grant J. J. Batistic O. Li L. Kim B.-G. (2007 ). Two calcineurin B-like calcium sensors, interacting with protein kinase CIPK23, regulate leaf transcription and root potassium uptake in Arabidopsis. Plant J. 52 223 –239 . 10.1111/j.1365-313X.2007.03236.x 17922773 Chmielowska-Bak J. Gryl J. Rucinska-Sobkowiak R. Arasimowicz-Jelonek M. Deckert J. (2014 ). The new insights into cadmium sensing. Front. Plant Sci. 5 :245 10.3389/fpls.2014.00245 Choi W.-G. Toyota M. Kim S.-H. Hilleary R. Gilroy S. (2014 ). Salt stress-induced Ca2+ waves are associated with rapid, long-distance root-to-shoot signaling in plants. Proc. Natl. Acad. Sci. U.S.A. 111 6497 –6502 . 10.1073/pnas.1319955111 24706854 Christensen A. Svensson K. Persson S. Jung J. Michalak M. Widell S. (2008 ). Functional characterization of Arabidopsis calreticulin1a: A key alleviator of endoplasmic reticulum. Plant Cell Physiol. 49 912 –924 . 10.1093/pcp/pcn065 18436549 Chung J. S. Zhu J. K. Bressan R. A. Hasagawa P. M. Shi H. (2008 ). Reactive oxygen species mediate Na+-induced SOS1 mRNA stability in Arabidopsis. Plant J. 53 554 –565 . 10.1111/j.1365-313X.2007.03364.x 17996020 Clemens S. Antosiewicz D. M. Ward J. M. Schachtman D. B. Schroeder J. I. (1998 ). The plant cDNA LCT1 mediates the uptake of calcium and cadmium in yeast. Proc. Natl. Acad. Sci. U.S.A. 95 12043 –12048 . 10.1073/pnas.95.20.12043 9751787 Connorton J. M. Webster R. E. Cheng N. H. Pittman J. K. (2012 ). Knockout of multiple Arabidopsis cation/H+ exchangers suggests isoform-specific roles in metal stress response, germination and see mineral nutrition. PLoS ONE 7 :e47455 10.1371/journal.pone.0047455 Dark A. M. Demidchik V. Richards S. L. Shabala S. N. Davies J. M. (2011 ). Release of extracellular purines from plant roots and effect on ion fluxes. Plant Signal. Behav. 6 1855 –1857 . 10.4161/psb.6.11.17014 22057319 Davies J. M. (2014 ). Annexin-mediated calcium signaling in plants. Plants 3 128 –140 . 10.3390/plants3010128 27135495 Demidchik V. Bowen H. C. Maathuis F. J. M. Shabala S. N. Tester M. A. White P. J. (2002 ). Arabidopsis thaliana root non-selective cation channels mediate calcium uptake and are involved in growth. Plant J. 32 799 –808 . 10.1046/j.1365-313X.2002.01467.x 12472694 Demidchik V. Cuin T. A. Svistunenko D. Smith S. J. Miller A. J. Shabala S. (2010 ). Arabidopsis root K+-efflux conductance activated by hydroxyl radicals: single-channel properties, genetic basis and involvement in stress-induced death. J. Cell Sci. 123 1468 –1479 . 10.1242/jcs.064352 20375061 Demidchik V. Shabala S. N. Coutts K. B. Tester M. A. Davies J. M. (2003 ). Free oxygen radicals regulate plasma membrane Ca2+ and K+-permeable channels in plant root cells. J. Cell Sci. 116 81 –88 . 10.1242/jcs.00201 12456718 Demidchik V. Shabala S. N. Davies J. M. (2007 ). Spatial variation in H2O2 response of Arabidopsis thaliana epidermal Ca2+ flux and plasma membrane Ca2+ channels. Plant J. 49 377 –386 . 10.1111/j.1365-313X.2006.02971.x 17181775 Demidchik V. Shang Z. Shin R. Thompson E. P. Rubio L. Laohavisit A. (2009 ). Plant extracellular ATP signalling by plasma membrane NADPH oxidase and Ca2+ channels. Plant J. 58 903 –913 . 10.1111/j.1365-313X.2009.03830.x 19220789 Deng Y. Srivatava R. Howell S. H. (2013 ). Protein kinase and ribonuclease domians of IRE1 confer stress tolerance, vegetative growth and reproductive development. Proc. Natl. Acad. Sci. U.S.A. 110 19633 –19638 . 10.1073/pnas.1314749110 24145452 DeWald D. B. Torabinejad J. Jones C. A. Shope J. C. Cangelosi A. R. Thompson J. E. (2001 ). Rapid accumulation of phosphatidylinositol 4,5-bisphosphate and inositol 1,4,5-trisphosphate correlates with calcium mobilization in salt-stressed Arabidopsis. Plant Physiol. 126 759 –769 . 10.1104/pp.126.2.759 11402204 Dinneny J. R. Long T. A. Wang J. Y. Jung J. W. Mace D. Pointer S. (2008 ). Cell identity mediates the response of Arabidopsis roots to abiotic stress. Science 320 942 –945 . 10.1126/science.1153795 18436742 Doherty C. J. Van Buskirk H. A. Myers S. J. Thomashow M. F. (2009 ). Roles for Arabidopsis CAMTA transcription factors in cold-regulated gene expression and freezing tolerance. Plant Cell 21 972 –984 . 10.1105/tpc.108.063958 19270186 Donaldson L. Ludidi N. Knight M. R. Gehring C. Denby K. (2004 ). Salt and osmotic stress cause rapid increases in Arabidopsis thaliana cGMP levels. FEBS Letts. 569 317 –320 . 10.1016/j.febslet.2004.06.016 15225654 Drerup M. M. Schluecking K. Hashimoto K. Manishankar P. Steinhorst L. Kuchitsu K. (2013 ). The Calcineurin B-Like calcium sensors CBL1 and CBL9 together with their interacting protein kinase CIPK26 regulate the Arabidopsis NADPH oxidase RBOHF. Mol. Plant 6 559 –569 . 10.1093/mp/sst009 23335733 Dressler L. Golbik R. Ulbrich-Hofmann R. (2014 ). Lanthanides as substitutes for calcium ions in the activation of plant alpha-type phospholipase D. Biol. Chem. 395 791 –799 . 10.1515/hsz-2014-0112 24643913 Eremina M. Rozhon W. Poppenberger B. (2016 ). Hormonal control of cold stress responses in plants. Cell. Mol. Life Sci. 73 797 –810 . 10.1007/s00018-015-2089-6 26598281 Fagerstedt K. V. (2010 ). “Programmed cell death and aerenchyma formation under hypoxia ,” in Waterlogging Signalling and Tolerance in Plants , eds Mancuso S. Shabala S. (Berlin : Springer ), 99 –118 . Fan J.-L. Wei X.-Z. Wan L.-C. Zhang L.-Y. Zhao X.-Q. Liu W.-Z. (2011 ). Disarrangement of actin filaments and Ca2+ gradient by CdCl2 alters cell wall construction in Arabidopsis thaliana root hairs by inhibiting vesicular trafficking. J. Plant Physiol. 168 1157 –1167 . 10.1016/j.jplph.2011.01.031 21497412 Fischer C. Kugler A. Hoth S. Dietrich P. (2013 ). An IQ domain mediates the interaction with calmodulin in a plant cyclic nucleotide-gated channel. Plant Cell Physiol. 54 573 –584 . 10.1093/pcp/pct021 23385145 Foreman J. Demidchik V. Bothwell J. H. F. Mylona P. Miedema H. Torres M. A. (2003 ). Reactive oxygen species produced by NADPH oxidase regulate plant cell growth. Nature 422 442 –446 . 10.1038/nature01485 12660786 Fry S. C. Miller J. G. Dumville J. C. (2002 ). A proposed role for copper ions in cell wall loosening. Plant Soil 247 57 –67 . 10.1023/A:1021140022082 Furuya T. Matsuoka D. Nanmori T. (2013 ). Phosphorylation of Arabidopsis thaliana MEKK1 via Ca2+ signaling as a part of the cold stress response. J. Plant Res. 126 833 –840 . 10.1007/s10265-013-0576-0 23857079 Furuya T. Matsuoka D. Nanmori T. (2014 ). Membrane rigidification functions upstream of the MEKK1-MKK2-MPK4 cascade during cold acclimation in Arabidopsis thaliana. FEBS Lett. 588 2025 –2030 . 10.1016/j.febslet.2014.04.032 24801600 Gobert A. Park G. Amtmann A. Sanders D. Maathuis F. J. M. (2006 ). Arabidopsis thaliana cyclic nucleotide-gated channel 3 forms a non-selective ion transporter involved in germination and cation transport. J. Expt. Bot. 57 791 –800 . 10.1093/jxb/erj064 Goldbach H. E. Yu Q. Wingender R. Schulz M. Wimmer M. Findeklee P. (2001 ). Rapid response reactions of roots to boron deprivation. J. Plant Nutr. Soil Sci. 164 173 –181 . 10.1002/1522-2624(200104)164:2<173::AID-JPLN173>3.3.CO;2-6 González-Fontes A. Herrera-Rodriguez M. B. Martin-Rejano E. M. Navarro-Gochicoa M. T. Rexach J. Camacho-Cristobal J. J. (2016 ). Root responses to boron deficiency mediated by ethylene. Front. Plant Sci. 6 :1103 10.3389/fpls.2015.01103 Gonzali S. Loreti E. Cardarelli F. Novi G. Parlanti S. Pucciariello C. (2015 ). Universal stress protein HRU1 mediates ROS homeostasis under anoxia. Nat. Plants 1 15151 10.1038/NPLANTS.2015.151 Guerra F. Duplessis S. Kohler A. Martin F. Tapia J. Lebed P. (2009 ). Gene expression analysis of Populus deltoides roots subjected to copper stress. Environ. Exp. Bot. 67 335 –344 . 10.1016/j.envexpbot.2009.08.004 Guo J. Zeng W. Chen Q. Lee C. Chen L. Yang Y. (2016 ). Structure of the voltage-gated two-pore channel TPC1 from Arabidopsis thaliana. Nature 531 196 –201 . 10.1038/nature16446 26689363 Guo K.-M. Babourina O. Christopher D. A. Borsics T. Rengel Z. (2008 ). The cyclic nucleotide-gated channel, AtCNGC10, influences salt tolerance in Arabidopsis. Physiol. Plant. 134 499 –507 . 10.1111/j.1399-3054.2008.01157.x 18823330 Haswell E. S. Peyronnet R. Barbier-Brygoo H. Meyerowitz E. M. Frachisse J. M. (2008 ). Two MscS homologs provide mechanosensitive channel activities in the Arabidopsis root. Curr. Biol. 18 730 –734 . 10.1016/j.cub.2008.04.039 18485707 Ho C.-H. Lin S.-H. Hu H.-C. Tsay Y.-F. (2009 ). CHL1 functions as a nitrate sensor in plants. Cell 138 1184 –1194 . 10.1016/j.cell.2009.07.004 19766570 Hossain M. A. Henriquez-Valencia C. Gómez-Páez M. Medina J. Orellana A. Vicente-Carbajosa J. (2016 ). Identification of novel components of the unfolded protein response in Arabidopsis. Front. Plant Sci. 7 :650 10.3389/fpls.2016.00650 Hou C. C. Tian W. Kleist T. He K. Garcia V. Bai F. (2014 ). DUF221 proteins are a family of osmosensitive calcium-permeable cation channels conserved across eukaryotes. Cell Res. 24 632 –635 . 10.1038/cr.2014.14 24503647 Hou Q. Ufer G. Bartels D. (2016 ). Lipid signalling in plant responses to abiotic stress. Plant Cell Environ. 39 1029 –1048 . 10.1111/pce.12666 26510494 Hu H. C. Wang Y. Y. Tsay Y. F. (2009 ). AtCIPK8, a CBL-interacting protein kinase, regulates the low-affinity phase of the primary nitrate response. Plant J. 57 264 –278 . 10.1111/j.1365-313X.2008.03685.x 18798873 Hu Y. F. Zhou G. Na X. F. Yang L. Nan W. B. Liu X. (2013 ). Cadmium interferes with maintenance of auxin homeostasis in Arabidopsis seedlings. J. Plant Physiol. 170 965 –975 . 10.1016/j.jplph.2013.02.008 23683587 Hua B. G. Mercier R. W. Zielinski R. E. (2003 ). Functional interaction of calmodulin with a plant cyclic nucleotide gated cation channel. Plant Physiol. Biochem. 41 945 –954 . 10.1016/j.plaphy.2003.07.006 Huang T.-L. Huang H.-J. (2008 ). ROS and CDPK-like kinase-mediated activation of MAP kinase in rice roots exposed to lead. Chemosphere 71 1377 –1385 . 10.1016/j.chemosphere.2007.11.031 18164745 Huang W.-C. Huang D.-D. Chien P.-S. Yeh C.-M. Chen P.-Y. Chi W.-C. (2007 ). Protein tyrosine dephosphorylation during copper-induced cell death in rice roots. Chemosphere 69 55 –62 . 10.1016/j.chemosphere.2007.04.073 17585992 Hunt L. Otterhag L. Lee J. C. Lasheen T. Hunt J. Seki M. (2004 ). Gene-specific expression and calcium activation of Arabidopsis thaliana phospholipase C isoforms. New Phytol. 162 643 –654 . 10.1111/j.1469-8137.2004.01069.x Isner J. C. Nühse T. Maathuis F. J. M. (2012 ). The cyclic nucleotide cGMP is involved in plant hormone signalling and alters phosphorylation of Arabidopsis thaliana root proteins. J. Expt. Bot. 63 3199 –3205 . 10.1093/jxb/ers045 Jakubowska D. Janicka-Russak M. Kabala K. Migocka M. Reda M. (2015 ). Modification of plasma membrane NADPH oxidase activity in cucumber seedling roots in response to cadmium stress. Plant Sci. 234 50 –59 . 10.1016/j.plantsci.2015.02.005 25804809 Jeon J. Kim J. (2013 ). Cold stress signaling networks in Arabidopsis. J. Plant Biol. 56 69 –76 . 10.1007/s12374-013-0903-y Jiang H. B. Wang S. F. Yang F. Zhang Z. H. Qiu H. C. Yi Y. (2015 ). Plant growth, nitrate content and Ca signaling in wheat (Triticum aestivum L.) roots under different nitrate supply. Plant Sci. J. 33 362 –368 . Jiao Y. Sun L. Song Y. Wang L. Liu L. Zhang L. (2013 ). AtrbohD and AtrbohF positively regulate abscisic acid-inhibited primary root growth by affecting Ca2+ signalling and auxin response of roots in Arabidopsis. J. Expt. Bot. 64 4183 –4192 . 10.1093/jxb/ert228 Jin Y. K. Jing W. Zhang Q. Zhang W. H. (2015 ). Cyclic nucleotide gated channel 10 negatively regulates salt tolerance by mediating Na+ transport in Arabidopsis. J. Plant Res. 128 211 –220 . 10.1007/s10265-014-0679-2 25416933 Kamano S. Kume S. Iida K. Lei K.-J. Nakano M. Nakayama Y. (2015 ). Transmembrane topologies of Ca2+-permeable mechanosensitive channels MCA1 and MCA2 in Arabidopsis thaliana. J. Biol. Chem. 290 30901 –30909 . 10.1074/jbc.M115.692574 26555262 Kanehara K. Yu C.-Y. Cho Y. Cheong W.-F. Torta F. Shui G. (2015 ). Arabidopsis AtPLC2 is a primary phosphoinositide-specific phospholipase C in phopshoinositide metabolism and the endoplasmic reticulum stress response. PLOS Genet. 11 :e1005511 10.1371/journal.pgen.1005511 Kasajima I. Ide Y. Hirai M. Y. Fujiwara T. (2010 ). WRKY6 is involved in the response to boron deficiency in Arabidopsis thaliana. Physiol. Plant. 139 80 –92 . 10.1111/j.1399-3054.2010.01349.x 20059736 Kawarazaki T. Kimura S. Iizuka A. Hanamata S. Nibori H. Michikawa M. (2013 ). A low temperature-inducible protein AtSRC2 enhances the ROS-producing activity of NADPH oxidase AtRbohF. Biochim. Biophys. Acta Mol. Cell Res. 1833 2775 –2780 . 10.1016/j.bbamcr.2013.06.024 Kiegle E. Moore C. A. Haseloff J. Tester M. A. Knight M. R. (2000 ). Cell-type-specific calcium responses to drought, salt and cold in the Arabidopsis root. Plant J. 23 267 –278 . 10.1046/j.1365-313x.2000.00786.x 10929120 Kim K. N. Cheong Y. H. Grant J. J. Pandey G. K. Luan S. (2003 ). CIPK3, a calcium sensor-associated protein kinase that regulates abscisic acid and cold signal transduction in Arabidopsis. Plant Cell 15 411 –423 . 10.1105/tpc.006858 12566581 Kim Y. Wang M. Bai Y. Zeng Z. Guo F. Han N. (2014 ). Bcl-2 suppresses activation of VPEs by inhibiting cytosolic Ca2+ level with elevated K+ efflux in NaCl-induced PCD in rice. Plant Physiol. Biochem. 80 168 –175 . 10.1016/j.plaphy.2014.04.002 24787501 Kintzer A. F. Stroud R. M. (2016 ). Structure, inhibition and regulation of two-pore channel TPC1 from Arabidopsis thaliana. Nature 531 258 –264 . 10.1038/nature17194 26961658 Knight H. Knight M. R. (2000 ). Imaging spatial and cellular characteristics of low temperature calcium signature after cold acclimation in Arabidopsis. J. Expt. Bot. 51 1679 –1686 . 10.1093/jexbot/51.351.1679 Knight H. Trewavas A. J. Knight M. R. (1996 ). Cold calcium signaling in Arabidopsis involves two cellular pools and a change in calcium signature after acclimation. Plant Cell 8 489 –503 . 10.1105/tpc.8.3.489 8721751 Knight H. Trewavas A. J. Knight M. R. (1997 ). Calcium signalling in Arabidopsis thaliana responding to drought and salinity. Plant J. 12 1067 –1078 . 10.1046/j.1365-313X.1997.12051067.x 9418048 Knight M. R. Knight H. (2012 ). Low-temperature perception leading to gene expression and cold tolerance in higher plants. New Phytol. 195 737 –751 . 10.1111/j.1469-8137.2012.04239.x 22816520 Kobayashi B. Matoh T. Azuma J. (1996 ). Two chains of rhamnogalacturonan II are cross-linked by borate-diol ester bonds in higher plant cell walls. Plant Physiol. 110 1017 –1020 .12226238 Kolbert Z. Petö A. Lehotai N. Feigl G. Erdei L. (2012 ). Long-term copper (Cu2+) exposure impacts on auxin, nitric oxide (NO) metabolism and morphology of Arabidopsis thaliana L. Plant Growth Regul. 68 151 –159 . 10.1007/s10725-012-9701-7 Koshiba T. Ishihara A. Matoh T. (2010 ). Boron nutrition of cultured Tobacco BY-2 Cells. VI. Calcium is involved in early responses to boron deprivation. Plant Cell Physiol. 51 323 –327 . 10.1093/pcp/pcp179 20008940 Kugler A. Koehler B. Palme K. Wolff P. Dietrich P. (2009 ). Salt-dependent regulation of a CNG channel subfamily in Arabidopsis. BMC Plant Biol. 9 :140 10.1186/1471-2229-9-140 Kurusu T. Nishikawa D. Yamazaki Y. Gotoh M. Nakano M. Hamada H. (2012 ). Plasma membrane protein OsMCA1 is involved in regulation of hypo-osmotic shock-induced Ca2+ influx and modulates generation of reactive oxygen species in cultured rice cells. BMC Plant Biol. 12 :11 10.1186/1471-2229-12-11 Laohavisit A. Davies J. M. (2011 ). Annexins. New Phytol. 189 40 –53 . 10.1111/j.1469-8137.2010.03533.x 21083562 Laohavisit A. Richards S. L. Shabala L. Chen C. Colaço R. Swarbreck S. M. (2013 ). Salinity-induced calcium signaling and root adaptation in Arabidopsis require the regulatory protein annexin1. Plant Physiol. 163 253 –262 . 10.1104/pp.113.217810 23886625 Laohavisit A. Shang Z.-L. Rubio L. Cuin T. A. Véry A. A. Wang A. H. (2012 ). Arabidopsis annexin 1 mediates the radical-activated plasma membrane Ca2+ and K+-permeable conductance in root cells. Plant Cell 24 1522 –1533 . 10.1105/tpc.112.097881 22523205 Latz A. Mehlmer N. Zapf S. Mueller T. D. Wurzinger B. Pfister B. (2013 ). Salt stress triggers phosphorylation of the Arabidopsis vacuolar K+ channel TPK1 by calcium-dependent protein kinases (CDPKs). Mol. Plant 6 1274 –1289 . 10.1093/mp/sss158 23253603 Legue V. Blancaflor E. Wymer C. Perbal G. Fahtin D. Gilroy S. (1997 ). Cytoplasmic free calcium in Arabidopsis roots changes in response to touch but not gravity. Plant Physiol. 114 789 –800 . 10.1104/pp.114.3.789 9232870 Lequeux H. Hermans C. Lutts S. Verbruggen N. (2010 ). Response to copper excess in Arabidopsis thaliana: impact on the root system architecture, hormone distribution, lignin accumulation and mineral profile. Plant Physiol. Biochem. 48 673 –682 . 10.1016/j.plaphy.2010.05.005 20542443 Léran S. Edel K. H. Pervent M. Hashimoto K. Corratge-Faillie C. Offenborn J. N. (2015 ). Nitrate sensing and uptake in Arabidopsis are enhanced by ABI2, a phosphatase inactivated by the stress hormone abscisic acid. Sci. Signal. 8 ra43. 10.1126/scisignal.aaa4829 Li G. Boudsocq M. Hem S. Vialaret J. Rossignol M. Maurel C. (2015 ). The calcium-dependent protein kinase CPK7 acts on root hydraulic conductivity. Plant Cell Environ. 38 1312 –1320 . 10.1111/pce.12478 25366820 Li J. Long Y. Qi G. N. Xu Z. J. Wu W. H. Wang Y. (2014 ). The OsAKT1 channel is critical for K+ uptake in rice roots and is modulated by the rice CBL1-CIPK23 complex. Plant Cell 26 3387 –3402 . 10.1105/tpc.114.123455 25096783 Li L. Kim B.-G. Cheong Y. H. Pandey G. K. Luan S. (2006 ). A Ca2+ signalling pathway regulates a K+ channel for low K+ response in Arabidopsis. Proc. Natl. Acad. Sci. U.S.A. 103 12625 –12630 . 10.1073/pnas.0605129103 16895985 Li P. Zhao C. Z. Zhang Y. Q. Wang X. M. Wang X. Y. Wang J. F. (2016 ). Calcium alleviates cadmium-induced inhibition on root growth by maintaining auxin homeostasis in Arabidopsis seedlings. Protoplasma 253 185 –200 . 10.1007/s00709-015-0810-9 25837011 Li P. H. Zhang G. Y. Gonzales N. Guo Y. Q. Hu H. H. Park S. H. (2016 ). Ca2+-regulated and diurnal rhythm-regulated Na+/Ca2+ exchanger AtNCL affects flowering time and auxin signalling in Arabidopsis. Plant Cell Environ. 39 377 –392 . 10.1111/pce.12620 26296956 Li S. Yu J. Zhu M. Zhao F. Luan S. (2012 ). Cadmium impairs ion homeostasis by altering K+ and Ca2+ channel activities in rice root hairs. Plant Cell Environ. 35 1998 –2013 . 10.1111/j.1365-3040.2012.02532.x 22563739 Lin C. Y. Trinh N. N. Fu S. F. Hsiung Y. C. Chia L. C. Lin C. W. (2013 ). Comparison of early transcriptome responses to copper and cadmium in rice roots. Plant Mol. Biol. 81 507 –522 . 10.1007/s11103-013-0020-9 23400832 Liu J. P. Ishitani M. Halfter U. Kim C. S. Zhu J. K. (2000 ). The Arabidopsis thaliana SOS2 gene encodes a protein kinase that is required for salt tolerance. Proc. Natl. Acad. Sci. U.S.A. 97 3730 –3734 . 10.1073/pnas.97.7.3730 10725382 Liu J.-X. Srivastava R. Che P. Howell S. H. (2007 ). Salt stress responses in Arabidopsis utilize a signal transduction pathway related to endoplasmic reticulum stress signaling. Plant J. 51 897 –909 . 10.1111/j.1365-313X.2007.03195.x 17662035 Liu L. Li Q. Yin B. Zhang H. Lin B. Wu Y. (2011 ). The endoplasmic reticulum-associated degradation is necessary for plant salt tolerance. Cell Res. 21 957 –969 . 10.1038/cr.2010.181 21187857 Liu Q. Zheng L. He F. Zhao F.-J. Shen Z. Zheng L. (2015 ). Transcriptional and physiological analyses identify a regulatory role for hydrogen peroxide in the lignin biosynthesis of copper-stressed rice roots. Plant Soil 387 323 –336 . 10.1007/s11104-014-2290-7 Lokdarshi A. Conner W. C. McClintock C. Li T. Roberts D. M. (2016 ). Arabidopsis CML38, a calcium sensor that localizes to ribonucleoprotein complexes under hypoxia stress. Plant Physiol. 170 1046 –1059 . 10.1104/pp.15.01407 26634999 Lu G. Wang X. Liu J. Yu K. Gao Y. Liu H. (2014 ). Application of T-DNA activation tagging to identify glutamate receptor-like genes that enhance drought tolerance in plants. Plant Cell Rep. 33 617 –631 . 10.1007/s00299-014-1586-7 24682459 Lyalin O. O. Ktitorova I. N. (1969 ). Resting potential of the root hair of Trianea bogotensis. Sov. Plant Phys. 16 214 –221 . Ma L. Zhang H. Sun L. Jiao Y. Zhang G. Miao C. (2012 ). NADPH oxidase AtrbohD and AtrbohF function in ROS-dependent regulation of Na+/K+ homeostasis in Arabidopsis under salt stress. J. Expt. Bot. 63 305 –317 . 10.1093/jxb/err280 Ma Q. Tang R. J. Zheng X. J. Wang S. M. Luan S. (2015 ). The calcium sensor CBL7 modulates plant responses to low nitrate in Arabidopsis. Biochem. Biophys. Res. Commun. 468 59 –65 . 10.1016/j.bbrc.2015.10.164 26549233 Ma Y. Dai X. Y. Xu Y. Y. Luo W. Zheng X. M. Zeng D. L. (2015 ). COLD1 confers chilling tolerance in rice. Cell 160 1209 –1221 . 10.1016/j.cell.2015.01.046 25728666 Maathuis F. J. M. (2014 ). Sodium in plants: perception, signalling and regulation of sodium fluxes. J. Expt. Bot. 65 849 –858 . 10.1093/jxb/ert326 Mazars C. Thion L. Thuleau P. Graziana A. Knight M. R. Moreau M. (1997 ). Organization of cytoskeleton controls the changes in cytosolic calcium of cold-shocked Nicotiana plumbaginifolia protoplasts. Cell Calcium 22 413 –420 . 10.1016/S0143-4160(97)90025-7 9448947 McAinsh M. R. Pittman J. K. (2009 ). Shaping the calcium signature. New Phytol. 181 275 –294 . 10.1111/j.1469-8137.2008.02682.x 19121028 McLoughlin F. Galvan-Ampudia C. S. Julkowska M. M. Caarls L. van der Does D. Lauriere C. (2012 ). The Snf1-related protein kinases SnRK2.4 and SnRK2.10 are involved in maintenance of root system architecture during salt stress. Plant J. 72 436 –449 . 10.1111/j.1365-313X.2012.05089.x 22738204 Mehlmer N. Wurzinger B. Stael S. Hofmann-Rodrigues D. Csaszar E. Pfister B. (2010 ). The Ca2+-dependent protein kinase CPK3 is required for MAPK-independent salt-stress acclimation in Arabidopsis. Plant J. 63 484 –498 . 10.1111/j.1365-313X.2010.04257.x 20497378 Mei H. Cheng N. H. Zhao J. Park S. Escareno R. A. Pittman J. K. (2009 ). Root development under metal stress in Arabidopsis thaliana requires the H+/cation antiporter CAX4. New Phytol. 183 95 –105 . 10.1111/j.1469-8137.2009.02831.x 19368667 Meringer M. V. Villasuso A. L. Peppino M. M. Usorach J. Pasquare S. J. Giusto N. M. (2016 ). Saline and osmotic stresses stimulate PLD/diacylglycerol kinase activities and increase the level of phosphatidic acid and proline in barley roots. Environ. Expt. Bot. 128 69 –78 . 10.1016/j.envexpbot.2016.03.011 Miedema H. Demidchik V. Véry A.-A. Bothwell J. H. F. Brownlee C. Davies J. M. (2008 ). Two voltage-dependent calcium channels co-exist in the apical plasma membrane of Arabidopsis thaliana root hairs. New Phytol. 179 378 –385 . 10.1111/j.1469-8137.2008.02465.x 19086288 Miller G. Schlauch K. Tam R. Cortes D. Torres M. A. Shulaev V. (2009 ). The plant NADPH oxidase RBOHD mediates rapid systemic signaling in response to diverse stimuli. Sci. Signal. 2 ra45. 10.1126/scisignal.2000448 Minorsky P. V. Spanswick R. M. (1989 ). Electrophysiological evidence for a role for calcium in temperature sensing by roots of cucumber seedlings. Plant Cell Env. 12 137 –143 . 10.1111/j.1365-3040.1989.tb01925.x Monshausen G. B. Bibikova T. N. Weisenseel M. H. Gilroy S. (2009 ). Ca2+ regulates reactive oxygen species production and pH during mechanosensing in Arabidopsis roots. Plant Cell 21 2341 –2356 . 10.1105/tpc.109.068395 19654264 Moore C. A. Bowen H. C. Scrase-Field S. Knight M. R. White P. J. (2002 ). The deposition of suberin lamellae determines the magnitude of cytosolic Ca2+ elevations in root endodermal cells subjected to cooling. Plant J. 30 457 –465 . 10.1046/j.1365-313X.2002.01306.x 12028575 Mortimer J. C. Laohavisit A. Miedema H. Davies J. M. (2008 ). Voltage, reactive oxygen species and the influx of calcium. Plant Sig. Behav. 3 698 –699 . 10.4161/psb.3.9.6405 Mühling K. H. Wimmer M. Goldbach H. E. (1998 ). Apoplastic and membrane associated Ca2+ in leaves and roots as affected by boron deficiency. Physiol. Plant. 102 179 –184 . 10.1034/j.1399-3054.1998.1020204.x Nakagawa Y. Katagiri T. Shinozaki K. Qi Z. Tatsumi H. Furuichi T. (2007 ). Arabidopsis plasma membrane protein crucial for Ca2+ influx and touch sensing in roots. Proc. Natl. Acad. Sci. U.S.A. 104 3639 –3644 . 10.1073/pnas.0607703104 17360695 Nieves-Cordones M. Miller A. J. Aleman F. Martinez V. Rubio F. (2008 ). A putative role for the plasma membrane potential in the control of the expression of the gene encoding the tomato high-affinity potassium transporter HAK5. Plant Mol. Biol. 68 521 –532 . 10.1007/s11103-008-9388-3 18726559 Oiwa Y. Kitayama K. Matoh T. (2013 ). Boron deprivation immediately causes cell death in growing roots of Arabidopsis thaliana (L.) Heynh. Soil Sci. Plant Nutr. 59 621 –627 . 10.1080/00380768.2013.813382 Ordenes V. R. Moreno I. Maturana D. Norambuena L. Trewavas A. J. Orellana A. (2012 ). In vivo analysis of the calcium signature in the plant Golgi apparatus reveals unique dynamics. Cell Calcium 52 397 –404 . 10.1016/j.ceca.2012.06.008 22831913 Orvar B. L. Sangwan V. Omann F. Dhindsa R. S. (2000 ). Early steps in cold sensing by plant cells: the role of actin cytoskeleton and membrane fluidity. Plant J. 23 785 –794 . 10.1046/j.1365-313x.2000.00845.x 10998189 Pandey N. Ranjan A. Pant P. Tripathi R. K. Ateek F. Pandey H. P. (2013 ). CAMTA 1 regulates drought responses in Arabidopsis thaliana. BMC Genomics 14 :216 10.1186/1471-2164-14-216 Peiter E. Maathuis F. J. M. Mills L. N. Knight H. Pelloux J. Hetherington A. M. (2005 ). The vacuolar Ca2+-activated channel TPC1 regulates germination and stomatal movement. Nature 434 404 –408 . 10.1038/nature03381 15772667 Perfus-Barbeoch L. Leonhardt N. Vavasseur A. Forestier C. (2002 ). Heavy metal toxicity: cadmium permeates through calcium channels and disturbs the plant water status. Plant J. 32 539 –548 . 10.1046/j.1365-313X.2002.01442.x 12445125 Petrov V. Hille J. Mueller-Roeber B. Gechev T. S. (2015 ). ROS-mediated abiotic stress-induced programmed cell death in plants. Front. Plant Sci. 6 :69 10.3389/fpls.2015.00069 Plieth C. Hansen U. P. Knight H. Knight M. R. (1999 ). Temperature sensing by plants: the primary characteristics of signal perception and calcium response. Plant J. 18 491 –497 . 10.1046/j.1365-313X.1999.00471.x 10417699 Pottosin I. Maria Velarde-Buendia A. Bose J. Fuglsang A. T. Shabala S. (2014 ). Polyamines cause plasma membrane depolarization, activate Ca2+-, and modulate H+-ATPase pump activity in pea roots. J. Expt. Bot. 65 2463 –2472 . 10.1093/jxb/eru133 Probst A. V. Scheid O. M. (2015 ). Stress-induced structural changes in plant chromatin. Curr. Opin. Plant Biol. 27 8 –16 . 10.1016/j.pbi.2015.05.011 26042538 Qin W. S. Pappan K. Wang X. M. (1997 ). Molecular heterogeneity of phospholipase D (PLD) – Cloning of PLD gamma and regulation of plant PLD gamma, -beta, and -alpha by polyphosphoinositides and calcium. J. Biol. Chem. 272 28267 –28273 . 10.1074/jbc.272.45.28267 9353280 Qiu Q. S. Guo Y. Quintero F. J. Pardo J. M. Schumaker K. S. Zhu J. K. (2004 ). Regulation of vacuolar Na+/H+ exchange in Arabidopsis thaliana by the salt-overly-sensitive (SOS) pathway. J. Biol. Chem. 279 207 –215 . 10.1074/jbc.M307982200 14570921 Quiles-Pando C. Rexach J. Navarro-Gochicoa M. T. Camacho-Cristóbal J. J. Herrera-Rodríguez M. B. González-Fontes A. (2013 ). Boron deficiency increases the level of cytosolic Ca2+ and expression of Ca2+-related genes in Arabidopsis thaliana roots. Plant Physiol. Biochem. 65 55 –60 . 10.1016/j.plaphy.2013.01.004 23416496 Quintero F. J. Martinez-Atienza J. Villalta I. Jiang X. Kim W. Y. Ali Z. (2011 ). Activation of the plasma membrane Na/H antiporter Salt-Overly-Sensitive 1 (SOS1) by phosphorylation of an auto-inhibitory C-terminal domain. Proc. Natl. Acad. Sci. U.S.A. 108 2611 –2616 . 10.1073/pnas.1018921108 21262798 Ragel P. Rodenas R. Garcia-Martin E. Andres Z. Villalta I. Nieves-Cordones M. (2015 ). SOS the CBL-interacting protein kinase CIPK23 regulates HAK5-mediated high-affinity K+ uptake in Arabidopsis roots. Plant Physiol. 169 2863 –2873 . 10.1104/pp.15.01401 26474642 Richards S. L. Wilkins K. A. Swarbreck S. M. Anderson A. Habib N. McAinsh M. (2015 ). The hydroxyl radical; from seed to seed. J. Expt. Bot. 66 37 –46 . 10.1093/jxb/eru398 Richter G. L. Monshausen G. B. Krol A. Gilroy S. (2009 ). Mechanical stimuli modulate lateral root organogenesis. Plant Physiol. 151 1855 –1866 . 10.1104/pp.109.142448 19794120 Riveras E. Alvarez J. M. Vidal E. A. Oses C. Vega A. Gutierrez R. A. (2015 ). The calcium ion is a second messenger in the nitrate signaling pathway of Arabidopsis. Plant Physiol. 169 1397 –1404 . 10.1104/pp.15.00961 26304850 Rodrigo-Moreno A. Andres-Colas N. Poschenrieder C. Gunse B. Penarrubia L. Shabala S. (2013 ). Calcium- and potassium-permeable plasma membrane transporters are activated by copper in Arabidopsis root tips: linking copper transport with cytosolic hydroxyl radical production. Plant Cell Environ. 36 844 –855 . 10.1111/pce.12020 23046313 Rodriguez-Hernandez M. C. Bonifas I. Alfaro-De la Torre M. C. Flores-Flores J. I. Bañuelos-Hernández B. Patiño-Rodríguez O. (2015 ). Increased accumulation of cadmium and lead under Ca and Fe deficiency in Typha latifolia: a case study of two pore channel (TPC1) gene. Environ. Exp. Bot. 115 38 –48 . 10.1016/j.envexpbot.2015.02.009 Roy S. J. Gilliham M. Berger B. Essah P. A. Cheffings C. Miller A. J. (2008 ). Investigating glutamate receptor-like gene co-expression in Arabidopsis thaliana. Plant Cell Environ. 31 861 –871 . 10.1111/j.1365-3040.2008.01801.x 18284583 Ruberti C. Kim S. J. Stefano G. Brandizzi F. (2015 ). Unfolded protein responses in plants: one master, many questions. Curr. Opin. Plant Biol. 27 59 –66 . 10.1016/j.pbi.2015.05.016 26149756 Ruelland E. Kravets V. Derevyanchuk M. Martinec J. Zachowski A. Pokotylo I. (2015 ). Role of phospholipid signalling in plant environmental responses. Environ. Exp. Bot. 114 129143 10.1016/j.envexpbot.2014.08.009 Sagor G. H. M. Chawla P. Kim D. W. Berberich T. Kojima S. Niitsu M. (2015 ). The polyamine spermine induces the unfolded protein response via the MAPK cascade in Arabidopsis. Front. Plant Sci. 6 :687 10.3389/fpls.2015.00687 Sani E. Herzyk P. Perrella G. Colot V. Amtmann A. (2013 ). Hyperosmotic priming of Arabidopsis seedlings establishes a long-term somatic memory accompanied by specific changes of the epigenome. Genome Biol. 14 R59. 10.1186/gb-2013-14-6-r59 Sarwat M. Naqvi A. R. (2013 ). Heterologous expression of rice calnexin (OsCNX) confers drought tolerance. Mol. Biol. Rep. 40 5451 –5464 . 10.1007/s11033-013-2643-y 23681551 Shabala S. Pottosin I. (2014 ). Regulation of potassium transport in plants under hostile conditions: implications for abiotic and biotic stress tolerance. Physiol. Plant 151 257 –279 . 10.1111/ppl.12165 24506225 Shabala S. Shabala L. Barcelo J. Poschenrieder C. (2014 ). Membrane transporters mediating root signalling and adaptive responses to oxygen deprivation and soil flooding. Plant Cell Environ. 37 2216 –2233 . 10.1111/pce.12339 24689809 Shabala S. Wu H. Bose J. (2015 ). Salt stress sensing and early signalling in plant roots: current knowledge and hypothesis. Plant Sci. 241 109 –119 . 10.1016/j.plantsci.2015.10.003 26706063 Shi Y. T. Ding Y. L. Yang S. H. (2015 ). Cold signal transduction and its interplay with phytohormones during cold acclimation. Plant Cell Physiol. 56 7 –15 . 10.1093/pcp/pcu115 25189343 Shih H. W. Depew C. L. Miller N. D. Monshausen G. B. (2015 ). The cyclic nucleotide-gated channel CNGC14 regulates root gravitropism in Arabidopsis thaliana. Curr. Biol. 25 1 –7 . 10.1016/j.cub.2015.10.025 25532895 Shih H.-W. Miller N. D. Dai C. Spalding E. P. Monshausen G. B. (2014 ). The receptor-like kinase FERONIA is required for mechanical signal transduction in Arabidopsis seedlings. Curr. Biol. 24 1887 –1892 . 10.1016/j.cub.2014.06.064 25127214 Shin R. Schachtman D. P. (2004 ). Hydrogen peroxide mediates plant root cell response to nutrient deprivation. Proc. Natl. Acad. Sci. U.S.A. 101 8827 –8832 . 10.1073/pnas.0401707101 15173595 Shorrocks V. M. (1997 ). The occurrence and correction of boron deficiency. Plant Soil 193 121 –148 . 10.1023/A:1004216126069 Singh S. Parihar P. Singh R. Singh V. P. Prasad S. M. (2016 ). Heavy metal tolerance in plants: role of transcriptomics, proteomics, metabolomics, and ionomics. Front. Plant Sci. 6 :1143 10.3389/fpls.2015.01143 Stael S. Wurzinger B. Mair A. Mehlmer N. Vothknecht U. C. Teige M. (2012 ). Plant organellar calcium signalling: an emerging field. J. Expt. Bot. 63 1525 –1542 . 10.1093/jxb/err394 Straltsova D. Chykun P. Subramaniam S. Sosan A. Kolbanov D. Sokolik A. (2015 ). cation channels are involved in brassinosteroid signalling in higher plants. Steroids 97 98 –106 . 10.1016/j.steroids.2014.10.008 25449770 Sulaiman Y. Knight M. R. Kataky R. (2012 ). Non-invasive monitoring of temperature stress in Arabidopsis thaliana roots, using ion amperometry. Anal. Meth. 4 1656 –1661 . 10.1039/c2ay05747f Sunkar R. Kaplan B. Bouche N. Arazi T. Dolev D. Talke I. N. (2000 ). Expression of a truncated tobacco NtCBP4 channel in transgenic plants and disruption of the homologous Arabidopsis CNGC1 gene confer Pb2+ tolerance. Plant J. 24 533 –542 . 10.1046/j.1365-313x.2000.00901.x 11115134 Swarbreck S. M. Colaço R. Davies J. M. (2013 ). Update on plant calcium-permeable channels. Plant Physiol. 163 1514 –1522 . 10.1104/pp.113.220855 Tai F. Yuan Z. Li S. P. Wang Q. Liu F. Wang W. (2016 ). ZmCIPK8, a CBL-interacting protein kinase, regulates maize response to drought stress. Plant Cell Tiss. Org. Cul. 124 459 –469 . 10.1007/s11240-015-0906-0 Teardo E. Carraretto L. De Bortoli S. Costa A. Behera S. Wagner R. (2015 ). Alternative splicing-mediated targeting of the Arabidopsis GLUTAMATE RECEPTOR3.5 to mitochondria affects organelle morphology. Plant Physiol. 167 216 –227 . 10.1104/pp.114.242602 25367859 Thion L. Mazars C. Nacry P. Bouchez D. Moreau M. Ranjeva R. (1998 ). Plasma membrane depolarization-activated calcium channels, stimulated by microtubule-depolymerizing drugs in wild-type Arabidopsis thaliana protoplasts, display constitutively large activities and a longer half-life in ton 2 mutant cells affected in the organization of cortical microtubules. Plant J. 13 603 –610 .9681002 Thoday-Kennedy E. L. Jacobs A. J. Roy S. J. (2015 ). The role of the CBL-CIPK calcium signalling network in regulating ion transport in response to abiotic stress. Plant Growth Regul. 76 3 –12 . 10.1007/s10725-015-0034-1 Toyota M. Gilroy S. (2013 ). Gravitropism and mechanical signalling in plants. Am. J. Bot. 100 111 –125 . 10.3732/ajb.1200408 23281392 Tracy F. E. Gilliham M. Dodd A. N. Webb A. A. R. Tester M. (2008 ). NaCl-induced changes in cytosolic free Ca2+ in Arabidopsis thaliana are heterogeneous and modified by external ionic composition. Plant Cell Environ. 31 1063 –1073 . 10.1111/j.1365-3040.2008.01817.x 18419736 Véry A. A. Davies J. M. (2000 ). Hyperpolarization-activated calcium channels at the tip of Arabidopsis root hairs. Proc. Natl. Acad. Sci. U.S.A. 97 9801 –9806 . 10.1073/pnas.160250397 10920194 Villiers F. Jourdain A. Bastien O. Leonhardt N. Fujioka S. Tichtincky G. (2012 ). Evidence for functional interaction between brassinosteroids and cadmium response in Arabidopsis thaliana. J. Expt. Bot. 63 1185 –1200 . 10.1093/jxb/err335 Virdi A. S. Singh S. Singh P. (2015 ). Abiotic stress responses in plants: roles of calmodulin-regulated proteins. Front. Plant Sci. 6 :809 10.3389/fpls.2015.00809 Virolainen E. Blokhina O. Fagerstedt K. (2002 ). Ca2+-induced high amplitude swelling and cytochrome c release from wheat (Triticum aestivum L.) mitochondria under anoxic stress. Ann. Bot. 90 509 –516 . 10.1093/aob/mcf221 12324275 Wagner S. Behera S. De Bortoli S. Logan D. C. Fuchs P. Carraretto L. (2015 ). The EF-Hand Ca2+-binding protein MICU choreographs mitochondrial Ca2+ dynamics in Arabidopsis. Plant Cell 27 3190 –3212 . 10.1105/tpc.15.00509 26530087 Wan S. Jiang L. (2016 ). Endoplasmic reticulum (ER) stress and the unfolded protein response (UPR) in plants. Protoplasma 273 753 –764 . 10.1007/s00709-015-0842-1 26060134 Wang F. F. Chen Z. H. Liu X. H. Colmer T. D. Zhou M. X. Shabala S. (2016 ). Tissue-specific root ion profiling reveals essential roles of the CAX and ACA calcium transport systems in response to hypoxia in Arabidopsis. J. Expt. Bot. 67 3747 –3762 . 10.1093/jxb/erw034 Wang P. Li Z. W. Wei J. S. Zhao Z. L. Sun D. Y. Cui S. J. (2012 ). A Na+/Ca2+ exchanger-like protein (AtNCL) involved in salt stress in Arabidopsis. J. Biol. Chem. 287 44062 –44070 . 10.1074/jbc.M112.351643 23148213 Weiland M. Mancuso S. Baluska F. (2016 ). Signalling via glutamate and GLRs in Arabidopsis thaliana. Funct. Plant Biol. 43 1 –25 . Whalley H. J. Knight M. R. (2013 ). Calcium signatures are decoded by plants to give specific gene responses. New Phytol. 197 690 –693 . 10.1111/nph.12087 23190495 Whalley H. J. Sargeant A. W. Steele J. F. Lacoere T. Lamb R. Saunders N. J. (2011 ). Transcriptomic analysis reveals calcium regulation of specific promoter motifs in Arabidopsis. Plant Cell 23 4079 –4095 . 10.1105/tpc.111.090480 22086087 White P. J. (2009 ). Depolarization-activated calcium channels shape the calcium signatures induced by low-temperature stress. New Phytol. 183 7 –8 . 10.1111/j.1469-8137.2009.02857.x Wigge P. A. (2013 ). Ambient temperature signalling in plants. Curr. Opn. Plant Biol. 16 661 –666 . 10.1016/j.pbi.2013.08.004 Wu J. Y. Jin C. Qu H. Y. Tao S. T. Xu G. H. Wu J. (2012 ). Low temperature inhibits pollen viability by alteration of actin cytoskeleton and regulation of pollen plasma membrane ion channels in Pyrus pyrifolia. Environ. Expt. Bot. 78 70 –75 . 10.1016/j.envexpbot.2011.12.021 Xu Q. T. Fan H. Y. Jiang Z. Zhou Z. Q. Yang L. Mei F. Z. (2013 ). Cell wall degradation and the dynamic changes of Ca2+ and related enzymes in the developing aerenchyma of wheat (Triticum aestivum L.) under waterlogging. Acta Biol. Hung. 64 328 –340 . 10.1556/ABiol.64.2013.3.6 24013894 Yamada N. Theerawitaya C. Ca-Um S. Kirdmanee C. Takabe T. (2014 ). Expression and functional analysis of putative vacuolar Ca2+-transporters (CAXs and ACAs) in roots of salt tolerant and sensitive rice cultivars. Protoplasma 251 1067 –1075 . 10.1007/s00709-014-0615-2 24482191 Yamauchi T. Watanabe K. Fukazawa A. Mori H. Abe F. Kawaguchi K. (2014 ). Ethylene and reactive oxygen species are involved in root aerenchyma formation and adaptation of wheat seedlings to oxygen-deficient conditions. J. Expt. Bot. 65 261 –273 . 10.1093/jxb/ert371 Yang C.-Y. Hong C.-P. (2015 ). The NADPH oxidase RbohD is involved in primary hypoxia signalling and modulates expression of hypoxia-inducible genes under hypoxic stress. Env. Expt. Bot. 115 63 –72 . 10.1016/j.envexpbot.2015.02.008 Yang T. B. Chaudhuri S. Yang L. H. Du L. Q. Poovaiah B. W. (2010 ). A calcium/calmodulin-regulated member of the receptor-like kinase family confers cold tolerance in plants. J. Biol. Chem. 285 7119 –7126 . 10.1074/jbc.M109.035659 20026608 Yeh C. M. Chien P. S. Huang H. J. (2007 ). Distinct signalling pathways for induction of MAP kinase activities by cadmium and copper in rice roots. J. Expt. Bot. 58 659 –671 . 10.1093/jxb/erl240 Yemelyanov V. V. Shishova M. F. Chirkova T. V. Lindberg S. M. (2011 ). Anoxia-induced elevation of cytosolic Ca2+ concentration depends on different Ca2+ sources in rice and wheat protoplasts. Planta 234 271 –280 . 10.1007/s00425-011-1396-x 21424536 Yu H. Q. Yong T. M. Li H. J. Liu Y. P. Zhou S. F. Fu F. L. (2015 ). Overexpression of a phospholipase D alpha gene from Ammopiptanthus nanus enhances salt tolerance of phospholipase D alpha 1-deficient Arabidopsis mutant. Planta 242 1495 –1509 . 10.1007/s00425-015-2390-5 26318308 Yu L. J. Nie J. N. Cao C. Y. Jin Y. K. Yan M. Wang F. Z. (2010 ). Phosphatidic acid mediates salt stress response by regulation of MPK6 in Arabidopsis thaliana. New Phytol. 188 762 –773 . 10.1111/j.1469-8137.2010.03422.x 20796215 Yu Y. Q. Assmann S. M. (2015 ). The heterotrimeric G-protein beta subunit, AGB1, plays multiple roles in the Arabidopsis salinity response. Plant Cell Environ. 38 2143 –2156 . 10.1111/pce.12542 25808946 Yuan F. Yang H. Xue Y. Kong D. Ye R. Li C. (2014 ). OSCA1 mediates osmotic-stress-evoked Ca2+ increases vital for osmosensing in Arabidopsis. Nature 514 367 –371 . 10.1038/nature13593 25162526 Yuan H.-M. Huang X. (2016 ). Inhibition of root meristem growth by cadmium involves nitric oxide-mediated repression of auxin accumulation and signalling in Arabidopsis. Plant Cell Environ. 39 120 –135 . 10.1111/pce.12597 26138870 Yuen C. C. Y. Christopher D. A. (2013 ). The group IV-A cyclic nucleotide-gated channels, CNGC19 and CNGC 20, localize to the vacuole membrane in Arabidopsis thaliana. AoB Plants 5 plt012. 10.1186/1471-2164-15-853 Zelinova V. Alemayehu A. Bocova B. Huttova J. Tamas L. (2015 ). Cadmium-induced reactive oxygen species generation, changes in morphogenic responses and activity of some enzymes in barley root tip are regulated by auxin. Biologia 70 356 –364 . 10.1515/biolog-2015-0035 Zeng F. Konnerup D. Shabala L. Zhou M. Colmer T. D. Zhang G. (2014 ). Linking oxygen availability with membrane potential maintenance and K+ retention of barley roots: implications for waterlogging stress tolerance. Plant Cell Environ. 37 2325 –2338 . 10.1111/pce.12422 25132404 Zeng Q. Ling Q. Fan L. Li Y. Hu F. Chen J. (2015 ). Transcriptome profiling of sugarcane roots in response to potassium stress. PLoS ONE 10 :e0126306 10.1371/journal.pone.0126306 Zepeda-Jazo I. Maria Velarde-Buendia A. Enriquez-Figueroa R. Bose J. Shabala S. Muniz-Murguia J. (2011 ). Polyamines interact with hydroxyl radicals in activating Ca2+ and K+ transport across the root epidermal plasma membranes. Plant Physiol. 157 2167 –2180 . 10.1016/j.plaphy.2012.09.002 21980172 Zhang L. Chen Z. Zhu C. (2012 ). Endogenous nitric oxide mediates alleviation of cadmium toxicity induced by calcium in rice seedlings. J. Environ. Sci. 24 940 –948 . 10.1016/S1001-0742(11)60978-9 Zhang X. Shen Z. D. Sun J. Yu Y. C. Deng S. R. Li Z. Y. (2015 ). NaCl-elicited, vacuolar Ca2+ release facilitates prolonged cytosolic Ca2+ signaling in the salt response of Populus euphratica cells. Cell Calcium 57 348 –365 . 10.1016/j.ceca.2015.03.001 25840638 Zhang Y. Wang Y. Taylor J. L. Jiang Z. Zhang S. Mei F. (2015 ). Aequorin-based luminescence imaging reveals differential calcium signalling responses to salt and reactive oxygen species in rice roots. J. Expt. Bot. 66 2535 –2545 . 10.1093/jxb/erv043 Zhao M. G. Tian Q. Y. Zhang W. H. (2007 ). Ethylene activates a plasma membrane Ca2+-permeable channel in tobacco suspension cells. New Phytol. 174 507 –515 . 10.1111/j.1469-8137.2007.02037.x 17447907 Zhao R. Sun H. M. Zhao N. Jing X. S. Shen X. Chen S. (2015 ). The Arabidopsis Ca2+-dependent protein kinase CPK27 is required for plant response to salt-stress. Gene 563 203 –214 . 10.1016/j.gene.2015.03.024 25791495 Zhao Y. Pan Z. Zhang Y. Qu X. L. Zhang Y. G. Yang Y. G. (2013 ). The Actin-Related Protein2/3 complex regulates mitochondrial-associated calcium signaling during salt stress in Arabidopsis. Plant Cell 25 4544 –4559 . 10.1105/tpc.113.117887 24280386 Zhu S.-Y. Yu X.-C. Wang X.-J. Zhao R. Li Y. Fan R.-C. (2007 ). Two calcium-dependent protein kinases, CPK4 and CPK11, regulate abscisic acid signal transduction in Arabidopsis. Plant Cell 19 3019 –3036 . 10.1105/tpc.107.050666 17921317 Zhu X. H. Feng Y. Liang G. M. Liu N. Zhu J. K. (2013 ). Aequorin-based luminescence imaging reveals stimulus- and tissue-specific Ca2+ dynamics in Arabidopsis plants. Mol. Plant 6 444 –455 . 10.1093/mp/sst013 23371933

PMC005xxxxxx/PMC5002412.txt

==== Front Front PhysiolFront PhysiolFront. Physiol.Frontiers in Physiology1664-042XFrontiers Media S.A. 10.3389/fphys.2016.00350PhysiologyOriginal ResearchComparison of High-Protein, Intermittent Fasting Low-Calorie Diet and Heart Healthy Diet for Vascular Health of the Obese Zuo Li 1†He Feng 23†Tinsley Grant M. 4Pannell Benjamin K. 1Ward Emery 3Arciero Paul J. 3*1Radiologic Sciences and Respiratory Therapy Division, School of Health and Rehabilitation Sciences, The Ohio State University College of Medicine, The Ohio State University Wexner Medical CenterColumbus, OH, USA2Department of Kinesiology, California State University, ChicoChico, CA, USA3Human Nutrition and Metabolism Laboratory, Health and Exercise Sciences Department, Skidmore CollegeSaratoga Springs, NY, USA4Department of Kinesiology and Sport Management, Texas Tech UniversityLubbock, TX, USAEdited by: Yih-Kuen Jan, University of Illinois at Urbana–Champaign, USA Reviewed by: Heikki Kainulainen, University of Jyväskylä, Finland; Andreas Bergdahl, Concordia University, Canada; Graziamaria Corbi, University of Molise, Italy; Chien-Liang Chen, I-Shou University, Taiwan *Correspondence: Paul J. Arciero parciero@skidmore.eduThis article was submitted to Clinical and Translational Physiology, a section of the journal Frontiers in Physiology †These authors have contributed equally to this work. 29 8 2016 2016 7 35009 5 2016 02 8 2016 Copyright © 2016 Zuo, He, Tinsley, Pannell, Ward and Arciero.2016Zuo, He, Tinsley, Pannell, Ward and ArcieroThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.Aim: It has been debated whether different diets are more or less effective in long-term weight loss success and cardiovascular disease prevention among men and women. To further explore these questions, the present study evaluated the combined effects of a high-protein, intermittent fasting, low-calorie diet plan compared with a heart healthy diet plan during weight loss, and weight loss maintenance on blood lipids and vascular compliance of obese individuals. Methods: The experiment involved 40 obese adults (men, n = 21; women, n = 19) and was divided into two phases: (a) 12-week high-protein, intermittent fasting, low-calorie weight loss diet comparing men and women (Phase 1) and (b) a 1-year weight maintenance phase comparing high-protein, intermittent fasting with a heart healthy diet (Phase 2). Body weight, body mass index (BMI), blood lipids, and arterial compliance outcomes were assessed at weeks 1 (baseline control), 12 (weight loss), and 64 (12 + 52 week; weight loss maintenance). Results: At the end of weight loss intervention, concomitant reductions in body weight, BMI and blood lipids were observed, as well as enhanced arterial compliance. No sex-specific differences in responses were observed. During phase 2, the high-protein, intermittent fasting group demonstrated a trend for less regain in BMI, low-density lipoprotein (LDL), and aortic pulse wave velocity than the heart healthy group. Conclusion: Our results suggest that a high-protein, intermittent fasting and low-calorie diet is associated with similar reductions in BMI and blood lipids in obese men and women. This diet also demonstrated an advantage in minimizing weight regain as well as enhancing arterial compliance as compared to a heart healthy diet after 1 year. arterial compliancecholesterollipidsweight lossweight relapse ==== Body Introduction In the United States, the prevalence of overweight and obesity is over 60% in adults (Flegal et al., 2010), which subsequently contributes to metabolic and cardiovascular diseases (CVD). Obesity is often understood to coexist with numerous cardiovascular risk factors (Wilson et al., 2002), and is associated with multiple inflammatory markers and cytokines that potentially contribute to the adverse cardiovascular outcomes in individuals with obesity (Van Gaal et al., 2006). Men suffer from a higher prevalence of these disorders than women, indicating that sex-based differences may play a role in cardiometabolic health (Lönnqvist et al., 1997; Blaak, 2001). Many short-term weight loss (WL) interventions are effective at improving cardiovascular/metabolic disease risk factors [e.g., body fat, total cholesterol (TC), and triglycerides (TG; Kelishadi et al., 2008; Clifton et al., 2009; Klempel et al., 2012; Mozaffarian et al., 2015; Pedersen et al., 2015)]. One popular form of WL diet is intermittent fasting (IF), which utilizes repeated short-term fasts to reduce energy intake, promote WL, and improve the lipid profile (Tinsley and La Bounty, 2015). IF has been shown to be as effective as continuous energy restriction in terms of WL, reducing TG, low-density lipoprotein (LDL) cholesterol, and blood pressure (BP; Harvie et al., 2011; Varady, 2011). However, the impact of IF on vascular compliance, an indicator of cardiovascular disease risk (Vlachopoulos et al., 2010), has not been examined. There is also limited information concerning whether men and women differ in their responses to short-term IF. Additionally, IF diets have not been examined in combination with high protein intake aimed at promoting cardiovascular health. Thus, the first aim of this study is to examine if there are sex differences in terms of lipid profile, and arterial compliance following a short-term high-protein, intermittent fasting, low-calorie (HP-IF-LC) diet. WL induced by nutritional interventions, such as a HP-IF and a traditional heart healthy (HH) diet, has demonstrated beneficial effects on minimizing cardiovascular risk factors (Arciero et al., 2006; Camhi et al., 2010; Song et al., 2016). HH diets exert a protective effect on the heart, as evidenced by reduced rates of myocardial infarction (Hansen-Krone et al., 2012). On the other hand, HP diets contribute to WL and weight loss maintenance (WL-M) by increasing the metabolic rate (Leidy et al., 2015). HP diets also induce extra energy expenditure via protein and urea synthesis as well as gluconeogenesis (Westerterp-Plantenga et al., 2009). Although multiple studies have demonstrated the short-term benefits of HP and IF diets, there is little research focusing on the combined or long-term effects of HP-IF diet on blood lipids, arterial compliance and CVD risk reduction in adults with obesity during and after a short-term WL intervention (Mattson and Wan, 2005; Varady, 2011). Clinical trials have demonstrated the independent effects of either HP, IF, or LC diets on improving cardiovascular outcomes in individuals with obesity (Katare et al., 2009; Harvie et al., 2011; Damsgaard et al., 2013; de Luis et al., 2016). However, few studies emphasize the combined effects of each of these dietary strategies on long-term WL-M and cardiovascular health. Higher protein intake during IF with low energy intake may be beneficial for WL and WL-M due to improved preservation of fat-free mass and resting metabolic rate. Therefore, the second major aim of the current study was to compare the effects of two different WL-M strategies (HP-IF vs. HH) on blood lipids, and arterial compliance in adults with obesity following an initial short-term HP-IF-LC WL diet. Materials and methods Participants Eligible volunteers were randomly recruited from the Saratoga Springs, NY area. Participants were eligible for inclusion in the study if they were healthy non-smokers, but were overweight or obese. A comprehensive medical examination/history assessment was performed by their physicians. Individuals were excluded from participation in the study if they had any previous cardiovascular or metabolic disease, or were receiving hormone therapy which could influence weight status, central adiposity and CVD risk factors measured in this study. Additionally, only individuals who were either sedentary or lightly active (<30 min, 2 days/week of organized physical activity), weight stable (± 2 kg during the past 6 months), middle-aged, and non-alcoholic, based on self-report, were eligible for inclusion in the study. Every participant provided written informed consent in accordance with the Skidmore College Human Subjects Review Board prior to participation, and the study was approved by the Human Subjects Institutional Review Board of Skidmore College (IRB #: 1307-347). All experimental procedures were performed in adherence with New York State regulations and the Federal Wide Assurance, which are consistent with the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, and in agreement with the Helsinki Declaration (revised in 1983). This trial was registered at clinicaltrials.gov as NCT02525419. Experimental design and study timeline Forty subjects with obesity were enrolled as a single cohort in this 64-week dietary regimen, splitting into two consecutive intervention phases: (a) 12-week WL (Phase 1) with HP-IF-LC diet (1-week baseline control, 10-week WL, 1-week post-testing) comparing men and women, and (b) a 1-year WL-M (Phase 2) comparing two diets (HP-IF vs. HH). All laboratory testing procedures were completed following baseline control (week 0), WL (week 12), and WL-M (week 64), as shown in Figure 1. During baseline control, the height, and body weight were recorded for each participant. In order to maintain stable weight, subjects were asked to maintain routine eating patterns and record dietary food logs for 2 days during baseline control period. In order to verify sedentary/low activity levels, all participants wore an Actical accelerometer (Bio-Lynx Scientific Equipment Inc., Montreal Quebec, Canada) around their waist for 2 days during weeks 0, 5, and 10. At weeks 0, 12, and 64, body weight was assessed between 6:00 and 9:00 a.m., after an overnight fast. Subjects then rested in a supine position for 15-min in a quiet and dimly lit room before a fasted blood draw was performed for assessment of blood lipids and measures of arterial compliance were obtained (see Laboratory Testing Procedures). Following control baseline testing, participants were provided with detailed instructions on their WL dietary guidelines (see Dietary Intervention) and scheduled their weekly meeting with a licensed registered dietitian. At the beginning of WL-M (Phase 2), all subjects continued to meet with a dietitian, but on a monthly basis. Figure 1 Schematic illustrating the experiment timeline comprised of a 12-week WL (phase 1), followed by a 52-week WL-M (phase 2). HP-IF-LC, high-protein, intermittent fasting, low-calorie; HH diet, heart healthy diet; WL, weight loss; WL-M, weight loss maintenance. Rationale A HH diet that meets the guidelines of the American Heart Association is comprised of 50–60% of energy from carbohydrates and approximately 15% from protein, with the remaining 25–35% from healthy fats. This diet is often recommended to improve cardiovascular health independent of WL (Fagerberg et al., 1984; Trumbo et al., 2002; Lichtenstein et al., 2006). Thus, it is of great interest to systematically compare the relative effectiveness of this HH diet and other diets, such as a HP-IF diet, in individuals with obesity during WL and long-term WL-M. In this context, using a two-phase study design, our first objective was to determine whether men and women with obesity demonstrate similar cardiovascular health improvements following a 12-week HP-IF-LC diet. Thereafter, we aimed to quantitatively compare the effectiveness of 52 weeks of a HP-IF vs. HH diet on WL-M and maintenance of cardiovascular enhancements produced by WL. Dietary intervention WL (Phase 1, weeks 1–12): HP-IF-LC diet Beginning at week 1, subjects consumed a HP-LC diet 6 days per week, along with an IF diet on the remaining day of the week. This diet was adhered to in conjunction with weekly counseling sessions with a registered dietitian. The composition of the HP-LC diet and the timing of meals are shown in Table 1. In order to induce an energy deficit without compromising lean body mass, the daily macronutrient distribution (30% protein, 45% carbohydrate, and 25% fat) has been used in this phase based on our previous study (Arciero et al., 2014). Total energy intake during HP-LC was 1200 and 1500 kcal per day for females and males, respectively. The eating schedule was designed to produce a regular frequency of meals and protein consumption during the WL and WL-M interventions. During WL, each meal consisted of approximately 20–30 g servings of high-quality protein in either whole food or supplement form. Subjects were told to eat 4–5 meals per day on HP-LC days, as depicted in Table 1. On the one IF day per week, daily energy intake consisted of 330 and 430 kcal for women and men, respectively. The composition of the diet for the IF day is depicted in Table 2. Table 1 Composition of high-protein low-calorie diet. Meal and recommended time of consumption Energy (kcal)a Description Breakfast (6:00–8:00 A.M.) 240 Liquid meal replacementb Lunch (11:00 A.M.–1 P.M.) 240 Liquid meal replacementb Afternoon snackc (2:00–4:00 P.M.) 150 Low-glycemic protein wafers or whole-food high protein snackfg Dinnerd (5:00–7:00 P.M.) 450 or 600 Evening snack (9:00–10:00 P.M.) 270 Meal replacement bare a The total daily macronutrient distribution for the HP-LC diet shown above was 30% protein, 45% carbohydrate, and 25% fat, and total energy intake for males and females was ~1500 and 1200 kcal, respectively. In addition to following this HP-LC diet 6 days per week, an IF diet was followed for one day per week. b Isalean Shake, Isagenix LLC, Chandler, AZ, USA. c Afternoon snack was consumed by males only. d Females consumed 450 kcal meal while males consumed 600 kcal meal. e Isalean Bar, Isagenix LLC, Chandler, AZ, USA. f Snack components for men: 2 hard-boiled egg with 1/3 cup fruit; 6 oz. cottage cheese with 1/3 cup fruit; 1 tbsp of nuts and 1/3 cup of fresh veggies or fruit; 1 slice of whole grain bread with 1.5 tsp nut butter; 1/2 cup Fresh fruit or veggies with 1/3 cup hummus; 8 oz plain greek yogurt; ¼ cup granola. g Snack components for women:1 hard-boiled egg with ¼ cup fruit; 4 oz. cottage cheese; ¼ cup fruit; 4 oz. plain non-fat greek yogurt; ¼cup granola; ½ slice bread; ¾ tsp nut butter; ¼ cup fruit or veggies with 1/5 tbsp hummus. Table 2 Composition of intermittent fasting diet. Food item/dietary supplement Consumption frequency Descriptiona Whole-food high protein snack 1/day 100 or 200 kcal for females and males, respectively Anti-oxidant rich powderb 6/day 120 kcal total Low-glycemic protein wafersb 3/day 90 kcal total Micronutrient supplementc 2/day Contains vitamins, minerals, phytonutrients, antioxidants, and essential fatty acids Herbal supplementd 1/day Multiple ingredients, such as wolfberry, kiwi, rhodiola root, harada, tribulus, and maca root a During the Phase 1 (weight loss) of the study, participants performed one day of IF per week, which consisted of a total energy intake of 330 kcal/d for women and 430 kcal/d for men. b Isalean Shake, Isagenix LLC, Chandler, AZ, USA. c Ageless Essentials with Product B, AM & PM, Isagenix LLC, Chandler, AZ, USA; consumed on IF and non-IF days. d Ionix Supreme, Isagenix LLC, Chandler, AZ, USA; consumed on IF and non-IF days. WL-M (Phase 2, weeks 13–64): HP-IF and HH diets Starting at week 13, participants chose whether to enter the HP-IF or HH diet group for Phase 2 of the study. Both groups (n = 10 for HP-IF; n = 14 for HH) received monthly dietary counseling from a registered dietitian, and subjects in both groups reported continued desire to lose weight after the initial 12 weeks. In order to resemble free-living conditions (i.e., conditions without excessive supervision), subjects were instructed to follow the guidelines of their respective diets without restrictions on physical activity or total food intake. However, they were encouraged to stay at an intake level necessary for weight maintenance, based on their calculated energy needs [measured resting metabolic rate (RMR) X Activity Factor]. The HP-IF group were provided 2 meal replacements per day (either two protein powder packets or one protein powder packet and one meal replacement bar) while the remaining 2–3 meals were whole foods, as guided by a dietitian. HP-IF subjects also performed IF 1 to 2 days per month. The HH group followed the dietary guidelines that are in compliance with the National Cholesterol Education Program Therapeutic Lifestyle Changes diet (i.e., < 35% of kcal as fat; 50–60% of kcal as carbohydrates; < 200 mg/dL of dietary cholesterol; and 20–30 g/day of fiber). Both groups (HH and HP-IF) had monthly meetings with a registered dietitian to establish healthy eating choices that were compliant with their meal plans. Additional counseling with the registered dietitian was made available to participants if necessary. During WL-M, the same timing of meals as during WL was implemented (i.e., 4–5 meals per day evenly spaced throughout the day). The only exception was during IF days for the HP-IF subjects. Compliance To encourage compliance, all subjects had weekly meetings with a registered dietitian during WL and monthly meetings during WL-M to incorporate healthy eating strategies while consuming their appropriate diet. Additionally, all subjects were given detailed written and verbal instructions for each diet plan (HP-IF-LC; HP-IF; HH). For example, participants were informed that HP-IF was designed to maintain intake close to the 1.8 g protein/kg body weight (BW), whereas the HH was designed to deliver 1.0 g protein/kg BW. Furthermore, as stated above, there were no differences between the subjects choosing HH or HP-IF diets for the WL-M phase. Monitoring of the meal plans was performed through daily subject-researcher interaction (e.g., telephone conversations, 2-day food diary analysis, weekly dietary intake journals inspections, distribution of weekly meal/supplement containers, return of empty packets and containers, and monthly group meetings). The PI (PJA) and/or an investigator held weekly meetings with all participants to verify compliance with the dietary meal plans, clarify dietary guidelines, and answer questions. Participants demonstrated a high compliance rate (>90%), which was defined as consuming more than 85% of their respective meals/supplemented feedings. Subjects were considered non-compliant if they were absent from more than two consecutive dietitian meetings or consumed ≥3 inappropriate meal/supplement servings a week for ≥2 consecutive weeks at a time. A 2-day food record was utilized to verify compliance to each respective diet (HP-IF-LC; HP-IF; HH). Based on our experience, a representative sample of only 2 days was adequate to assess subjects' stable and consistent intake during each stage. If necessary, subjects were instructed to record food intake every day. Food records were filled out by the participants at weeks 0, 11, and 63. A registered dietitian and a research team member provided instructions to the participants on making detailed records of portion sizes and food items. The dietary information was subsequently recorded into the food analysis program, The Food Processor SQL Edition (version 10.2.0 ESHA Research, Sale, OR, 2012). Single trained operators (E.W.) performed the analysis in order to eliminate inter-investigator variation. Each participant was also given a checklist in an effort to help adhere to the IF day regimen. Laboratory testing procedures BW and physical activities assessment Body weight was measured using an electronic scale during each testing visit without shoes and in minimal clothing. Standard BMI measurements were obtained by dividing the participant's weight (kg) by the square of their height (m2). Free-living, daily activity was monitored by an Actical accelerometer secured on the waist to ensure consistent levels of activity between the participants (Esliger et al., 2007; Hooker et al., 2011). Blood lipid determination A 12-h fasted venous blood sample (~20 ml) was obtained at baseline (week 0) and post-intervention (weeks 12 and 64). Blood was collected into EDTA-coated vacutainer tubes and centrifuged (Hettich Rotina 46R5) for 15 min at 2500 rpm at 4°C. After separation, plasma was stored at −70°C in small aliquots until analyzed. TC, high-density lipoprotein (HDL) cholesterol, and TG were assessed using the Cholestech LDX blood analysis system (Hayward, CA). Test-retest intraclass correlation (r) and coefficient of variation (CV) with n = 15 is: r = 0.95, CV = 3.2%, and r = 0.97, CV = 5.3%, for TC and HDL cholesterol, respectively. Vascular compliance measurement Vascular compliance can be measured non-invasively and gives information regarding cardiovascular disease risk, even in healthy individuals (Ring et al., 2014). Resting heart rate (HR) and systolic and diastolic BP (SBP; DBP) were obtained in the supine position as previously described (4, 6, 7). HR and BP were obtained by the same investigator (E.W.) following a minimum of 10 min of quiet resting. The Arteriograph (version 1.10.0.1, TensioMed Kft., Budapest, Hungary) device uses an upper arm cuff inflated to >35 mmHg above the subjects' actual systolic pressure. This causes a small diaphragm in the brachial artery to develop along the upper border of the over-pressurized cuff. A pulse wave is created as the central pressure changes, forming an early (direct) systolic wave (P1), late (reflected) systolic wave (P2), and diastolic wave(s) (P3). The device is able to record each of these suprasystolic pressure changes. Initially, the Arteriograph measures the systolic and diastolic BP oscillometrically, and then decompresses the cuff. Within a few seconds, the device re-inflates the cuff, first to the actual measured diastolic BP followed by the suprasystolic pressure, which is 35 mmHg over the actual systolic BP. The device records the signals from both pressure levels for 8 s. A computer receives all of the signals sent wirelessly by the device. Using the software, the augmentation index is determined by the following formula: Aix(%)=(P2-P1)/PP×100 where P1 reflects the early direct wave's amplitude; P2 refers to the late reflected systolic wave's amplitude; and PP equals the pulse pressure. Augmentation index was calculated for the brachial artery (brachial augmentation index; BAIX) and for the aorta (central augmentation index; CAIX). CAIX values are produced by the Arteriograph based on the correlation between previous simultaneously measured brachial and aortic augmentation indices. The aortic pulse wave velocity (PWVao) is determined by the wave reflection generated from the early direct pulse wave as it is reflected back primarily from the aortic bifurcation. Return time (RT) is determined by measuring the time interval between peaks from the early direct (P1) and reflected late (P2) systolic waves. The PWVao calculations were measured using the distance from the upper edge of the pubic bone to the sternal notch (Jugulum-Symphisis¼), as this provides the closest approximation of the true aortic length (Sugawara et al., 2008; Horváth et al., 2010). The parallel straight-line distance between these anatomical points was measured to allow for the calculation of the PWVao with the following formula: PWVao(m/s) = [Jug – Sy(m)]/[(RT/2(s)] Finally, calculation of the blood pressure using the Arteriograph was based on algorithms that have been previously validated (Németh et al., 2002; Horváth et al., 2010). Statistical analysis Statistical analysis was performed using SPSS software (Ver. 21; IBM-SPSS). All values are reported as means ± SE. Before the start of the study, sample size was determined through power analysis based on the major outcome variables (blood lipids and arterial compliance) to achieve an effect size of 0.25 with 80% power at alpha 0.05 based on the preliminary data. This analysis determined that n = 12 were required to detect significant differences between groups. A 2 × 2 factor repeated measures ANOVA (RMANOVA) was performed for the WL (HP-IF-LC, weeks 0–12) phase (sex; M/F and time; control baseline vs. 12 Weeks) and the WL-M (weeks 13–64) phase (group; HP-IF/HH and time; 13 weeks vs. 64 weeks) to determine main effects as shown in the results. Bonferroni's method was performed if there was an interaction between variables. A multivariate ANOVA was also performed as an additional analysis. Pearson's correlation coefficients were used to assess the relationships between body fat, blood lipids, cardiovascular function, and arterial compliance during WL and WL-M phases. The significance was set at p < 0.05, and trends were noted for 0.05 < p < 0.1. Percent change for dependent variables was calculated as (measurement 2) - (measurement 1)measurement 1*100. Results WL (Phase 1, HP-IF-LC diet) Forty participants completed Phase 1 of this study. Descriptive baseline characteristics of the participants are shown in Table 3. Results of the repeated-measures ANOVAs for our dependent variables are displayed in Table 4. No gender-by-time interactions were found for any variables after Phase 1. The average loss of BW reached 10% of original BW following the 12-week WL (10.4 ± 0.6%, p < 0.001). BMI was also significantly decreased post Phase 1 compared to baseline (BMI: 37.5 ± 0.9 vs. 33.7 ± 0.8 kg/m2, p < 0.001; Figures 2A, 4A). However, there was no difference observed in BMI at week 12 between those who chose to enter either the HH or HP-IF groups. Table 3 Baseline (week 0) characteristics of participants for WL (Phase 1). Male (n = 21) Female (n = 19) Total (n = 40) Characteristics Mean SE Mean SE p-value Mean SE Age (years) 46.1 1.5 50.0 2.3 0.163 48.0 1.4 Weight (kg) 120.1 4.8 99.5 2.8 0.001 110.3 3.3 Height (cm) 179.0 1.7 163.1 1.0 0.000 171.4 1.6 BMI (kg/m2) 37.5 1.5 37.4 1.1 0.945 37.5 0.95 HR (bpm) 65.0 1.9 64.7 2.8 0.430 64.9 1.6 SBP (mmHg) 127.7 2.1 121.9 2.7 0.034 125.2 1.7 DBP (mmHg) 81.7 2.4 76.9 2.6 0.098 79.5 1.8 BMI, body mass index; HR, heart rate; SBP, systolic blood pressure; DBP, diastolic blood pressure. Table 4 Results of weight loss phase (Phase 1). Gender Baseline 12 weeks p-value (interaction) p-value (time) p-value (gender) BMI (kg/m2) M 38.3 ± 1.6 34.3 ± 1.3 0.715 < 0.001* 0.598 F 37.4 ± 1.1 33.7 ± 1.0 TC (mg/dL) M 188.2 ± 7.9 160.8 ± 6.6 0.670 < 0.001* 0.219 F 199.6 ± 9.6 168.7 ± 8.2 LDL (mg/dL) M 114.6 ± 6.7 104.1 ± 6.4 0.505 0.036* 0.644 F 121.0 ± 6.9 104.0 ± 5.8 HDL (mg/dL) M 48.2 ± 3.1 43.5 ± 2.8 0.846 0.006* 0.550 F 51.1 ± 3.1 45.7 ± 2.7 TG (mg/dL) M 134.2 ± 17.7 84.9 ± 10.7 0.686 < 0.001* 0.307 F 144.3 ± 13.7 102.7 ± 9.1 HR (bpm) M 65.5 ± 2.2 60.8 ± 2.3 0.878 0.017* 0.867 F 64.8 ± 2.8 60.6 ± 2.3 SBP (mmHg) M 126.3 ± 2.0 116.5 ± 2.5 0.426 0.002* 0.195 F 121.9 ± 2.7 115.3 ± 2.0 DBP (mmHg) M 80.1 ± 1.8 72.6 ± 1.8 0.783 < 0.001 0.158 F 76.9 ± 2.6 68.7 ± 1.8 PWVao (m/s) M 7.8 ± 0.4 7.2 ± 0.3 0.571 0.001* 0.764 F 7.8 ± 0.3 6.9 ± 0.3 BAIX (%) M −27.0 ± 6.5 −37.3 ± 4.4 0.289 0.134 0.035** F −10.6 ± 7.4 −12.6 ± 7.3 CAIX (%) M 24.0 ± 3.3 18.8 ± 2.2 0.289 0.133 0.035** F 32.3 ± 3.7 31.2 ± 3.7 Return Time (s) M 142.4 ± 6.0 157.4 ± 4.9 0.752 < 0.001* 0.101 F 128.3 ± 5.4 146.1 ± 5.1 Data presented as mean ± SE. Return time refers to the time intervals between peaks from the early direct and reflected late systolic waves. * Significant difference based on time (p < 0.05); ** Significant difference between genders (p < 0.05). BMI, body mass index; TC, total cholesterol; LDL, low-density lipoprotein; HDL, high-density lipoprotein; TG, triglycerides; HR, heart rate; SBP, systolic blood pressure; DBP, diastolic blood pressure; PWVao, pulse wave velocity; BAIX, brachial augmentation index; CAIX, central augmentation index; M, male; F, female. Figure 2 Dietary effects on BMI during phase 1 (A) and phase 2 (B) collapsed across genders. (A) Effect of 12 week HP-IF-LC intervention (Phase 1) on BMI (n = 40). (B) Percent change in BMI between HP-IF and HH groups during Phase 2 (n = 24). ***Significant difference compared to baseline (p < 0.001). *Trend for significant difference compared to HH group (p = 0.069). BMI, body mass index; HP-IF, high-protein, intermittent fasting; HH, heart healthy. Significant decreases were also found in the levels of plasma TG, LDL, and TC following the WL (Figures 3A–D). As expected, HR and BP were also significantly decreased following the 12 weeks WL (Figures 4B–D); however, HR only decreased in male subjects. Figure 3 Effect of 12 week HP-IF-LC intervention (Phase 1) on (A) triglycerides, (B) LDL, (C) total cholesterol, (D) total cholesterol/HDL ratio of males and females. Significant difference compared to baseline of the same gender indicated by: *p < 0.05; **p < 0.01; ***p < 0.001. LDL, low-density-lipoprotein; HDL, high-density-lipoprotein. Figure 4 Effect of 12 week HP-IF-LC intervention (Phase 1) on (A) BMI, (B) heart rate, (C) systolic blood pressure, (D) diastolic blood pressure, (E) PWVao, (F) return time (the time intervals between peaks from the early direct and reflected late systolic waves), (G) BAIX, and (H) CAIX of males and females. Significant difference compared to baseline of the same gender indicated by: *p < 0.05; **p < 0.01; ***p < 0.001. BMI, body mass index; PWVao, aortic pulse wave velocity; BAIX, brachial augmentation index; CAIX, central augmentation index. Moreover, our results revealed that PWVao, an important measurement of arterial stiffness, was significantly decreased whereas the return time (RT) was increased following Phase 1 (Figures 4E,F). There were not significant changes in BAIX or CAIX (Figures 4G,H). Percent changes in anthropometrics (BW and BMI) were significantly positively correlated with the corresponding changes in plasma TG (r = 0.30, 0.30, p < 0.05, respectively) and TC (r = 0.45, 0.48, p < 0.01, respectively), but not with the ratio of TC/HDL (Table 5). Interestingly, the percent change in CAIX was significantly correlated with HR, systolic BP, and diastolic BP (r = −0.29, r = −0.31, r = 0.29, p < 0.05, respectively). Table 5 Pearson correlation coefficients for the percent change in body weight/BMI and blood lipid profiles following the phase 1 study. TRG (% Δ) LDL (% Δ) TC (% Δ) TC/HDL (% Δ) Body metrics r-value p-value r-value p-value r-value p-value r-value p-value Body weight (% Δ) 0.30 0.03* 0.22 0.09 0.45 0.00** 0.22 0.09 BMI (% Δ) 0.30 0.03* 0.27 0.05 0.48 0.00** 0.19 0.12 BMI, body mass index; TRG, triglycerides; LDL, low density lipoprotein; TC, total cholesterol; TC/HDL, total cholesterol/high density lipoprotein ratio. * Significantly correlated (p < 0.05). ** Significantly correlated (p < 0.01). WL-M (Phase 2, HP-IF, and HH diets) Male and female participants were pooled for Phase 2. Twenty-four participants successfully completed Phase 2, while 16 were excluded due to drop-out and non-compliance. Descriptive characteristics of the participants are shown in Table 6, and the results of the repeated measures ANOVAs for our dependent variables are displayed in Table 7. TC, LDL, HDL, TG, HR, SBP, DBP, and return time increased over time in both groups. However, there were trends (0.05 < p < 0.1) for less regain in BMI (p = 0.069; Figure 2B) and LDL (p = 0.068) in the HP-IF group. There was also a reduced PWVao in the HP-IF group (p = 0.045) and increased CAIX over time in both groups (p = 0.084). Table 6 Baseline (week 12) characteristics of participants for WL-M (Phase 2). HP-IF (n = 10) HH (n = 14) Total (n = 24) Characteristics Mean SE Mean SE p-value Mean SE Age (years) 50.9 3.1 50.0 1.8 0.791 50.4 1.6 Weight (kg) 90.4 2.4 95.1 3.7 0.306 93.1 2.4 Height (cm) 169.6 3.1 170.2 3.3 0.903 169.9 2.3 BMI (kg/m2) 31.6 0.82 33.1 1.3 0.076 32.5 0.81 HR (bpm) 57.4 2.8 58.7 2.7 0.607 58.2 1.9 SBP (mmHg) 114.0 3.3 110.9 2.5 0.534 112.2 2.0 DBP (mmHg) 68.9 2.6 69.3 2.0 0.859 69.1 1.5 BMI, body mass index; HR, heart rate; SBP, systolic blood pressure; DBP, diastolic blood pressure. Table 7 Results of weight loss maintenance phase (Phase 2). Diet Baseline (week 12) 64 weeks p-value (interaction) p-value (time) p-value (diet) BMI (kg/m2) HP-IF 31.6 ± 0.8 31.8 ± 1.2 0.069* 0.102 0.264 HH 32.2 ± 1.4 34.0 ± 1.4 TC (mg/dL) HP-IF 157.6 ± 8.4 184.4 ± 11.9 0.339 <0.001** 0.664 HH 158.0 ± 12.8 195.2 ± 11.1 LDL (mg/dL) HP-IF 101.5 ± 7.7 113.1 ± 8.6 0.068* 0.002** 0.715 HH 90.0 ± 8.5 116.8 ± 11.3 HDL (mg/dL) HP-IF 38.4 ± 3.2 50.1 ± 3.1 0.107 0.001** 0.365 HH 46.8 ± 4.7 51.6 ± 4.5 TG (mg/dL) HP-IF 96.1 ± 12.0 124.1 ± 21.3 0.281 0.007** 0.764 HH 91.0 ± 15.5 142.4 ± 22.9 HR (bpm) HP-IF 57.4 ± 2.8 60.0 ± 2.9 0.367 0.008** 0.677 HH 57.5 ± 2.7 62.5 ± 1.7 SBP (mmHg) HP-IF 114.0 ± 3.3 130.8 ± 4.4 0.706 0.001** 1.000 HH 112.8 ± 2.9 132.0 ± 5.3 DBP (mmgHg) HP-IF 68.9 ± 2.6 80.7 ± 2.9 0.987 <0.001** 0.514 HH 70.8 ± 2.2 82.5 ± 3.8 PWVao (m/s) HP-IF 7.4 ± 0.6 7.0 ± 0.3 0.094* 0.665 0.992 HH 6.9 ± 0.3 7.5 ± 0.4 BAIX (%) HP-IF −9.2 ± 14.1 −10.2 ± 14.5 0.156 0.025 0.641 HH −36.8 ± 7.4 −15.9 ± 6.7 CAIX (%) HP-IF 30.9 ± 6.0 31.6 ± 5.8 0.288 0.084* 0.812 HH 26.3 ± 3.7 32.7 ± 4.0 Return Time (s) HP-IF 146.6 ± 8.2 138.3 ± 6.8 0.220 0.002** 0.797 HH 156.0 ± 7.4 134.7 ± 7.2 Data presented as mean ± SE. Return time refers to the time intervals between peaks from the early direct and reflected late systolic waves. * Trend for significant difference (0.05 < p < 0.1); ** Significant difference based on time (p < 0.05). BMI, body mass index; TC, total cholesterol; LDL, low-density lipoprotein; HDL, high-density lipoprotein; TG, triglycerides; HR, heart rate; SBP, systolic blood pressure; PWVao, pulse wave velocity; BAIX, brachial augmentation index; CAIX, central augmentation index; HP-IF, high-protein intermittent fasting; HH, heart healthy. No significant difference was observed at the beginning of Phase 2 between the HP-IF and HH groups for any blood lipid components. Although there were no statistically significant differences in respect to the changes in blood lipid profiles between the HP-IF and HH groups following Phase 2, the HP-IF group demonstrated a tendency toward lower percentage increases compared to HH group in all lipid profiles except HDL (Figures 5A–F). Figure 5 Percent changes in (A) triglycerides, (B) LDL, (C) HDL, (D) non HDL lipids, (E) total cholesterol, and (F) total cholesterol/HDL ratio between the HP-IF and HH groups following Phase 2 (52 weeks). HP-IF, high-protein, intermittent fasting; HH, heart healthy; LDL, low-density lipoprotein; HDL, high-density-lipoprotein. Our results did not show a meaningful difference between the HP-IF and HH groups in terms of systolic BP, diastolic BP, or HR following phase 2. It is worth noting that the percent change in PWVao was much less in the HP-IF group than in the HH group following Phase 2 (−2.5 ± 4.1% vs. 11.2 ± 4.6%; Figure 6A). The difference in return time following the WL-M phase was also noticeable (HP-IF: −4.6 ± 3.8% vs. HH: −11.8 ± 3.0%, p = 0.220), although it was not statistically significant (Figure 6B). BAIX and CAIX percent changes are depicted in Figures 6C,D. Figure 6 Percent changes in (A) PWVao, (B) return time (the time intervals between peaks from the early direct and reflected late systolic waves), (C) BAIX, and (D) CAIX between the HP-IF and HH groups following Phase 2 (52 weeks). *Significant difference from the HH group (p = 0.045), as assessed by repeated-measures ANOVA. PWVao, aortic pulse wave velocity; HP-IF, high-protein, intermittent fasting; HH, heart healthy; BAIX, brachial augmentation index; CAIX, central augmentation index. The percent change in systolic and diastolic BP was positively correlated with changes in BAIX (r = 0.52, p < 0.05; r = 0.63, p < 0.01, respectively). We also found that changes in TRG were positively correlated with changes in LDL, BAIX, and CAIX (r = 0.42, r = 0.49, r = 0.50, p < 0.05, respectively; Table 8). Table 8 Pearson correlation coefficients for the percent change in variables following the phase 2 study. LDL (% Δ) BAIX (% Δ) CAIX (% Δ) PWVao (% Δ) r-value p-value r-value p-value r-value p-value r-value p-value Systolic BP (% Δ) −0.22 0.17 0.52 0.01* 0.05 0.41 0.11 0.32 Diastolic BP (% Δ) −0.19 0.20 0.63 0.00** 0.30 0.10 −0.28 0.12 TRG (% Δ) 0.42 0.03* 0.49 0.02* 0.50 0.01* −0.27 0.12 LDL, low density lipoprotein; BAIX, brachial augmentation index; CAIX, central augmentation index; PWVao, aortic pulse wave velocity; TRG, triglycerides; BP, blood pressure. * Significantly correlated (p < 0.05). ** Significantly correlated (p < 0.01). Discussion Previous studies have supported the independent effects of HP, HH, LC, or IF diets on WL success and cardiometabolic improvement (Arciero et al., 2006; Kelishadi et al., 2008; Clifton et al., 2009; Camhi et al., 2010; Klempel et al., 2012; Song et al., 2016). The current study provides convincing evidence that the combined HP-IF-LC diet successfully induces marked body WL, and is likely associated with reduced blood lipid levels and enhanced arterial compliance among men and women with obesity. During WL (Phase 1), the percent change in BW and BMI was significantly correlated with the changes in certain blood lipid variables such as TRG and TC. Subsequently in WL-M (Phase 2), subjects who adhered to the HP-IF diet experienced reduced weight regain and demonstrated better arterial compliance than those consuming the HH diet. Collectively, these data suggest that the HP-IF diet may be a new type of healthy diet which could be advantageous in maintaining the long-term health benefits from initial WL in overweight and obese individuals. Although significant WL occurred in both men and women, we found no sex effect in the parameters of our study. The alarming rate of obesity in the United States is often attributed to the consumption of low-quality, high caloric diets (Bruemmer, 2012). Obesity was established as an independent risk contributor for CVD in the 26 plus-year follow-up from the original Framingham cohort study (Hubert et al., 1983). Numerous studies have investigated WL and the corresponding alterations in cardiovascular health and disease risk (Kelishadi et al., 2008; Clifton et al., 2009; Klempel et al., 2012; Mozaffarian et al., 2015; Pedersen et al., 2015). However, a major challenge we currently face is effectively maintaining successful WL and the accompanying health benefits. Therefore, it is paramount that researchers and clinicians develop WL strategies that are also safe and effective for aiding in long-term heart health (Bruemmer, 2012). While recidivism often appears to be unavoidable in long-term WL studies, maintaining even modest WL can be clinically significant (Stevens et al., 2001; Harsha and Bray, 2008). Research has reported that short-term interventions similar to our HP-IF-LC diet effectively promote WL (Clifton et al., 2009; Klempel et al., 2012). One novel aspect of our study was comparing a traditional HH diet with a HP-IF diet using a 1-year follow-up to the initial 12-week WL period. We tracked changes in BMI, blood lipids, as well as cardiovascular and arterial compliance measures during both WL and WL-M phases. This could be attributed to the relatively higher protein content in the HP-IF diet, but the 1 to 2 days per month of IF likely exerted beneficial effects by counteracting normal weight gain. The individual and combined effects of HP and IF warrant further investigation. The combined diet plan (HP-IF-LC) utilized in this study is not overly complex, and the modified IF that was included during the WL-M phase was only employed on 1 to 2 days per month. The HP-IF diet provided sufficient energy intake and was not intended for continued weight loss during the WL-M phase. Rather, it was included simply to more closely mimic the diet during the WL phase without the intended WL. Moreover, both HP and IF have been reported to induce significant health benefits, thus this was a logical dietary intervention to compare to the HH diet. Previous research has shown that elevated resting HR can predict cardiovascular mortality in men and women (Fox et al., 2007; Cooney et al., 2010). Hypertension, which is associated with obesity, is also known to predispose individuals to CVD (Aucott et al., 2005). These serve as additional support for the necessity of long-term WL interventions which can sustain improvements in these significant cardiovascular variables. In addition, it is well-known that WL in hypertensive individuals is associated with a decrease in BP in short-term interventions either by caloric restriction, exercise, or both (Neter et al., 2003; Elmer et al., 2006; Harsha and Bray, 2008). Nevertheless, long-term studies on the effects of WL induced by different diets on BP are still lacking. As such, we measured systolic and diastolic BP, as well as resting heart rate, at weeks 0, 12, and 64. We observed a decline in systolic BP and diastolic BP following the 12-week WL Phase. However, the participants in the present study were, on average, categorized as having normal BP at study commencement, so the complete impact of these BP reductions is not entirely clear, although they could potentially aid in preventing a rise in BP over time. Some degree of weight relapse occurred during the WL-M phase, although body weight and BMI still remained below baseline levels. During the weight relapse, simultaneous increases in HR and BP were seen, but neither diet (HP-IF or HH) was shown to be more advantageous in lessening the recidivism in either BP or HR during the WL-M phase. In the future, extensive studies requiring a larger sample and longer-term interventions, possibly combined with exercise regimens, are needed to evaluate the advantages of the health-promoting effects of HP-IF on BP and HR changes. The degree of WL has previously been linked to the reduction of CVD risk factors (Van Gaal et al., 2005). Consistent with previous studies (Melanson et al., 2003; Harder et al., 2004), we found that a 12-week WL program markedly reduced triglyceride levels and moderately lowered LDL (−25 and −11% in the present study), without improvements in HDL. In the future, moderate exercise combined with the HP-IF dietary intervention might be a promising approach to increase HDL levels while still reducing triglycerides and LDL. Following the 1 year WL-M phase, the improved lipid risk factors rebounded slightly, which was associated with recidivism of the weight gain. Although the weight relapse seems to be common for long-term WL studies, Wing et al. concluded that large or modest WL at 1 year is still clinically beneficial for diabetic patients with obesity (Wing et al., 2011), and it has been reported that acute weight loss affects long-term cardiovascular function (Corbi et al., 2002). Further studies should address the additional factors associated with CVD risk, such as the presence of diabetes and race difference, as well as the use of Framingham risk score to understand the beneficial effects of HP-IF diet. An augmentation index is used as a measure of arterial stiffness, which has been demonstrated to adversely affect cardiovascular health (Nürnberger et al., 2002). For example, type 1 and type 2 diabetes are both characterized by increased arterial stiffness (Wilkinson et al., 2000). Our data revealed non-significant reductions in BAIX and CAIX after Phase 1, as well as a reduction in PWVao in the HP-IF group compared to the HH group in Phase 2. This finding suggests that the participants in the HP-IF group may have benefits from reduced arterial stiffness as a result of continuing the HP-IF diet for 1 year. Our data also suggest a positive relationship between the changes of BAIX, CAIX, and the change in TRG during phase 2. The beneficial outcome of the HP-IF diet may have clinical significance as studies have shown arterial stiffness to have a strong predictive value for cardiovascular events and all-cause mortality (Vlachopoulos et al., 2010). In accordance with our results, a study which examined patients with atherosclerotic disease found that the augmentation index was significantly correlated with both systolic BP and diastolic BP (Nürnberger et al., 2002). PWVao, another index of arterial stiffness, increases as a result of hypertension, atherosclerosis, or aging (Willum-Hansen et al., 2006). Research has shown that PWVao is a useful biomarker that can be successfully used to predict future cardiovascular risk in patients (Ben-Shlomo et al., 2014). Our findings are consistent with other studies showing that WL is associated with improved PWVao (Wildman et al., 2005; Barinas-Mitchell et al., 2006; Rider et al., 2010). Moreover, after the 1-year follow up, there is a lower PWVao over time in the HP-IF group as compared to the HH group, which relapsed to higher PWVao values. Several significant strengths characterize the present study. First, there were two separate phases, WL and WL-M, with two different baselines which allowed for comparison between and within both interventions. Second, we diligently monitored the physical activity level and diet of all participants through both Phase 1 and Phase 2. On the other hand, the study also had some limitations. The age range of the cohort crosses over several reproductive stages of women (i.e., premenopausal, perimenopausal, and postmenopausal), which could affect the results. As is often the case in long-term human studies, the total number of our participants in phase 2 was less than that of phase 1 due to factors such as scheduling conflicts and non-compliance of subjects. Presumably, participants who finished the second phase were more motivated and committed to the study than those who dropped out. Therefore, we were unable to collect some follow-up data from drop outs, potentially increasing the possibility of false positive results. However, significant differences in baseline measurements did not exist between groups, meaning that the dropout and non-compliance did not produce systematic differences in the individuals in each group. The intensity of our dietary intervention may have been more stringent compared to other standard dietary interventions. As a result, future studies may need to be directed to a more practical protocol modified from current plans. Perspective and significance In summary, a 12-week HP-IF-LC diet effectively improved BMI, lowered cholesterol, favorably altered cardiovascular variables, and reduced resting HR and BP. A novel finding is that a 1-year HP-IF intervention minimizes weight regain and improves BMI, leading to enhanced cardiovascular parameters. A comparison between this diet and other WL and WL-M trials should be performed in future studies, especially given that the HP-IF diet appears to be “advantageous” in maintaining the long-term health benefits from initial WL. A comparison WL group would be valuable in informing us whether the HP-IF-LC diet results in better metabolic and vascular outcomes than other traditional WL diets. Researchers should investigate the fasting glucose and HbA1C levels of the subjects, as diabetic and pre-diabetic subjects are prone to cardiovascular disease. In future studies, we will determine the exact effect of feedings during HP-IF-LC on gastrointestinal tract function, postprandial thermogenesis, RMR, nitrogen balance and glycosuria as compared to other diets. Furthermore, as previous studies have confirmed the involvement of reactive oxygen species in multiple CVD (Zuo et al., 2011; Zhu and Zuo, 2013), assessing the potential effects of different diets on the redox balance in the body may be an area of interest for future dietary studies at the molecular level. Further identification of an optimal role of cardiac preconditioning regimens and resistance training may be necessary as these methods may likely modulate lean body mass and muscle strength during WL (Zuo et al., 2013; Ataee et al., 2014). Author contributions PA conception and design of research; PA performed experiments; LZ, GT, BP, EW, and PA analyzed data; LZ, FH, and PA interpreted results of experiments; LZ and BP prepared figures; LZ, FH, BP, and PA drafted manuscript; LZ, FH, BP, GT, and PA edited and revised manuscript; LZ, FH, GT, BP, EW, and PA approved final version of manuscript. Funding This study was primarily supported (90%) by Isagenix International Grant #13-086 and (http://www.isagenix.com/?sc_lang=en-US) to PA. Additional support (10%) was provided by OSU-HRS Fund 013000. Conflict of interest statement Isagenix International, a multilevel marketing company, funded this study to PA. However, the company did not have access to the data throughout the study, and no patents, products in development or marketed product is to declare. This does not alter our adherence to all the Frontiers in Physiology policies on sharing data and materials. All the other authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. We appreciate the time and effort given by all of our volunteers. We respectfully acknowledge the assistance of Tingyang Zhou, Chia-Chen Chuang, Michael Chien, Alex Ziegler, Joshua Stringer, and Nicole Otenbaker during the manuscript preparation. We are grateful for the assistance of Olivia Minicucci, Bradley Schuler, Allison Keller, Karen Arciero DPT, Caitlin Sheridan, Rebecca Mahoney RD, and Elisa Kline RD throughout the entire intervention. ==== Refs References Arciero P. J. Baur D. Connelly S. Ormsbee M. J. (2014 ). Timed-daily ingestion of whey protein and exercise training reduces visceral adipose tissue mass and improves insulin resistance: the PRISE study . J. Appl. Physiol. 117 , 1 –10 . 10.1152/japplphysiol.00152.2014 24833780 Arciero P. J. Gentile C. L. Martin-Pressman R. Ormsbee M. J. Everett M. Zwicky L. . (2006 ). Increased dietary protein and combined high intensity aerobic and resistance exercise improves body fat distribution and cardiovascular risk factors . Int. J. Sport Nutr. Exerc. Metab. 16 , 373 –392 . 10.1123/ijsnem.16.4.373 17136940 Ataee J. Koozehchian M. S. Kreider R. B. Zuo L. (2014 ). Effectiveness of accommodation and constant resistance training on maximal strength and power in trained athletes . PeerJ 2 :e441 . 10.7717/peerj.441 25024910 Aucott L. Poobalan A. Smith W. C. Avenell A. Jung R. Broom J. (2005 ). Effects of weight loss in overweight/obese individuals and long-term hypertension outcomes: a systematic review . Hypertension 45 , 1035 –1041 . 10.1161/01.HYP.0000165680.59733.d4 15897373 Barinas-Mitchell E. Kuller L. H. Sutton-Tyrrell K. Hegazi R. Harper P. Mancino J. . (2006 ). Effect of weight loss and nutritional intervention on arterial stiffness in type 2 diabetes . Diabetes Care 29 , 2218 –2222 . 10.2337/dc06-0665 17003296 Ben-Shlomo Y. Spears M. Boustred C. May M. Anderson S. G. Benjamin E. J. . (2014 ). Aortic pulse wave velocity improves cardiovascular event prediction: an individual participant meta-analysis of prospective observational data from 17,635 subjects . J. Am. Coll. Cardiol. 63 , 636 –646 . 10.1016/j.jacc.2013.09.063 24239664 Blaak E. (2001 ). Gender differences in fat metabolism . Curr. Opin. Clin. Nutr. Metab. Care 4 , 499 –502 . 10.1097/00075197-200111000-00006 11706283 Bruemmer D. (2012 ). Targeting angiogenesis as treatment for obesity . Arterioscler. Thromb. Vasc. Biol. 32 , 161 –162 . 10.1161/ATVBAHA.111.241992 22258895 Camhi S. M. Stefanick M. L. Ridker P. M. Young D. R. (2010 ). Changes in C-reactive protein from low-fat diet and/or physical activity in men and women with and without metabolic syndrome . Metab. Clin. Exp. 59 , 54 –61 . 10.1016/j.metabol.2009.07.008 19709693 Clifton P. M. Bastiaans K. Keogh J. B. (2009 ). High protein diets decrease total and abdominal fat and improve CVD risk profile in overweight and obese men and women with elevated triacylglycerol . Nutr. Metab. Cardiovasc. Dis. 19 , 548 –554 . 10.1016/j.numecd.2008.10.006 19179060 Cooney M. T. Vartiainen E. Laatikainen T. Juolevi A. Dudina A. Graham I. M. (2010 ). Elevated resting heart rate is an independent risk factor for cardiovascular disease in healthy men and women . Am. Heart J. 159 , 612 –619 .e3. 10.1016/j.ahj.2009.12.029 20362720 Corbi G. M. Carbone S. Ziccardi P. Giugliano G. Marfella R. Nappo F. . (2002 ). FFAs and QT intervals in obese women with visceral adiposity: effects of sustained weight loss over 1 year . J. Clin. Endocrinol. Metab. 87 , 2080 –2083 . 10.1210/jcem.87.5.8516 11994344 Damsgaard C. T. Papadaki A. Jensen S. M. Ritz C. Dalskov S. M. Hlavaty P. . (2013 ). Higher protein diets consumed ad libitum improve cardiovascular risk markers in children of overweight parents from eight European Countries . J. Nutr. 143 , 810 –817 . 10.3945/jn.112.173427 23596158 de Luis D. A. Aller R. Izaola O. Romero E. (2016 ). Effect of -55CT Polymorphism of UCP3 on insulin resistance and cardiovascular risk factors after a high protein/low carbohydrate versus a standard hypocaloric diet . Ann. Nutr. Metab. 68 , 157 –163 . 10.1159/000444150 26848765 Elmer P. J. Obarzanek E. Vollmer W. M. Simons-Morton D. Stevens V. J. Young D. R. . (2006 ). Effects of comprehensive lifestyle modification on diet, weight, physical fitness, and blood pressure control: 18-month results of a randomized trial . Ann. Intern. Med. 144 , 485 –495 . 10.7326/0003-4819-144-7-200604040-00007 16585662 Esliger D. W. Probert A. Connor Gorber S. Bryan S. Laviolette M. Tremblay M. S. (2007 ). Validity of the Actical accelerometer step-count function . Med. Sci. Sports Exerc. 39 , 1200 –1204 . 10.1249/mss.0b013e3804ec4e9 17596790 Fagerberg B. Andersson O. K. Isaksson B. Bjorntorp P. (1984 ). Blood pressure control during weight reduction in obese hypertensive men: separate effects of sodium and energy restriction . Br. Med. J. (Clin. Res. Ed). 288 , 11 –14 . 10.1136/bmj.288.6410.11 6418295 Flegal K. M. Carroll M. D. Ogden C. L. Curtin L. R. (2010 ). Prevalence and trends in obesity among US adults, 1999-2008 . JAMA 303 , 235 –241 . 10.1001/jama.2009.2014 20071471 Fox K. Borer J. S. Camm A. J. Danchin N. Ferrari R. Lopez Sendon J. L. . (2007 ). Resting heart rate in cardiovascular disease . J. Am. Coll. Cardiol. 50 , 823 –830 . 10.1016/j.jacc.2007.04.079 17719466 Hansen-Krone I. J. Enga K. F. Njolstad I. Hansen J. B. Braekkan S. K. (2012 ). Heart healthy diet and risk of myocardial infarction and venous thromboembolism The Tromso Study . Thromb Haemostasis 108 , 554 –560 . 10.1160/TH11-11-0818 22739999 Harder H. Dinesen B. Astrup A. (2004 ). The effect of a rapid weight loss on lipid profile and glycemic control in obese type 2 diabetic patients . Int. J. Obes. Relat. Metab. Disord. 28 , 180 –182 . 10.1038/sj.ijo.0802529 14610532 Harsha D. W. Bray G. A. (2008 ). Weight loss and blood pressure control (Pro) . Hypertension 51 , 1420 –1425 ; discussion 1425. 10.1161/hypertensionaha.107.094011 18474829 Harvie M. N. Pegington M. Mattson M. P. Frystyk J. Dillon B. Evans G. . (2011 ). The effects of intermittent or continuous energy restriction on weight loss and metabolic disease risk markers: a randomized trial in young overweight women . Int. J. Obes. (Lond). 35 , 714 –727 . 10.1038/ijo.2010.171 20921964 Hooker S. P. Feeney A. Hutto B. Pfeiffer K. A. Mciver K. Heil D. P. . (2011 ). Validation of the actical activity monitor in middle-aged and older adults . J. Phys. Act. Health 8 , 372 –381 . 10.1123/jpah.8.3.372 21487136 Horváth I. G. Nemeth A. Lenkey Z. Alessandri N. Tufano F. Kis P. . (2010 ). Invasive validation of a new oscillometric device (Arteriograph) for measuring augmentation index, central blood pressure and aortic pulse wave velocity . J. Hypertens. 28 , 2068 –2075 . 10.1097/HJH.0b013e32833c8a1a 20651604 Hubert H. B. Feinleib M. Mcnamara P. M. Castelli W. P. (1983 ). Obesity as an independent risk factor for cardiovascular disease: a 26-year follow-up of participants in the Framingham Heart Study . Circulation 67 , 968 –977 . 10.1161/01.CIR.67.5.968 6219830 Katare R. G. Kakinuma Y. Arikawa M. Yamasaki F. Sato T. (2009 ). Chronic intermittent fasting improves the survival following large myocardial ischemia by activation of BDNF/VEGF/PI3K signaling pathway . J. Mol. Cell. Cardiol. 46 , 405 –412 . 10.1016/j.yjmcc.2008.10.027 19059263 Kelishadi R. Hashemi M. Mohammadifard N. Asgary S. Khavarian N. (2008 ). Association of changes in oxidative and proinflammatory states with changes in vascular function after a lifestyle modification trial among obese children . Clin. Chem. 54 , 147 –153 . 10.1373/clinchem.2007.089953 17998270 Klempel M. C. Kroeger C. M. Bhutani S. Trepanowski J. F. Varady K. A. (2012 ). Intermittent fasting combined with calorie restriction is effective for weight loss and cardio-protection in obese women . Nutr. J. 11 :98 . 10.1186/1475-2891-11-98 23171320 Leidy H. J. Clifton P. M. Astrup A. Wycherley T. P. Westerterp-Plantenga M. S. Luscombe-Marsh N. D. (2015 ). The role of protein in weight loss and maintenance . Am. J. Clin. Nutr. 101 , 1320s –1329s . 10.3945/ajcn.114.084038 Lichtenstein A. H. Appel L. J. Brands M. Carnethon M. Daniels S. Franch H. A. . (2006 ). Diet and lifestyle recommendations revision 2006: a scientific statement from the American Heart Association Nutrition Committee . Circulation 114 , 82 –96 . 10.1161/CIRCULATIONAHA.106.176158 16785338 Lönnqvist F. Thörne A. Large V. Arner P. (1997 ). Sex differences in visceral fat lipolysis and metabolic complications of obesity . Arterioscler. Thromb. Vasc. Biol. 17 , 1472 –1480 . 10.1161/01.ATV.17.7.1472 9261282 Mattson M. P. Wan R. Q. (2005 ). Beneficial effects of intermittent fasting and caloric restriction on the cardiovascular and cerebrovascular systems . J. Nutr. Biochem. 16 , 129 –137 . 10.1016/j.jnutbio.2004.12.007 15741046 Melanson K. Gootman J. Myrdal A. Kline G. Rippe J. M. (2003 ). Weight loss and total lipid profile changes in overweight women consuming beef or chicken as the primary protein source . Nutrition 19 , 409 –414 . 10.1016/S0899-9007(02)01080-8 12714091 Mozaffarian D. Benjamin E. J. Go A. S. Arnett D. K. Blaha M. J. Cushman M. (2015 ). Executive summary: heart disease and stroke statistics-2015 update a report from the American Heart Association . Circulation 131 , 434 –441 . 10.1161/CIR.0000000000000157 Németh Z. Moczar K. Deak G. (2002 ). Evaluation of the Tensioday ambulatory blood pressure monitor according to the protocols of the British Hypertension Society and the Association for the Advancement of Medical Instrumentation . Blood Press. Monit. 7 , 191 –197 . 10.1097/00126097-200206000-00008 12131077 Neter J. E. Stam B. E. Kok F. J. Grobbee D. E. Geleijnse J. M. (2003 ). Influence of weight reduction on blood pressure: a meta-analysis of randomized controlled trials . Hypertension 42 , 878 –884 . 10.1161/01.HYP.0000094221.86888.AE 12975389 Nürnberger J. Keflioglu-Scheiber A. Opazo Saez A. M. Wenzel R. R. Philipp T. Schafers R. F. (2002 ). Augmentation index is associated with cardiovascular risk . J. Hypertens. 20 , 2407 –2414 . 10.1097/00004872-200212000-00020 12473865 Pedersen L. R. Olsen R. H. Jurs A. Astrup A. Chabanova E. Simonsen L. . (2015 ). A randomised trial comparing weight loss with aerobic exercise in overweight individuals with coronary artery disease: the CUT-IT trial . Eur. J. Prev. Cardiol. 22 , 1009 –1017 . 10.1177/2047487314545280 25082954 Rider O. J. Tayal U. Francis J. M. Ali M. K. Robinson M. R. Byrne J. P. . (2010 ). The effect of obesity and weight loss on aortic pulse wave velocity as assessed by magnetic resonance imaging . Obesity (Silver. Spring). 18 , 2311 –2316 . 10.1038/oby.2010.64 20360756 Ring M. Eriksson M. J. Zierath J. R. Caidahl K. (2014 ). Arterial stiffness estimation in healthy subjects: a validation of oscillometric (Arteriograph) and tonometric (SphygmoCor) techniques . Hypertens. Res. 37 , 999 –1007 . 10.1038/hr.2014.115 25056681 Song X. Kestin M. Schwarz Y. Yang P. Hu X. Lampe J. W. . (2016 ). A low-fat high-carbohydrate diet reduces plasma total adiponectin concentrations compared to a moderate-fat diet with no impact on biomarkers of systemic inflammation in a randomized controlled feeding study . Eur. J. Nutr . 55 , 237 –246 . 10.1007/s00394-015-0841-1 25648736 Stevens V. J. Obarzanek E. Cook N. R. Lee I. M. Appel L. J. Smith West D. . (2001 ). Long-term weight loss and changes in blood pressure: results of the Trials of Hypertension Prevention, phase II . Ann. Intern. Med. 134 , 1 –11 . 10.7326/0003-4819-134-1-200101020-00007 11187414 Sugawara J. Hayashi K. Yokoi T. Tanaka H. (2008 ). Age-associated elongation of the ascending aorta in adults . JACC Cardiovasc. Imaging 1 , 739 –748 . 10.1016/j.jcmg.2008.06.010 19356510 Tinsley G. M. La Bounty P. M. (2015 ). Effects of intermittent fasting on body composition and clinical health markers in humans . Nutr. Rev. 73 , 661 –674 . 10.1093/nutrit/nuv041 26374764 Trumbo P. Schlicker S. Yates A. A. Poos M. (2002 ). Dietary reference intakes for energy, carbohydrate, fiber, fat, fatty acids, cholesterol, protein and amino acids . J. Am. Diet. Assoc. 102 , 1621 –1630 . 10.1016/S0002-8223(02)90346-9 12449285 Van Gaal L. F. Mertens I. L. Ballaux D. (2005 ). What is the relationship between risk factor reduction and degree of weight loss? Eur. Heart J. Suppl. 7 , L21 –L26 . 10.1093/eurheartj/sui082 Van Gaal L. F. Mertens I. L. De Block C. E. (2006 ). Mechanisms linking obesity with cardiovascular disease . Nature 444 , 875 –880 . 10.1038/nature05487 17167476 Varady K. A. (2011 ). Intermittent versus daily calorie restriction: which diet regimen is more effective for weight loss? Obes. Rev. 12 , E593 –E601 . 10.1111/j.1467-789X.2011.00873.x 21410865 Vlachopoulos C. Aznaouridis K. Stefanadis C. (2010 ). Prediction of cardiovascular events and all-cause mortality with arterial stiffness: a systematic review and meta-analysis . J. Am. Coll. Cardiol. 55 , 1318 –1327 . 10.1016/j.jacc.2009.10.061 20338492 Westerterp-Plantenga M. S. Nieuwenhuizen A. Tome D. Soenen S. Westerterp K. R. (2009 ). Dietary protein, weight loss, and weight maintenance . Annu. Rev. Nutr. 29 , 21 –41 . 10.1146/annurev-nutr-080508-141056 19400750 Wildman R. P. Farhat G. N. Patel A. S. Mackey R. H. Brockwell S. Thompson T. . (2005 ). Weight change is associated with change in arterial stiffness among healthy young adults . Hypertension 45 , 187 –192 . 10.1161/01.HYP.0000152200.10578.5d 15596570 Wilkinson I. B. Maccallum H. Rooijmans D. F. Murray G. D. Cockcroft J. R. Mcknight J. A. . (2000 ). Increased augmentation index and systolic stress in type 1 diabetes mellitus . QJM 93 , 441 –448 . 10.1093/qjmed/93.7.441 10874053 Willum-Hansen T. Staessen J. A. Torp-Pedersen C. Rasmussen S. Thijs L. Ibsen H. . (2006 ). Prognostic value of aortic pulse wave velocity as index of arterial stiffness in the general population . Circulation 113 , 664 –670 . 10.1161/CIRCULATIONAHA.105.579342 16461839 Wilson P. W. D'agostino R. B. Sullivan L. Parise H. Kannel W. B. (2002 ). Overweight and obesity as determinants of cardiovascular risk: the Framingham experience . Arch. Intern. Med. 162 , 1867 –1872 . 10.1001/archinte.162.16.1867 12196085 Wing R. R. Lang W. Wadden T. A. Safford M. Knowler W. C. Bertoni A. G. . (2011 ). Benefits of modest weight loss in improving cardiovascular risk factors in overweight and obese individuals with type 2 diabetes . Diabetes Care 34 , 1481 –1486 . 10.2337/dc10-2415 21593294 Zhu X. Zuo L. (2013 ). Characterization of oxygen radical formation mechanism at early cardiac ischemia . Cell Death Dis. 4 :e787 . 10.1038/cddis.2013.313 24008731 Zuo L. Roberts W. J. Tolomello R. C. Goins A. T. (2013 ). Ischemic and hypoxic preconditioning protect cardiac muscles via intracellular ROS signaling . Front. Biol. 8 , 305 –311 . 10.1007/s11515-012-1225-z Zuo L. Youtz D. J. Wold L. E. (2011 ). Particulate matter exposure exacerbates high glucose-induced cardiomyocyte dysfunction through ROS generation . PLoS ONE 6 :e23116 . 10.1371/journal.pone.0023116 21850256

PMC005xxxxxx/PMC5002413.txt

==== Front Front Plant SciFront Plant SciFront. Plant Sci.Frontiers in Plant Science1664-462XFrontiers Media S.A. 10.3389/fpls.2016.01265Plant ScienceEditorialEditorial: Multiple Facets of H+-Pyrophosphatase and Related Enzymes Ferjani Ali 1*Maeshima Masayoshi 2*1Department of Biology, Tokyo Gakugei UniversityKoganei, Japan2Laboratory of Cell Dynamics, Graduate School of Bioagricultural Sciences, Nagoya UniversityNagoya, JapanEdited by: Simon Gilroy, University of Wisconsin-Madison, USA Reviewed by: Kendal Hirschi, Baylor College of Medicine, USA *Correspondence: Ali Ferjani ferjani@u-gakugei.ac.jpMasayoshi Maeshima maeshima@agr.nagoya-u.ac.jpThis article was submitted to Plant Cell Biology, a section of the journal Frontiers in Plant Science 29 8 2016 2016 7 126501 7 2016 09 8 2016 Copyright © 2016 Ferjani and Maeshima.2016Ferjani and MaeshimaThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.The Editorial on the Research Topic Multiple Facets of H+-Pyrophosphatase and Related Enzymes H+-pyrophosphatasepyrophosphateindolebutyric acid (IBA)Enoyl-CoA Hydratase 2uncoupling H+-PPase variantscompensated cell enlargementnitrogen metabolismamine fungicides ==== Body Nearly 200 anabolic reactions liberate pyrophosphate (PPi), a byproduct of NTP hydrolysis (Kornberg, 1962; Maeshima, 2000; Heinonen, 2001; Ferjani et al., 2014a,b). In living cells, PPi must be hydrolyzed to orthophosphate by pyrophosphatase (PPase), because it suppresses the above reactions. PPases fall into two major classes, soluble PPases (sPPases) and membrane-bound H+-PPases (H+-PPases). In plants, vacuolar H+-PPase uses the energy released by PPi hydrolysis to acidify the vacuole, and its activity is particularly high in young tissues (Martinoia et al., 2007). Nevertheless, the physiological roles of PPases remain unclear due to severe phenotypes in loss-of-function mutants in various organisms (Ferjani et al., 2011 and references therein). Due to the importance of PPi homeostasis for life, this paper presents the most recent findings in this field and discusses the present situation along with future directions. The Arabidopsis vacuolar H+-PPase AVP1 has been extensively characterized. Under physiological conditions, AVP1 is specifically localized to the tonoplast (Segami et al., 2014). It is also localized to the plasma membrane in sieve element companion cells, and upon overexpression in other cell types (Pizzio et al., 2015 and references therein; Khadilkar et al., 2016). In addition to its role in maintaining vacuolar pH, AVP1 controls auxin transport and, consequently, auxin-dependent development, based on avp1-1 mutant analyses (Li et al., 2005). Recent studies using fugu5 mutant alleles favored a role for AVP1 in PPi removal and did not confirm auxin-related phenotypes (Ferjani et al., 2011, 2014a,b; Bertoni, 2011). Genome sequencing revealed that a second T-DNA insertion in GNOM, which is required for auxin polar transport, is responsible for the avp1-1 phenotype (Kriegel et al., 2015). Most importantly, Kriegel et al. (2015) further demonstrated that even plants lacking both tonoplast proton pumps (V-ATPase and AVP1) are viable and have significantly acidified vacuoles, highlighting for the first time the role of the TGN/EE-localized V-ATPase in vacuolar pH. More recently, a unique approach adopted by Asaoka et al. allowed the independent evaluation of two AVP1 functions. First, they showed that overexpression of H+-PPase is not correlated with increased biomass (Asaoka et al., 2016), in agreement with Kriegel et al. (2015). Second, they produced a number of single-residue mutagenized H+-PPases, and successfully identified uncoupling mutant variants that retained PPi-hydrolyzing capacity, but failed to translocate protons. Interestingly, even such uncoupling variants complemented the developmental defects in fugu5, reinforcing the importance of PPi homeostasis in plant development. This timely contribution elegantly supports the importance of H+-PPase-mediated PPi hydrolysis in planta. Recently, a curious relationship between nitrogen metabolism and PPi has also emerged. In fact, local cell death at the leaf petiole-blade junction was consistently observed in H+-PPase loss-of-function alleles upon growth on ammonia-free media (Fukuda et al.). Such phenotypes were totally suppressed when PPi was specifically removed by the yeast sPPase IPP1, by uncoupling H+-PPase variants, or by the simple addition of ammonium. Thus, it appears that high intracellular Pi levels somehow inhibit PPi hydrolysis, creating a cellular environment in which actively dividing cells can hardly cope. In Arabidopsis, efficient usage of seed storage lipids is crucial for seedling establishment until the acquisition of photoautotrophy. Lack of H+-PPase in fugu5 increases cytosolic PPi levels, partially reduces de novo sucrose synthesis, and inhibits cell division. In contrast, post-mitotic cell expansion in cotyledons is unusually enhanced, a phenotype called “compensation.” Therefore, it appears that PPi inhibits several cellular functions, including cell cycling, and it triggers compensated cell enlargement (CCE), whose mechanism remains elusive. Katano et al. reported that mutations in enoyl-CoA hydratase 2 (ECH2) significantly and specifically reduced CCE in the fugu5 background. Together, defects in either H+-PPase or ECH2 reduce cell proliferation due to reduced seed storage lipid mobilization, but ECH2 alone likely promotes post-mitotic cell expansion in cotyledons, probably through the conversion of indolebutyric acid (IBA) to indole acetic acid (IAA). This provides a strong genetic and phenotypic basis to propose that the phytohormone auxin promotes CCE in fugu5, but not in other classes of compensation. ECH2 has a dual function, i.e., seed oil reserve mobilization and IBA-to-IAA conversion. Therefore, future studies using mutants with defects only in IBA-to-IAA conversion (Strader et al., 2010) should clarify the above issue. Fungal pests are a major biological threat since they include the largest number of plant pathogens. While plant vacuoles have a dual set of proton pumps (V-ATPase and H+-PPase), fungi possess only V-ATPase. Amine fungicides are believed to inhibit postlanosterol sterol biosynthesis in fungi resulting in the accumulation of toxic abnormal sterols. Abnormal sterols affect the H+-pumping capacity of V-ATPases in fungi, and this has been proposed as a major determinant of fungicide action. Using yeast as a model fungus, Hernández et al. showed that amine fungicide treatment induced cell death by apoptosis, that apoptosis was concomitant with impaired H+-pumping capacity in vacuolar membrane vesicles dependent on vacuolar proteases, and that heterologous expression of Arabidopsis AVP1 in yeast cells increased resistance to amine fungicides. This paper challenged a long-standing issue in plant biology, and suggested that vacuolar H+-PPase is a major determinant of plant tolerance to amine fungicides. Since the discovery of PPi in rat liver extracts in 1941 (see Cori et al., 1951), a large amount of work has aimed to uncover the biological roles of PPi-hydrolyzing enzymes. These discoveries do not herald the end of the H+-PPase road, which has become more complex. Now that there are genuine H+-PPase mutations, we should maintain a broad enough vision to understand H+-PPase contributions in a wide range of organisms, but simultaneously narrow enough to uncover its detailed molecular and physiological functions. H+-PPase overexpression is beneficial under stressful conditions (Gaxiola et al., 2016 and references therein), yet how stress tolerance is increased needs to be mechanistically elucidated. The four papers published on this research topic have deepened our understanding of plant vacuolar H+-PPase and identified new challenges that should be addressed in the future. Author contributions AF and MM wrote and approved the final manuscript. Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. This work was supported by a Grant-in-Aid for Encouragement of Young Scientists (B) (21770036 to AF); Grant-in-Aid for Scientific Research (B) (16H04803 to AF); and Grant-in-Aid for Scientific Research on Innovative Areas (to AF and Hirokazu Tsukaya). The contribution to the fugu5 related research project of all past and present members of Tsukaya lab. (The University of Tokyo), Horiguchi lab. (Rikkyo University), Maeshima lab. (Nagoya University) and Ferjani lab. (Tokyo Gakugei University) is gratefully acknowledged. ==== Refs References Bertoni G. (2011 ). A surprising role for vacuolar pyrophosphatase . Plant Cell 23 , 2808 . 10.1105/tpc.111.230813 21862706 Cori G. T. Ochoa S. Slein M. W. Cori C. F. (1951 ). The metabolism of fructose in liver; isolation offructose-1-phosphate and inorganic pyrophosphate . Biochim. Biophys. Acta 7 , 304 –317 . 10.1016/0006-3002(51)90032-7 14858418 Ferjani A. Segami S. Asaoka M. Maeshima M. (2014a ). Regulation of PPi levels through vacuolar membrane H+-pyrophosphatase , in Progress in Botany , Vol. 75 , eds Lüttge U. Beyschlag W. Cushman J. (Heidelberg : Springer-Verlag ), 145 –166 . Ferjani A. Segami S. Horiguchi G. Muto Y. Maeshima M. Tsukaya H. (2011 ). Keep an eye on PPi: the vacuolar-type H+-pyrophosphatase regulates postgerminative development in Arabidopsis . Plant Cell 23 , 2895 –2908 . 10.1105/tpc.111.085415 21862707 Ferjani A. Segami S. Horiguchi G. Muto Y. Maeshima M. Tsukaya H. (2014b ). Roles of the vacuolar H+-PPase in seed storage oil mobilization and plant development . Plant Morphol. 26 , 45 –51 . 10.5685/plmorphol.26.45 Gaxiola R. A. Regmi K. Hirschi K. D. (2016 ). Moving on up: H+-PPase mediated crop improvement . Trends Biotechnol. 34 , 347 –349 . 10.1016/j.tibtech.2015.12.016 26818803 Heinonen J. K. (2001 ). Biological Role of Inorganic Pyrophosphate . Boston, MA : Kluwer Academic Publishers . Khadilkar A. S. Yadav U. P. Salazar C. Shulaev V. Paez-Valencia J. Pizzio G. A. . (2016 ). Constitutive and companion cell-specific overexpression of AVP1, encoding a proton-pumping pyrophosphatase, enhances biomass accumulation, phloem loading, and long-distance transport . Plant Physiol. 170 , 401 –414 . 10.1104/pp.15.01409 26530315 Kornberg A. (1962 ). On the metabolic significance of phosphorolytic and pyrophosphorolytic reactions , in Horizons in Biochemistry , eds Kasha H. Pullman B. (New York, NY : Academic Press ), 251 –264 . Kriegel A. Andrés Z. Medzihradszky A. Krüger F. Scholl S. Delang S. . (2015 ). Job sharing in the endomembrane system: Vacuolar acidification requires the combined activity of V-ATPase and V-PPase . Plant Cell 27 , 3383 –3396 . 10.1105/tpc.15.00733 26589552 Li J. Yang H. Peer W. A. Richter G. Blakeslee J. Bandyopadhyay A. . (2005 ). Arabidopsis H+-PPase AVP1 regulates auxin-mediated organ development . Science 310 , 121 –125 . 10.1126/science.1115711 16210544 Maeshima M. (2000 ). Vacuolar H+-pyrophosphatase . Biochim. Biophys. Acta 1465 , 37 –51 . 10.1016/s0005-2736(00)00130-9 10748246 Martinoia E. Maeshima M. Neuhaus H. E. (2007 ). Vacuolar transporters and their essential role in plant metabolism . J. Exp. Bot. 58 , 83 –102 . 10.1093/jxb/erl183 17110589 Pizzio G. A. Paez-Valencia J. Khadilkar A. S. Regmi K. Patron-Soberano A. Zhang S. . (2015 ). Arabidopsis type I proton-pumping pyrophosphatase expresses strongly in phloem, where it is required for pyrophosphate metabolism and photosynthate partitioning . Plant Physiol. 167 , 1541 –1453 . 10.1104/pp.114.254342 25681328 Segami S. Makino S. Miyake A. Asaoka M. Maeshima M. (2014 ). Dynamics of vacuoles and H+-pyrophosphatase visualized by monomeric green fluorescent protein in Arabidopsis: artifactual bulbs and native intravacuolar spherical structures . Plant Cell 26 , 3416 –3434 . 10.1105/tpc.114.127571 25118245 Strader L. C. Culler A. H. Cohen J. D. Bartel B. (2010 ). Conversion of endogenous indole-3-butyric acid to indole-3-acetic acid drives cell expansion in Arabidopsis seedlings . Plant Physiol. 153 , 1577 –1586 . 10.1104/pp.110.157461 20562230

PMC005xxxxxx/PMC5002414.txt

==== Front Front PsycholFront PsycholFront. Psychol.Frontiers in Psychology1664-1078Frontiers Media S.A. 10.3389/fpsyg.2016.01303PsychologyOriginal ResearchTask-Based and Questionnaire Measures of Inhibitory Control Are Differentially Affected by Acute Food Restriction and by Motivationally Salient Food Stimuli in Healthy Adults Bartholdy Savani 1*Cheng Jiumu 2Schmidt Ulrike 1Campbell Iain C. 1O'Daly Owen G. 31Section of Eating Disorders, Department of Psychological Medicine, Institute of Psychiatry, Psychology and Neuroscience, King's College LondonLondon, UK2Department of Neuroscience, Institute of Psychiatry, Psychology and Neuroscience, King's College LondonLondon, UK3Centre for Neuroimaging Sciences, Institute of Psychiatry, Psychology and Neuroscience, King's College LondonLondon, UKEdited by: Peter Hall, University of Waterloo, Canada Reviewed by: Laurence James Nolan, Wagner College, USA; Siegfried Dewitte, KU Leuven, Belgium *Correspondence: Savani Bartholdy savani.bartholdy@kcl.ac.ukThis article was submitted to Eating Behavior, a section of the journal Frontiers in Psychology 29 8 2016 2016 7 130327 5 2016 15 8 2016 Copyright © 2016 Bartholdy, Cheng, Schmidt, Campbell and O'Daly.2016Bartholdy, Cheng, Schmidt, Campbell and O'DalyThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.Adaptive eating behaviors are dependent on an interaction between motivational states (e.g., hunger) and the ability to control one's own behavior (inhibitory control). Indeed, behavioral paradigms are emerging that seek to train inhibitory control to improve eating behavior. However, inhibitory control is a multifaceted concept, and it is not yet clear how different types (e.g., reactive motor inhibition, proactive motor inhibition, reward-related inhibition) are affected by hunger. Such knowledge will provide insight into the contexts in which behavioral training paradigms would be most effective. The present study explored the impact of promoting a “need” state (hunger) together with motivationally salient distracting stimuli (food/non-food images) on inhibitory control in 46 healthy adults. Participants attended two study sessions, once after eating breakfast as usual and once after acute food restriction on the morning of the session. In each session, participants completed questionnaires on hunger, mood and inhibitory control, and undertook task-based measures of inhibitory control, and had physiological measurements (height, weight, and blood glucose) obtained by a researcher. Acute food restriction influenced task-based assessments but not questionnaire measures of inhibitory control, suggesting that hunger affects observable behavioral control but not self-reported inhibitory control. After acute food restriction, participants showed greater temporal discounting (devaluation of future rewards), and subjective hunger and these were inversely correlated with stop accuracy on the stop signal task. Finally, participants generally responded faster when food-related distractor images were presented, compared to non-food images, independent of state. This suggests that although food stimuli motivate approach behavior, stimulus relevance does not impact inhibitory control in healthy individuals, nor interact with motivational state. These findings may provide some explanation for poorer inhibitory control often reported in studies of individuals who practice restraint over eating. inhibitory controltemporal discountingfoodfedfastedNational Institute for Health Research10.13039/501100000272 ==== Body Introduction We live in an obesogenic environment. Thus, we need to be able to control our eating behavior and overcome the temptation toward unhealthy foods that are often readily accessible, or to avoid non-homeostatic eating (i.e., eating in the absence of a physiological energy deficit). The ability to adjust our behavior to adapt to our environment depends on the ability to stop/withhold inappropriate behaviors (broadly termed “inhibitory control”), as well as the ability to determine the salience and importance of environmental cues. This includes approaching items that fulfill a basic need or have rewarding properties (e.g., food), but only when this is contextually appropriate (i.e., when hungry compared to when satiated). Highly palatable and often calorific foods are easily accessible and the ability to exercise inhibitory control is crucial to resisting them. Inhibitory control is a multifaceted concept, and different types of inhibitory control may be relevant in such a context. For example, one needs to exercise proactive and reactive motor response inhibition (i.e., withholding a motor response in the context of uncertainty, or in reaction to a stop signal, respectively) to stop from purchasing and/or eating such foods. Similarly, reward-related inhibition (i.e., waiting for a larger delayed reward rather than choosing immediate gratification) is required to overcome the temptation of, for example, a highly calorific snack and wait to eat a more substantial meal when hungry. Thus, adaptive eating behaviors are dependent on an interaction between motivational/physiological states (e.g., hunger) and inhibitory control. Indeed, neuroimaging studies have revealed that brain activity in response to food stimuli differs according to hunger state, with satiety associated with a relatively reduced response in reward-related regions and an enhanced response in regions implicated in executive control in healthy individuals compared to a pre-meal state (Thomas et al., 2015), suggesting that the neural correlates of inhibitory control in the context of food is associated with hunger state. Altered inhibitory control has been proposed as a key factor in the aetiopathology of eating disorders and obesity (e.g., Brooks et al., 2012; Thamotharan et al., 2013; Berner and Marsh, 2014; Wierenga et al., 2014). There is evidence of reduced reward-related inhibitory control in individuals with bulimia nervosa, binge eating disorder and obesity (Davis et al., 2010; Manwaring et al., 2011; Galimberti et al., 2012; Mole et al., 2015; Kekic et al., 2016), and the opposite in anorexia nervosa (Steinglass et al., 2012; Decker et al., 2015) compared to healthy individuals (for review, see McClelland et al., In prep.). Similar findings have also been found with respect to reactive inhibition, although the findings have been less consistent than data on reward-related inhibition (for review, see Bartholdy et al., 2016). As a result, behavioral paradigms are emerging that seek to increase inhibitory control to improve eating behavior (Lawrence et al., 2015; for reviews, see Brockmeyer et al., 2015; Bartholdy et al., 2016; Turton et al., 2016), which may be useful both in the treatment of eating disorders/obesity, but also as possible preventative measures for individuals in the community. However, the relationships between hunger and the different elements of inhibitory control are not yet clear. For example, does being in a state of hunger affect all elements of inhibitory control affected in the same way? This information will be important in determining the contexts in which such training should be provided to improve each of the types of inhibitory control. There have been some studies exploring the relationship between hunger and inhibitory control. A relationship between hunger state and inhibitory control has also been reported in behavioral experiments. Nordgren et al. (2009) reported that hungry individuals had a weaker belief in their ability to control their impulses than satiated individuals and exposed themselves to less temptation (preferred snacks), argued to be a result of the weakened beliefs in their impulse-control abilities. In healthy overweight and obese women, temporal discounting (i.e., the devaluation of rewards over time) and hedonic hunger (i.e., a preoccupation with palatable foods or desire to eat for pleasure in the absence of a physical energy deficit; Witt et al., 2014; Manasse et al., 2015) were found to interact to predict food intake only in individuals high in temporal discounting (Appelhans et al., 2011; Manasse et al., 2015), suggesting that a greater ability to delay reward (i.e., lower temporal discounting) may be protective of overeating. Moreover, there is evidence that in healthy individuals, being hungry is associated with reduced temporal discounting but only for large rewards (De Ridder et al., 2014), which the authors argued may be due to increased reliance on emotion or intuition in decision making when in a motivational physical state. Reactive motor response inhibition in healthy individuals, however, does not appear to be influenced by hunger: one study reported no correlation between inhibitory control and hunger (Haynes et al., 2015), and another reported that the effect of inhibitory control on food intake was not affected by hunger (before/after lunch; Nederkoorn et al., 2015). However, reactive motor response inhibition has been reported to interact with hunger, with healthy individuals who are both hungry and more impulsive (i.e., have less inhibitory control) consuming the greatest number of calories in a subsequent taste test, and purchased more snack calories in a virtual supermarket, compared to participants who were either hungry or impulsive, separately (Nederkoorn et al., 2009). If hunger and inhibitory control interact to influence food intake, then it can be predicted that training paradigms aimed at improving inhibitory control would be most beneficial if completed in the context of hunger. However, while the above studies imply a relationship between hunger and inhibitory control, the employment of between-subjects designs only permits cross-sectional inferences to be drawn, as each participant was only assessed once at a particular level of hunger. It is important to replicate these findings using repeated measures designs to determine how hunger can affect inhibitory control within the same individual. This will improve our understanding of why some individuals may experience difficulty in exercising control over their eating behavior, and will provide insight into the contexts in which behavioral interventions targeting inhibitory control over eating would be most effective. In relation to contextual effects, it is important to explore whether motivationally salient stimuli additionally affect one's ability to exercise inhibitory control. For example, one study found that response times on a simple reaction time task correlated with a greater desire to eat when participants were asked to simultaneously imagine their favorite food, but this relationship was not observed when participants imagined a non-food-related scenario (Green et al., 2000). Food is a primary reinforcer, is typically appetitive and generates an approach-motivated emotional state (Rolls, 1999; Uher et al., 2014). Motivational salience attributed to food stimuli is easily manipulated by asking participants to abstain from eating. Thus, in a food deprived (fasted) state, food appears more pleasant, elicits less disgust, facilitates approach reactions, and is associated with enhanced startle reflexes and increased heart rate (Drobes et al., 2001; Seibt et al., 2007; Hoefling et al., 2009). In addition, increased attentional bias toward food has been demonstrated in people who are hungry/fasted (Castellanos et al., 2009) or are sensitive to external food cues (Brignell et al., 2009). Motivational state (i.e., hunger vs. satiety) may therefore modulate the personal relevance of food stimuli and have an impact on the salience of state-relevant cues (Oliveira et al., 2013). This present study investigated the influence of promoting a “need” state (hunger) and motivationally relevant distraction (food images) on several aspects of inhibitory control. Specifically, it examined the impact of acute food-restriction on questionnaire and task-based measures of inhibitory control, including proactive, reactive, and reward-based inhibition, using a repeated-measures design. In addition, the motivational relevance of distracting stimuli (food vs. non-food images) on proactive inhibition and reactive inhibition was investigated. It was broadly hypothesized that (a) acute food restriction would impair inhibitory control, (b) the presence of food stimuli compared to neutral (non-food) stimuli would promote “approach” behaviors resulting in poorer inhibitory control, and (c) a hunger state and the motivational relevance of the food images would interact, with the poorest inhibitory control observed on trials including food images in the fasted state. Materials and methods Participants Fifty seven healthy adults were recruited from King's College London for the study (31 students, 14 employed staff, 1 unemployed). Four participants met DSM-IV criteria for an eating disorder on the Eating Disorder Diagnostic Scale (Stice et al., 2000) and were therefore excluded from the analysis. One participant did not follow the study instructions during the first session so did not complete the second session. An estimate of weight and height was self-reported by participants during screening, however this was discrepant with the weight obtained during the study, with three participants being underweight (BMI < 18.5 kg/m2) and four participants being overweight (BMI > 25 kg/m2). As weight status may affect inhibitory control (Bartholdy et al., 2016), these individuals were excluded from analyses for a more homogeneous sample. The final sample consisted of 46 participants (39 women), with a mean age of 24.9 years (SD = 6.62 years, range = 18–49 years) and mean BMI (average across both sessions) of 21.6 kg/m2 (SD = 1.69 kg/m2). The majority (87.0%) of participants were right handed. Exclusion criteria included being under 18 years of age, a past or current axis I mental disorder, neurological disease, history of head trauma with loss of consciousness, current use of psychotropic medication, current use of illicit drugs, and drinking on average >18 units of alcohol per week. Computer tasks Temporal discounting task TD refers to the tendency of people to discount (reduce the magnitude of) the value of future rewards with increasing distance from the present. This was assessed using a hypothetical monetary choice task, in which participants were asked to indicate their preference between a smaller immediate reward (smaller-sooner reward) and a larger delayed reward (larger-later reward) on 80 binary choices, modeled on a paradigm developed by Steinglass et al. (2012). A monetary choice task was used to explore the degree to which any impact of hunger on inhibitory control generalized outside of the context of food. Greater discounting refers to an increased preference for smaller-sooner rewards. The rewards were presented using two decision frames: an Accelerate frame in which the smaller-sooner reward varied between £20 and £98 in £2 increments (40 choices) while the larger-later reward remained fixed at £100, and a Delay frame in which the smaller-sooner reward remained at £50 while the larger-later reward varied between £52 and £130 in £2 increments (40 choices). The larger-later delay was 3 months (0.25 years) for all choices. The paradigm involved a single block including all 80 trials, presented in a randomized order, however the order of trials was kept consistent across all participants. Discount rate was determined by calculating participants' discount factor, i.e., the degree to which the present subjective value of a future reward is reduced. This was calculated for each set (Accelerate, Delay) and a global DF was calculated as the mean discount factor of the two sets. Discount factors were quantified using a two-step procedure (Steinglass et al., 2012). The first step establishes the “indifference point” for that set, i.e., the point at which choices are considered equivalent. This is identified as the choice where the participant's preference switches from larger-later to smaller-sooner in the Accelerate set, and from smaller-sooner to larger-later in the Delay set (Steinglass et al., 2012). Second, the following formula is fit to the indifference point to calculate the discount factor: δ = (x1/x2)(1/(t2−t1)), where x1 is the smaller-sooner reward, x2 is the larger-later reward, and t2 − t1 is the delay to reward presentation (in years), which was 0.25 in this study (Weber et al., 2007; Steinglass et al., 2012). This procedure is a sensitive measure of temporal discounting that is independent of hyperbolic modeling and area under the curve analyses (Weber et al., 2007). The discount factor value ranges from 0 to 1, with smaller numbers indicating greater temporal discounting and thus a greater tendency to choose the smaller-sooner reward. In the case that no switch was made (i.e., participants consistently chose either the smaller-sooner or larger-later reward across all choices), the indifference point was unable to be calculated and therefore a respective default score of 0 (consistent smaller-sooner selection) or 1 (consistent larger-later selection) was assigned. Modified proactive inhibition task (PI) This task was a modified version of a simple cued-reaction time task based on previous paradigm designs used to assess proactive inhibition (Jaffard et al., 2008), in which the delay between “warning” cue and target onset was varied across trials. An image flanked by two empty boxes was presented on a computer screen. On each trial, a target (large yellow dot) appeared in one of the two boxes. Participants are instructed to react as quickly as possible to the location of the visual target by pressing the corresponding arrow key. On some trials, the target was preceded by a warning signal (“cued trials” compared to “non-cued trials”), which appeared in both boxes simultaneously and was consequently spatially uninformative. This study employed a blocked design. The image in the center of the screen changed on each trial, and consisted of either a high/low calorie food image, a neutral non-food image (household items such as stationary), or a fixation cross. The order of image presentation was randomized, with an equal number of each image type presented. The stimulus onset asynchrony was manipulated so that the cue-target delay (stimulus-onset asynchrony; SOA) varied randomly across four conditions: 0 (no cue), 100, 300, and 500 ms. This study was conducted in three blocks: one “pure” block (only non-cued trials) and two mixed blocks in which cued- and non-cued trials were presented randomly. Participants completed a practice block of 18 trials. Participants then completed three experimental blocks in a randomized, counterbalanced order. One block consisted only of non-cued trials (“pure” block; 63 trials). The SOA did not vary in the pure block. The remaining two blocks included a mixture of non-cued (0 ms SOA) and cued trials (at varying SOAs: 100, 300, 500 ms). A total of 63 non-cued trials and 63 cued trials (21 trials at each SOA) were presented in a random order across the mixed blocks. The playlists for each block were the same for all participants in all sessions. Across all blocks, the inter-trial interval was 3000 ms, cue duration was 300 ms and target onset (from start of trial) was 2000 ms. Two behavioral indices were calculated for entry into correlational analyses. The first was an index of the overall benefit gained from warning cues (“Warning benefit”), calculated by subtracting the mean reaction time (RT) on trials with the longest delay between cue and target (500 ms SOA) from the non-cued trials (0 ms SOA in the mixed blocks). The second was an index of proactive slowing in the context of uncertainty, calculated by subtracting the mean RT on trials in the pure block from non-cued trials in the mixed block (“Preparation cost”; Chikazoe et al., 2009). Faster reaction time during the pure block compared to 0 ms SOA trials in the mixed blocks should indicate that participants implicitly recognized a difference in the probability that the visual stimulus was the target, i.e., their threshold for responding is altered. Modified stop signal task The stop signal task is a reaction time task in which participants are asked to respond to one or more “go” stimuli. On each trial, participants were presented with a go signal in the form of a left- or right-pointing arrow in the direction of the required response. On 20% of trials, a stop signal (red dot) was presented at irregular intervals in order to minimize predictability. Following the presentation of this stop signal, participants were required to inhibit their response to the go signal. The delay between the go signal and stop signal is termed the stop signal delay (SSD). The ability to inhibit a response is dependent on the length of the SSD: the longer the delay, the harder it is to inhibit a response. The SSD was varied from trial to trial by 50 ms increments (ranging between 150 and 900 ms) in a staircase procedure, which intended to converge subjects toward an overall performance of 50% for each run. The initial SSD was set at 150 ms. The primary outcome variable of the stop signal task is the stop signal reaction time (SSRT), which is considered an index of inhibitory control ability. Typically, this is calculated by subtracting the SSD from the mean RT on go trials. However, due to the variability in stop accuracy and the small number of trials in this design, the SSRT in the present study was calculated by subtracting the mean SSD on stop trials from the nth percentile of the correct go RT distribution, using the formula: SSRT = RT (m)−SSD, where m = n(number of correct go responses) * probability(responding|signal) (Nederkoorn et al., 2012). In other words, if participants correctly stopped on 25% of the trials, the mean SSD was subtracted from the 25th% reaction time on correct go trials [0.25*n(correct go trials)] to calculate the SSRT. This study employed a blocked design. Participants first completed a practice block of 20 trials including 8 stop trials. This was followed by three experimental blocks of 100 trials each (20 stop trials): two blocks where irrelevant distractor stimuli (one block with high/low calorie food images, one block with neutral non-food images [e.g., household objects]) were presented beside the arrow (on the same side of the screen as the arrow was pointing to minimize congruency-related interference effects), and one block with no additional stimuli (no image condition). The blocks were presented in a randomized, counterbalanced order across participants, and the order of stop and go trials within each block was randomized. The playlists for each block were the same for all participants in all sessions. The arrow pointed to the left and right on an equal number of trials. On all trials, the arrow display duration (unless cut short by a stop signal) was 400 ms, stop signal duration was 300 ms. The maximum RT from the onset of the arrow was 1400 ms. The inter-trial interval was fixed at 1800 ms. Stimuli The food and non-food images were taken from an in-house image battery. standardized for size, resolution, color luminance, and complexity (Karra et al., 2013). State manipulation (fed/fasted) As has been done previously (e.g., Karra et al., 2013; Chechko et al., 2015), hunger was manipulated via an overnight fast. Adherence to the overnight fast was assessed via open-ended questions regarding time since last eaten (including drinks), what food/drink was last eaten, and the time since last consumption of a caffeinated beverage (Witt et al., 2014). A small blood sample was obtained through a finger prick to assess the difference in blood glucose in the fed and fasted state, as used by patients with diabetes for self-monitoring of blood glucose (for review, see Clarke and Foster, 2012). Procedure Individuals who expressed interest in the study completed a telephone screening to assess eligibility using a study-specific screening questionnaire developed by the researchers, including questions about their diet, substance use, caffeine intake, smoking, and history of neurological trauma and psychiatric disorder. Participants were asked to estimate their weight and height in order for the researchers to get an estimated BMI for assessment of eligibility. Eligible participants attended two experimental sessions occurring at the same time of day on 2 separate days spaced no more than 1 week apart. Participants were asked to fast (i.e., refrain from eating or drinking anything except water) on the morning of one of the study sessions and eat breakfast as usual for the other study session. The order of study conditions was randomized and counterbalanced across participants (54.7% fed and 45.3% fasted first condition). At both study sessions, participants were first asked to complete several questionnaires [Depression, Anxiety, and Stress Scale (Lovibond and Lovibond, 1995), Barratt Impulsiveness Scale (Patton et al., 1995), Delaying Gratification Inventory (Hoerger et al., 2011)]. Hunger was assessed using a 10 cm visual analog scale. Participants were also asked to state when they last ate, what their last meal consisted of, and when they next expected to eat. In the first session, participants additionally completed the Eating Disorder Examination Questionnaire (Fairburn, 2008) and the screening module of the Structured Clinical Interview for DSM-IV (First et al., 2002). Participants then engaged in three computerized tasks: the Temporal Discounting Task, the Stop Signal Task, and a simple cued-reaction time task assessing proactive inhibition. The order in which the tasks were completed was randomized and counterbalanced across participants and across sessions. After completing the tasks, the researcher measured the participants' height and weight, blood glucose level and temperature. Participants were thanked and received £20 compensation for their time and any additional reimbursement for travel. Ethics This study was reviewed and approved by the Psychiatry, Nursing, and Midwifery Research Ethics Subcommittee at King's College London (PNM/13/14-147). All participants gave written consent after the procedures were explained and were debriefed after the experiment. Data analysis Data was collected using in-house software and analyzed using SPSS and were corrected for multiple comparisons using Bonferroni correction. Statistical analyses were performed using IBM® SPSS® software (Version 22). All tests were two-tailed and the level of significance was set at α = 0.05. Questionnaires Inspection of histograms, skewness and kurtosis indicated that the majority of physiological and questionnaire data (hunger, temperature, time to next meal, and time since last meal) were skewed. Glucose, DGI total and physical subscales, and the BIS-11 attention subscale were the only questionnaire and physiological measures that did not show skewness or kurtosis in either the fed or fasted state. With respect to the physiological data, this was unsurprising: hunger and time since last meal were positively skewed in the fed state only, as participants were asked to ensure they had eaten breakfast as usual that day. Similarly, time to next meal in the fasted state was positively skewed. Thus, Wilcoxon Signed Rank tests were employed to compare questionnaire responses and physiological outcomes in the fed and fasted state. These tests were corrected for multiple comparisons using Bonferroni correction. Temporal discounting task While Global discount factor was normally distributed, Accelerate discount factor and Delay discount factor were positively skewed. Square root transformations were effective in normalizing the temporal discounting data. The main effects of and interaction between state and frame were analyzed using a repeated measures 2 × 2 (state × frame) ANOVA. Proactive inhibition task Gradual adjustment of behavioral inhibition was assessed using 2 × 3 × 4 ANOVA to evaluate the main effects and interactions between state (fed/fasted), stimulus (food/non-food/fixation cross) and SOA (0, 100, 300, 500 ms) on reaction time. The effect of uncertainty on non-cued trials was assessed using a 2 × 2 × 3 repeated measures to examine the main effects and interactions of block condition (pure vs. mixed), state and stimulus on reaction time. Post-hoc t-tests corrected for multiple comparisons using Bonferroni correction were employed to further assess statistically significant main effects and interactions. The 2 indices of proactive inhibition (warning benefit, preparation cost) were entered into correlational analyses (below). Stop signal task Mean RT and SSRT data from the stop signal task were also positively skewed. Square-root transformations were effective in normalizing the stop signal task data. Go accuracy could not be normalized through transformations due to almost perfect accuracy. Repeated measures 2 × 3 ANOVAs were conducted on the SSRT, mean RT and stop accuracy data to explore the main effects of and interactions between state (fed/fasted) and stimulus (food/non-food/no image). Statistically significant main effects were further explored using post-hoc t-tests corrected for multiple comparisons using Bonferroni correction. Correlations Correlations were conducted to assess the relationship between physiological measures, mood, and questionnaire and task-based measures of inhibitory control across states. In situations where a main effect of state is observed on both variables entered into the correlation, the relationship between the variables were assessed separately for the fed and fasted sessions due to the increased likelihood of spurious correlations associated with clustering due to state effects. Pearson's r and Spearman's rho tests were employed when data were parametric and non-parametric, respectively. Results State manipulation The length of fasting time in the fasted condition ranged from 8 to 17 h. To assess whether our manipulation of hunger state was successful, paired samples t-tests were conducted to examine differences in self-reported hunger, time since last meal and time to next meal and blood glucose in the fed compared to fasted state (Table 1). All comparisons were statistically significant after Bonferroni correction for multiple comparisons. Participants were significantly more hungry in the fasted condition [z(33) = 5.012, p < 0.001], had lower blood glucose levels [z(46) = −2.918, p = 0.004], had a longer time since last eating [z(42) = 5.639, p < 0.001] and a shorter intended waiting time before the next meal [z(42) = −3.921, p < 0.001] in the fasted compared to fed condition. Table 1 Differences in state measurements, questionnaire responses, and temporal discounting between fed and fasted states. Fed Mean(SD) Fasted Mean(SD) Wilcoxon z statistic N pa STATE MEASUREMENTS Hunger 0.2 (0.15) 0.6 (0.17) 5.012 33 <0.001 Glucose (mmol/liter) 5.1 (0.59) 4.8 (0.47) −2.918 46 0.004 Time since last meal (hours) 1.9 (1.80) 12.8 (2.00) 5.639 42 <0.001 Time to next meal (hours) 2.4 (0.95) 1.5 (0.70) −3.921 42 <0.001 DEPRESSION, ANXIETY, AND STRESS SCALE (DASS) Total 6.2 (4.90) 6.3 (4.94) 1.061 46 0.289 Depression 1.5 (1.52) 1.4 (1.50) −0.698 46 0.485 Anxiety 1.4 (1.89) 1.3 (1.48) 0.212 46 0.832 Stress 3.3 (2.50) 3.6 (3.08) 0.858 46 0.391 DELAYING GRATIFICATION INVENTORY (DGI) Total 136.4 (12.47) 134.5 (12.32) −1.717 45 0.086 Achievement 29.5 (3.85) 29.3 (4.33) −0.459 45 0.646 Food 24.9 (4.42) 24.5 (4.73) −0.657 45 0.511 Money 29.1 (4.38) 28.6 (4.37) −1.481 45 0.139 Physical 25.7 (3.64) 25.1 (3.95) −1.396 45 0.163 Social 27.3 (2.87) 27.0 (2.92) −0.783 45 0.434 BARRATT IMPULSIVENESS SCALE (BIS-11) Total 44.0 (9.22) 45.8 (8.27) 1.254 45 0.210 Attention 12.1 (2.86) 12.3 (2.80) 0.108 45 0.914 Motor 18.8 (4.08) 19.8 (3.49) 1.795 45 0.073 Non-planning 13.2 (4.08) 13.7 (3.82) 0.972 45 0.331 a Uncorrected p-values. Self-reported impulsivity and mood Differences in self-reported mood and impulsivity were assessed by comparing responses on the DASS-21 and on the BIS-11 and DGI questionnaires, respectively, in the fed and fasted state using paired samples t-tests. There was a trend for impulsivity assessed by the DGI questionnaire (DGI Total: z = −1.717, p = 0.086) and motor impulsivity on the BIS-11 (Motor subscale score: z = 1.795, p = 0.073) to be higher in the fasted state, however this was no longer observed after correction for multiple comparisons. The analyses did not yield any other significant differences between states (Table 1, all p > 0.193). Temporal discounting The impact of state and decision framing was assessed using a 2 × 2 repeated measures ANOVA using transformed discount factor scores (Figure 1). This analysis yielded a main effect of frame [F(1, 45) = 11.769, p = 0.001] and a trend toward a main effect of state [F(1, 45) = 3.587, p = 0.065], but no interaction between state and frame [F(1, 45) = 1.375, p = 0.247] (Figure 1). Post-hoc t-tests revealed the main effect of frame was driven by lower discount factors (indicating greater temporal discounting) in the delay frame (mean = 0.55, SD = 0.232) compared to the accelerate frame (mean = 0.61, SD = 0.262), t(45) = 3.341, p = 0.001. Discount factors were lower in the fasted (mean = 0.56, SD = 0.243) compared to the fed (mean = 0.60, SD = 0.255) state, however this trend did not survive correction for multiple comparisons, t(45) = 1.894, p = 0.065 (uncorrected). Figure 1 Discount factors for the accelerate and delay frame in the fed and fasted state. Error bars denote standard errors. Proactive inhibition This task was completed by 31 participants (27 women). Due to computer malfunctions during the pure block in the fed state for two female participants, these participants were excluded from analyses involving pure block data. Means and standard deviations for reaction times, accuracy and the behavioral indices of proactive inhibition for each trial type in each state are described in Table 2. Table 2 Mean (standard deviation) RT and accuracy for the non-cued trials (in the pure and mixed blocks) and for each SOA (in the mixed blocks), and mean behavioral index of proactive inhibition for each stimulus type in the fed and fasted state. SOA (ms) Fed Fasted Food Non-food Fixation Food Non-food Fixation RT (ms) Non-cued trials (Pure vs. 0ms; n = 29) Pure 347.71 (40.88) 357.72 (41.06) 371.95 (44.43) 352.98 (45.27) 363.84 (44.11) 380.45 (52.62) 0 396.08 (46.10) 392.97 (35.23) 409.50 (43.00) 405.14 (54.58) 411.06 (53.94) 423.38 (53.77) Mixed block SOA (n = 31) 0 396.82 (45.61) 393.05 (34.70) 410.94 (41.92) 403.25 (53.37) 409.10 (52.69) 421.54 (52.47) 100 359.92 (43.63) 357.01 (40.18) 370.25 (37.27) 362.84 (39.88) 374.89 (40.73) 375.68 (44.60) 300 364.34 (45.01) 371.17 (39.20) 374.14 (40.70) 374.47 (33.77) 375.83 (41.69) 387.42 (36.77) 500 333.19 (33.47) 328.77 (45.39) 329.19 (32.63) 337.07 (51.01) 339.33 (35.22) 342.86 (34.40) ACCURACY (%) Non-cued trials (Pure vs. 0ms; n = 29) Pure 99.01 (2.34) 98.69 (2.17) 98.03 (3.25) 99.01 (2.34) 99.01 (2.34) 98.85 (2.07) 0 99.06 (2.23) 98.97 (2.46) 99.67 (1.23) 99.37 (1.60) 98.62 (2.64) 98.85 (2.43) Mixed block SOA (n = 31) 0 99.12 (2.17) 99.03 (2.39) 99.69 (1.19) 99.41 (1.55) 98.71 (2.57) 98.92 (2.37) 100 94.01 (8.86) 97.24 (5.74) 97.24 (6.82) 94.01 (8.86) 95.85 (8.41) 97.7 (5.34) 300 98.62 (4.29) 96.77 (7.10) 100 (0.00) 99.08 (3.57) 97.7 (6.49) 99.54 (2.57) 500 95.85 (6.59) 98.16 (4.87) 97.7 (6.49) 97.7 (5.34) 98.16 (4.87) 96.77 (6.07) BEHAVIORAL INDEX (ms) Warning benefit 63.63 (38.43) 64.28 (33.85) 81.75 (29.38) 66.19 (45.65) 69.77 (37.86) 78.68 (45.78) Preparation cost 48.37 (36.88) 35.25 (26.88) 37.55 (30.28) 52.15 (38.63) 47.22 (33.21) 42.94 (30.11) Effect of uncertainty in cued trials: warning benefit A 2 × 3 × 4 ANOVA was employed to assess the main effects and interactions between SOA, state and stimulus on reaction times in the mixed blocks. This revealed a significant main effect of stimulus [F(1.677, 50.323*) = 17.277, p < 0.001] and SOA [F(2.047, 61.409*) = 126.083, p < 0.001], but no main effect of state [F(1, 30) = 2.728, p = 0.109]. No significant interactions were observed between state and stimulus [F(2, 60) = 1.981, p = 0.147], state and SOA [F(3, 90) = 0.071, p = 0.975], stimulus and SOA delay [F(5.399, 161.973) = 1.438, p = 0.209] or between all three factors [F(5.184, 155.5141) = 1.011, p = 0.415]. Bonferroni-corrected post-hoc t-tests revealed that participants responded significantly faster on food trials [t(30) = −4.578, p < 0.001] and non-food trials [t(30) = −4.762, p < 0.001] compared to fixation cross trials, with no difference between food and neutral trials [t(30) = −1.474, p = 0.151]. Statistically significant differences were observed between all SOAs, which remained after correction was applied. Participants had significantly longer reaction times for the non-cued trials compared to cued trials with SOAs of 100 ms [t(30) = 10.170, p < 0.001], 300 ms [t(30) = 7.065, p < 0.001], and 500 ms [t(30) = 14.226, p < 0.001]. Participants responded slowest at 500 ms compared to 100 ms SOA [t(30) = 11.938, p < 0.001] and 300 ms SOA [t(30) = 13.158, p < 0.001]. Unexpectedly, as can be seen in Figure 2, participants responded faster at 100 ms SOA compared to 300 ms SOA [t(30) = −3.397, p = 0.012]. Figure 2 Benefit of the warning cue in the (A) fed and (B) fasted states. Error bars denote standard error (ms). Effect of uncertainty in non-cued trials: preparation cost Reaction time during the pure block (only non-cued trials: no uncertainty) and the mixed block (mixture of cued and non-cued trials: some uncertainty) were compared using a 2 × 2 × 3 ANOVA to assess the impact of condition (pure/mixed), state (fed/fasted), and stimulus (food/non-food/fixation cross) on the employment of proactive inhibition as a goal-directed strategy (i.e., preparation cost). This yielded a significant main effect of condition, F(1, 28) = 101.48, p < 0.001. Participants responded more slowly during the mixed block (mean RT = 406.35 ms, SD = 41.19 ms), which involved preparation for slowing responses to distinguish the target from the cue, compared to the pure block (mean RT = 365.39 ms, SD = 40.34 ms). Post-hoc t-tests revealed that participants responded faster in the pure block compared to non-cued trials within the mixed blocks for all stimulus types in the fed state [food: t(28) = 7.064, p < 0.001; non-food: t(28) = 9.425, p < 0.001; fixation cross: t(28) = 6.677, p < 0.001] and fasted state [food: t(28) = 7.270, p < 0.001; non-food: t(28) = 7.657, p < 0.001; fix: t(28) = 7.678, p < 0.001], all of which remained at p < 0.001 after Bonferroni correction. The ANOVA also revealed a main effect of stimulus type, F(1.678,46.98*) = 39.612, p < 0.001, whereby average response time was fastest for food trials (mean = 380.69 ms, SD = 38.72 ms), followed by non-food trials (mean = 389.11 ms, SD = 39.58 ms) and slowest for fixation cross trials (mean = 395.29 ms, SD = 41.99 ms). These main effects were qualified by a significant interaction between condition and stimulus type [F(1.748,48.954*) = 4.194, p = 0.025; see Figure 3]. Post-hoc t-tests revealed that this interaction was driven by significantly faster reaction times to food stimuli compared to neutral stimuli in pure [t(28) = −5.222, p(corrected) < 0.001], but not the mixed block [t(28) = −0.399, p = 0.693]. Response times remained significantly faster during food trials and neutral trials compared to fixation cross trials [all t(28) ≥ 4.4, p(corrected) < 0.001]. In contrast, no main effect of state was observed, F(1, 28) = 2.269, p = 0.143, nor any interactions between state and condition, F(1, 28) = 1.498, p = 0.231, state and stimulus, F(2, 56) = 1.237, p = 0.298, or all three factors, F(2, 56) = 0.6, p = 0.553. Figure 3 Mean reaction times on non-cued food, non-food and fixation cross trials in the mixed and pure block of the proactive inhibition task. Error bars denote standard error (ms). ns = not statistically significant comparison. All other within-condition comparisons remained significant after applying Bonferroni correction for multiple comparisons. Reactive inhibition Index of inhibitory control: SSRT Participants showed poorer inhibitory control in the fed (mean = 238.87, SD = 97.60 ms) compared to fasted (mean = 217.63 ms, SD = 90.48 ms) state, and to food cues (mean = 228.76 ms, SD = 102.10 ms) compared to non-food (mean = 217.17 ms, SD = 141.58 ms) and fixation cross cues (mean = 210.66 ms, SD = 126.18 ms). The means and standard deviations of the main SST outcomes (SSRT, mean RT, stop accuracy, go accuracy, and stop delay) for each stimulus in the fed and fasted states are presented in Table 3. Table 3 Means (and standard deviations) of outcome variables on the stop signal task. Fed Fasted Food Non-food No image Food Non-food No image ME state ME stimulus Interaction SSRT (ms) 236.74 (101.15) 250.89 (127.82) 228.99 (89.76) 225.51 (115.16) 212.65 (91.71) 220.98 (87.74) F(1, 44) = 5.665, p = 0.022 F(2, 88) = 0.136, p = 0.873 F(2, 88) = 2.437, p = 0.093 Mean RT (ms) 578.75 (233.76) 574.54 (241.62) 584.01 (241.17) 547.87 (225.93) 552.65 (212.03) 536.38 (202.41) F(1, 44) = 4.390, p = 0.042 F(2, 88) = 0.086, p = 0.918 F(2, 88) = 1.084, p = 0.343 Stop accuracy (%) 54.94 (22.94) 54.00 (25.80) 45.22 (20.72) 47.22 (20.10) 49.56 (22.58) 43.44 (19.06) F(1, 44) = 7.442, p = 0.009 F(2, 88) = 11.546, p = 0.000 F(2, 88) = 1.767, p = 0.177 Go accuracy (%) 98.95 (1.60) 98.75 (1.77) 98.70 (1.89) 99.08 (1.50) 98.83 (2.34) 98.41 (2.46) Stop delay (ms) 329.93 (189.97) 311.74 (186.28) 315.71 (190.25) 297.42 (172.27) 307.82 (174.96) 278.69 (167.65) SST main outcomes The main outcomes of interest were the SSRT, mean RT, and stop accuracy in the fed and fasted state on each block. Repeated measures ANOVAs revealed a significant main effect of state on SSRT, mean RT, and stop accuracy. This was driven by poorer accuracy, faster reaction times, and smaller SSDs in the fed state (Table 3). A main effect of stimulus was observed for stop accuracy but not for SSRT or mean RT. Post-hoc t-tests revealed this effect was driven by greater stop accuracy on food trials [mean = 51.08% (SD = 20.40%), t(44) = 3.941, p = 0.001 (corrected)] and non-food trials [mean = 51.78% (SD = 22.06%), t(44) = 3.764, p = 0.001 (corrected)] compared to no image trials (mean = 44.33%, SD = 18.33%), with no difference between food and non-food trials [t(44) = −0.489, p > 0.05]. A trend toward a significant interaction between state and stimulus on SSRT was observed. Bonferroni-corrected post-hoc t-tests revealed that this interaction was driven by poorer inhibitory control (higher SSRT) to non-food cues in the fed compared to fasted state [t(44) = 2.981, p(corrected) = 0.015]. No interaction between state and stimulus was observed for mean RT or stop accuracy. Correlations between inhibitory control and state measurements Correlations were conducted to explore whether physiological state measurements were associated with self-reported mood, self-reported impulsivity and task-based measures of inhibitory control, whether self-reported impulsivity correlated with task-based impulsivity, and whether performance on different tasks correlated. As a main effect of state was observed for state measures (hunger, blood glucose) and some task-based measures of impulsivity [temporal discounting measures (Global discount factor, trend toward a main effect); stop signal task measures (stop accuracy, SSRT, mean RT)], correlations between these variables were calculated separately for the fed and fasted state. All other correlations were assessed across states. Correlations between state measurements and inhibitory control No associations were observed between state measurements (blood glucose, hunger) and assessments of either temporal discounting (global discount factor) in either state [fed (hunger: rs = −0.077, p = 0.667; glucose: rs = −0.205, p = 0.171); fasted (hunger: rs = 0.067, p = 0.712; glucose: rs = −0.138, p = 0.359)] or proactive inhibition [preparation cost (hunger: rs = −0.182, p = 0.344; glucose: r = −0.167, p = 0.386), warning benefit (hunger: rs = −0.065, p = 0.730; glucose: r = −0.067, p = 0.721)]. In contrast, while measures on the stop signal task were not related to state measurements in the fed state [mean RT (hunger: rs = 0.120, p = 0.506; glucose: rs = 0.106, p = 0.488), SSRT (hunger: rs = 0.103, p = 0.569; glucose: rs = 0.079, p = 0.606), stop accuracy (hunger: rs = 0.107, p = 0.554; glucose: r = 0.103, p = 0.500)], hunger statistically significantly negatively correlated with stop accuracy (%) (rs = −0.348, p = 0.047) in the fasted state. Self-reported hunger and glucose were not correlated with any questionnaire measures of impulsivity [Delaying Gratification Inventory (hunger: rs = −0.070, p = 0.705; glucose: r = −0.130, p = 0.395), Barratt Impulsiveness Scale motor subscale score (hunger: rs = 0.243, p = 0.173; glucose: rs = 0.135, p = 0.373)]. Correlations between task-based measures and questionnaire measures of inhibitory control and impulsivity Total scores on the Delaying Gratification Inventory were significantly negatively correlated to Barratt Impulsiveness Scale motor subscale scores (rs = −0.377, p = 0.011). Temporal discounting was related to both questionnaire measures: global discount factor positively correlated with Delaying Gratification Inventory total scores (trend: rs = 0.287, p = 0.056) and negatively correlated with Barratt Impulsiveness Scale Motor subscale scores (rs = −0.344, p = 0.019). With respect to proactive and reactive inhibition, few correlations with questionnaire measures of impulsivity were observed. Delaying Gratification Inventory total scores did not correlate with any outcome measure on the proactive inhibition task (preparation cost: rs = −0.190, p = 0.332; warning benefit: rs = 0.006, p = 0.976) or stop signal task (mean RT: rs = −0.112, p = 0.468; SSRT: rs = −0.101, p = 0.515; stop accuracy: rs = −0.075, p = 0.626). Similarly, Barratt Impulsiveness Scale motor subscale scores did not correlate with outcome measures on the proactive inhibition task (preparation cost: rs = −0.002, p = 0.991; warning benefit: rs = −0.058, p = 0.758). While Barratt Impulsiveness Scale motor subscale scores were positively correlated with mean RT (trend: rs = 0.258, p = 0.087) and SSRT (trend: rs = 0.255, p = 0.091) on the stop signal task, it did not correlate with stop accuracy (%) (rs = 0.245, p = 0.104). Correlations between task outcome measures indicated a consistent positive relationship between outcomes on the proactive inhibition and stop signal tasks. Warning benefit was significantly positively correlated with mean RT (rs ≤ −0.451, p = 0.012), SSRT (rs ≤ 0.394, p = 0.031), and stop accuracy (rs ≤ 0.367, p = 0.046) on the stop signal task. In contrast, preparation cost did not correlate with any stop signal task measure (all rs ≤ 0.095, p ≥ 0.630). No correlations were observed between temporal discounting and proactive inhibition task outcomes (all rs ≤ −0.066, p ≥ 0.725) or between temporal discounting and stop signal task outcomes in either state (all rs ≤ −0.168, p ≥ 0.173). Discussion This study examined the influence of promoting a “need” state (via acute food restriction) and motivationally relevant distraction (food images) on several aspects of inhibitory control. It was predicted that inhibitory control is impaired after food restriction, i.e., in the fasted compared to the fed state; that food stimuli will promote approach behaviors leading to poorer inhibitory control, and that food images will impair inhibitory control most strongly in the fasted condition due to motivational relevance. State effects Our data partially supported our hypothesis that acute food restriction would impair inhibitory control, as participants exhibited greater temporal discounting (preference for smaller immediate rewards) after acute fasting. However, food restriction had no effect on questionnaire measures of inhibitory control or on proactive inhibition. Moreover, reactive inhibitory control on the SST in fact improved after fasting. Contrary to De Ridder et al. (2014), our participants showed greater temporal discounting after acute food restriction compared to when satiated. This difference in findings may be due to the methods used in the two studies: we employed a repeated-measures design, whereas a between-subjects design was used by De Ridder et al. (2014). In addition, hunger did not correlate with temporal discounting in either state. In contrast, our data revealed improved reactive inhibitory control (lower SSRT and greater stop accuracy) in the fasted state compared to the fed state. However, this may be influenced by level of hunger, as hunger negatively correlated with stop accuracy in the fasted state. In other words, individuals who were less hungry after acute fasting showed greater stop accuracy than those who were more hungry after acute fasting. These findings hold implications for individuals who engage in dieting and food restraint, as well as for individuals with eating disorders. The observation that acute food restriction leads to poorer inhibitory control performance may provide some explanation as to why poorer inhibitory control is often reported in restrained eaters and has been reported in eating disorders (for review, see Bartholdy et al., 2016). As eating restraint was not assessed in this study, it is of interest to explore whether this is the case in restrained eaters and in successful and unsuccessful dieters to understand how inhibitory control in the context of various eating patterns/behavior is affected by food and hunger, and how this might translate into success or failure in dietary restraint, weight management and overall healthy and unhealthy eating patterns. Future studies may also wish to explore the impact of food-specific inhibitory control on later consumption of food to assess disinhibition effects in healthy non-dieters. Motivational relevance Stimulus type had an effect on proactive preparation cost and the accuracy of reactive inhibition (stop accuracy). When comparing the reaction time of non-cued trials in the pure and mixed blocks on the proactive task, a significant main effect of stimulus was observed, with overall faster reaction times on food trials (compared to non-food and fixation cross trials). This is in line with previous studies that have revealed faster approach behaviors to food stimuli (e.g., Seibt et al., 2007). Moreover, an interaction between stimulus and condition was observed: reaction times were significantly shorter on food trials than both non-food and fixation cross trials in the pure block, but were not significantly shorter than non-food trials in the mixed block. As the mixed block was associated with increased uncertainty (i.e., in terms of whether or not the upcoming visual stimulus was a target or a warning signal), this finding may reflect either additional effortful slowing in the context of food (i.e., greater inhibition of approach motivated behaviors) to ensure accurate responding, or that food stimuli facilitated approach behaviors but only in the context of reduced uncertainty. Alternatively, improved performance on food trials may reflect alertness or arousal, i.e., individuals may be more alert to food stimuli when hungry. The data also revealed a main effect of stimulus on stop accuracy, driven by poorer inhibition accuracy during no image trials. This is in contrast to our hypotheses as it was predicted that inhibitory control performance would be worse on food trials as food is thought to stimulate approach behaviors. However, it may have been that the presence of the distracting images had a facilitatory effect on performance. Both the arrow (target) and the distractors were removed from the screen and replaced by the stop signal during stop trials. This may have directed their attention away from the point/target of fixation which may have altered the detection of the stop signal. This is unlikely to fully explain this effect, as the arrow duration was only slightly higher than the mean SSDs across stimulus conditions and states, meaning this would have only provided a facilitatory effect at very short SSDs. Although not assessed in this study, external eating may have contributed to the observed effects of stimulus type on inhibitory control. External eating refers to the tendency to overeat in response to food cues, and is thought to be associated with both enhanced attentional bias to food cues and increased trait impulsivity (Brignell et al., 2009; Nijs et al., 2009; Hou et al., 2011). Thus, food stimuli may be particularly salient to external eaters, and therefore may have more of an influence on their ability to control their behavior compared to non-external eaters, relative to non-food stimuli. It would be of interest to explore the potential mediating role on inhibitory control in the context of food cues, and also how external eating may be influenced by state. Further research may wish to explore external eating by incorporating taste-tests in the laboratory (in which participants are invited to eat as much as they wish of the foods provided) to assess differences in subsequent food intake after exposure to food stimuli compared to non-food stimuli. Such research will improve our understanding of the mechanisms underlying the relationship between food stimuli and eating behaviors in non-clinical populations. It will also provide useful insight into the contexts in which inhibitory control is most influenced by external eating, and therefore the contexts in which behavioral training paradigms may be most effective. Relation of inhibitory control to physiological measures Our data revealed some evidence for a relationship between questionnaire or task-based measures of inhibitory control and physiological measures of hunger and glucose. Despite the notion that self-control is a stable predisposition over time (Davis et al., 2010), previous research has shown that blood glucose levels have an effect on executive function (Stephens and Tunney, 2004) and self-regulation. For example, one study revealed a negative relationship between temporal discounting and blood glucose (manipulated via drinks containing either sugar or sweetener): individuals with lower blood glucose levels showed a higher degree of temporal discounting (i.e., a greater inclination to choose smaller immediate rewards than larger, delayed rewards), whereas individuals with increased blood glucose after ingestion of the sugar-containing drink showed a greater valuation of future rewards (Wang and Dvorak, 2010). Reducing fluctuations in blood glucose levels has even been suggested as a means of improving temporal discounting in disorders characterized by impulsive or compulsive behaviors, such as eating disorders (Wang and Dvorak, 2010). With respect to motor inhibition, evidence from the literature on eating disorders suggests poorer inhibitory control (i.e., longer SSRTs) on the stop signal task in binge-related eating disorders, which are associated with largely fluctuating glucose levels (for review, see Bartholdy et al., 2016). Contrary to these previous findings, the present data did not reveal a relationship between glucose and task-based measures of temporal discounting, proactive inhibition, or reactive inhibition, which may indicate that state-related differences in performance were not underscored by physical differences in energy levels. The discrepancy between our data and previous temporal discounting findings may be related to differences in the procedure for manipulating blood glucose: our study involved an overnight fast and thus may reflect differences in resting glucose levels, whereas Wang and Dvorak (2010) manipulated their participants' blood glucose level through consumption of a soda drink containing either sugar or artificial sweetener. With respect to self-reported inhibitory control, our data revealed a trend toward a negative correlation between Delaying Gratification Inventory total scores and glucose, and a positive correlation between glucose and self-reported motor impulsivity on the Barratt Impulsiveness Scale, in the fasted but not fed state. Together our findings suggest that lower glucose levels are associated with greater self-reported inhibitory control in the fasted state but are not related to inhibitory control assessed using neuropsychological tasks. Methodological considerations This study benefited from its repeated measures design, inclusion of both subjective and objective measurements of inhibitory control and state (glucose, self-reported hunger, and eating), and by assessing several different types of inhibitory control. In addition, this study is the first to explicitly discuss proactive inhibition in the context of eating behavior (hunger state), and findings from this task can form the basis for future studies assessing proactive inhibition in eating disorders and obesity. This study has several limitations. With respect to paradigm design, the images in the stop signal task may have also facilitated go performance accuracy as they provided additional visual information that may have altered attention and subsequent detection of the stop signal. However, go accuracy was high across all conditions, thus the task was already sufficiently simple and any facilitation may be a small effect. In the temporal discounting task, only one time delay was assessed (3 months). Although we used a greater number of trials than typical designs that employ hyperbolic modeling of intertemporal choice behavior, future studies may wish to explore a range of delays to obtain a more comprehensive assessment of participants' valuation of monetary rewards over time. In addition, the impact of stimuli in the SST or proactive inhibition task may be more reliably distinguished using a blocked design in which stimuli are presented in different blocks to remove any lingering effects of previous stimuli on behavior in subsequent trials. Other limitations relate to confounding factors and generalizability. Potential confounding factors such as income and level of education were not assessed. Although this study was only assessing within-subjects factors, and thus the impact of these confounds on the present results should in theory be minimal, this may limit the generalizability of our findings to populations similar in age and education to our sample. While current dieters were excluded during screening, eating restraint tendencies were not assessed in our sample. This is important as evidence suggests that restrained eaters have poorer inhibitory control compared to unrestrained eaters (Bartholdy et al., 2016). Future studies may wish to explore whether inhibitory control in restrained eaters is differentially affected by state or stimuli compared to unrestrained eaters. Additionally, our sample was predominantly comprised of women. While there is some evidence to suggest gender differences in inhibitory control and impulsivity in adults and children, a meta-analysis of healthy adults reported that these gender differences are most evident with respect to reward and punishment sensitivity, risk taking and sensation seeking, whereas no differences between men and women were observed on tasks involving effortful inhibitory control, including assessments of reactive inhibition (go/no-go task, stop signal task), reward-based inhibition (temporal discounting), or cognitive inhibition (e.g., the Stroop task) (Cross et al., 2011). However, future studies may wish to include a more even distribution of men and women to control for possible effects of gender in their analyses. Conclusions The lack of correlations between inhibitory control measures in this study supports the notion that these neuropsychological tasks and questionnaires are tapping into distinct subcomponents of inhibitory control. Acute food restriction did not influence questionnaire measures of inhibitory control, but did appear to affect reactive and reward-based inhibitory control performance on neuropsychological tasks. Participants were on the whole faster at responding to food images in the proactive task, suggesting that food stimuli do motivate approach behavior. However, state did not interact with motivational relevance of distracting stimuli, suggesting that stimulus relevance does not have any additive influences over inhibitory control in non-dieting healthy individuals. Future studies may wish to replicate this study in external eaters, restrained eaters and current dieters to explore whether the influence of food restriction on an individual's ability to exhibit inhibitory control is associated with their responsiveness to food cues, success of eating restraint or weight regulation. Author contributions SB and OO conceived and designed the study. SB and JC collected the data. SB, JC, and OO analyzed the data. SB drafted the manuscript. JC, OO, IC, and US critically revised the manuscript. All authors were involved in interpretation of the data, drafting, critiquing and approving the manuscript and accept responsibility for the accuracy and integrity of this work. Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The views expressed herein are not necessarily those of the grant awarding bodies. SB is supported by a studentship awarded by the National Institute for Health Research (NIHR) [Mental Health Biomedical Research Centre and/or Dementia Biomedical Research Unit] at South London and Maudsley NHS Foundation Trust and King's College London. US and IC receive salary support from the NIHR Biomedical Research Centre for Mental Health, South London and Maudsley NHS Foundation Trust and Institute of Psychiatry, King's College London. OO receives salary support from the NIHR and a Wellcome Trust CRF infrastructure grant. 1Due to violation of the assumption of sphericity, Huynh Feldt degrees of freedom and p values are reported. ==== Refs References Appelhans B. M. Woolf K. Pagoto S. L. Schneider K. L. Whited M. C. Liebman R. (2011 ). Inhibiting food reward: delay discounting, food reward sensitivity, and palatable food intake in overweight and obese women . Obesity 19 , 2175 –2182 . 10.1038/oby.2011.57 21475139 Bartholdy S. Dalton B. O'Daly O. G. Campbell I. C. Schmidt U. (2016 ). A systematic review of the relationship between eating, weight and inhibitory control using the stop signal task . Neurosci. Biobehav. Rev. 64 , 35 –62 . 10.1016/j.neubiorev.2016.02.010 26900651 Berner L. A. Marsh R. (2014 ). Frontostriatal circuits and the development of bulimia nervosa . Front. Behav. Neurosci. 8 :395 . 10.3389/fnbeh.2014.00395 25452718 Brignell C. Griffiths T. Bradley B. P. Mogg K. (2009 ). Attentional and approach biases for pictorial food cues. Influence of external eating . Appetite 52 , 299 –306 . 10.1016/j.appet.2008.10.007 19027808 Brockmeyer T. Hahn C. Reetz C. Schmidt U. Friederich H.-C. (2015 ). Approach bias modification in food craving—A proof-of-concept study . Eur. Eat. Disord. Rev. 23 , 352 –360 . 10.1002/erv.2382 26136357 Brooks S. J. Rask-Andersen M. Benedict C. Schiöth H. B. (2012 ). A debate on current eating disorder diagnoses in light of neurobiological findings: is it time for a spectrum model? BMC Psychiatry 12 :76 . 10.1186/1471-244X-12-76 22770364 Castellanos E. H. Charboneau E. Dietrich M. S. Park S. Bradley B. P. Mogg K. . (2009 ). Obese adults have visual attention bias for food cue images: evidence for altered reward system function . Int. J. Obes. 33 , 1063 –1073 . 10.1038/ijo.2009.138 19621020 Chechko N. Vocke S. Habel U. Toygar T. Kuckartz L. Berthold-Losleben M. . (2015 ). Effects of overnight fasting on working memory-related brain network: an fMRI study . Hum. Brain Mapp. 36 , 839 –851 . 10.1002/hbm.22668 25393934 Chikazoe J. Jimura K. Hirose S. Yamashita K.-I. Miyashita Y. Konishi S. (2009 ). Preparation to inhibit a response complements response inhibition during performance of a stop-signal task . J. Neurosci. 29 , 15870 –15877 . 10.1523/JNEUROSCI.3645-09.2009 20016103 Clarke S. Foster J. (2012 ). A history of blood glucose meters and their role in self-monitoring of diabetes mellitus . Br. J. Biomed. Sci. 69 , 83 . 22872934 Cross C. P. Copping L. T. Campbell A. (2011 ). Sex differences in impulsivity: a meta-analysis . Psychol. Bull. 137 , 97 . 10.1037/a0021591 21219058 Davis C. Patte K. Curtis C. Reid C. (2010 ). Immediate pleasures and future consequences. A neuropsychological study of binge eating and obesity . Appetite 54 , 208 –213 . 10.1016/j.appet.2009.11.002 19896515 De Ridder D. Kroese F. Adriaanse M. Evers C. (2014 ). Always gamble on an empty stomach: hunger is associated with advantageous decision making . PLoS ONE 9 :e111081 . 10.1371/journal.pone.0111081 25340399 Decker J. H. Figner B. Steinglass J. E. (2015 ). On weight and waiting: delay discounting in anorexia nervosa pretreatment and posttreatment . Biol. Psychiatry 78 , 606 –614 . 10.1016/j.biopsych.2014.12.016 25641636 Drobes D. J. Miller E. J. Hillman C. H. Bradley M. M. Cuthbert B. N. Lang P. J. (2001 ). Food deprivation and emotional reactions to food cues: implications for eating disorders . Biol. Psychol. 57 , 153 –177 . 10.1016/S0301-0511(01)00093-X 11454438 Fairburn C. G. (2008 ). Appendix II: Eating Disorder Examination Questionnaire (EDE-Q Version 6.0), Cognitive Behaviour Therapy and Eating Disorders . New York, NY : Guilford Press . First M. B. Spitzer R. L. Gibbon M. Williams J. B. W. (2002 ). Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Non-patient Edition (SCID-I/NP) . New York, NY : Biometrics Research, New York State Psychiatric Institute . Galimberti E. Martoni R. M. Cavallini M. C. Erzegovesi S. Bellodi L. (2012 ). Motor inhibition and cognitive flexibility in eating disorder subtypes . Prog. Neuropsychopharmacol. Biol. Psychiatry 36 , 307 –312 . 10.1016/j.pnpbp.2011.10.017 22079108 Green M. W. Rogers P. J. Elliman N. A. (2000 ). Dietary restraint and addictive behaviors: the generalizability of Tiffany's Cue reactivity model . Int. J. Eat. Disord. 27 , 419 –427 . 10.1002/(SICI)1098-108X(200005)27:4<419::AID-EAT6>3.0.CO;2-Z 10744848 Haynes A. Kemps E. Moffitt R. (2015 ). The moderating role of state inhibitory control in the effect of evaluative conditioning on temptation and unhealthy snacking . Physiol. Behav. 152 , 135 –142 . 10.1016/j.physbeh.2015.09.020 26409213 Hoefling A. Likowski K. U. Deutsch R. Häfner M. Seibt B. Mühlberger A. . (2009 ). When hunger finds no fault with moldy corn: food deprivation reduces food-related disgust . Emotion 9 , 50 . 10.1037/a0014449 19186916 Hoerger M. Quirk S. W. Weed N. C. (2011 ). Development and validation of the delaying gratification inventory . Psychol. Assess. 23 , 725 . 10.1037/a0023286 21480721 Hou R. Mogg K. Bradley B. P. Moss-Morris R. Peveler R. Roefs A. (2011 ). External eating, impulsivity and attentional bias to food cues . Appetite 56 , 424 –427 . 10.1016/j.appet.2011.01.019 21256908 Jaffard M. Longcamp M. Velay J.-L. Anton J.-L. Roth M. Nazarian B. . (2008 ). Proactive inhibitory control of movement assessed by event-related fMRI . Neuroimage 42 , 1196 –1206 . 10.1016/j.neuroimage.2008.05.041 18588986 Karra E. O'Daly O. G. Choudhury A. I. Yousseif A. Millership S. Neary M. T. . (2013 ). A link between FTO, ghrelin, and impaired brain food-cue responsivity . J. Clin. Invest. 123 , 3539 –3551 . 10.1172/JCI44403 23867619 Kekic M. Bartholdy S. Cheng J. McClelland J. Boysen E. Musiat P. . (2016 ). Increased temporal discounting in bulimia nervosa . Int. J. Eat. Disord. . [Epub ahead of print]. 10.1002/eat.22571 27301653 Lawrence N. S. O'Sullivan J. Parslow D. Javaid M. Adams R. C. Chambers C. D. . (2015 ). Training response inhibition to food is associated with weight loss and reduced energy intake . Appetite 95 , 17 –28 . 10.1016/j.appet.2015.06.009 26122756 Lovibond S. H. Lovibond P. F. (1995 ). Manual for the Depression Anxiety Stress Scales, 2nd Edn . Sydney, NSW : Psychology Foundation . Manasse S. M. Espel H. M. Forman E. M. Ruocco A. C. Juarascio A. S. Butryn M. L. . (2015 ). The independent and interacting effects of hedonic hunger and executive function on binge eating . Appetite 89 , 16 –21 . 10.1016/j.appet.2015.01.013 25613129 Manwaring J. L. Green L. Myerson J. Strube M. J. Wilfley D. E. (2011 ). Discounting of various types of rewards by women with and without binge eating disorder: evidence for general rather than specific differences . Psychol. Rec. 61 , 4 . 24039301 Mole T. Irvine M. Worbe Y. Collins P. Mitchell S. Bolton S. . (2015 ). Impulsivity in disorders of food and drug misuse . Psychol. Med. 45 , 771 –782 . 10.1017/S0033291714001834 25118940 Nederkoorn C. Coelho J. S. Guerrieri R. Houben K. Jansen A. (2012 ). Specificity of the failure to inhibit responses in overweight children . Appetite 59 , 409 –413 . 10.1016/j.appet.2012.05.028 22664299 Nederkoorn C. Dassen F. C. Franken L. Resch C. Houben K. (2015 ). Impulsivity and overeating in children in the absence and presence of hunger . Appetite 93 , 57 –61 . 10.1016/j.appet.2015.03.032 25841646 Nederkoorn C. Guerrieri R. Havermans R. Roefs A. Jansen A. (2009 ). The interactive effect of hunger and impulsivity on food intake and purchase in a virtual supermarket . Int. J. Obes. 33 , 905 –912 . 10.1038/ijo.2009.98 19546869 Nijs I. M. T. Franken I. H. A. Muris P. (2009 ). Enhanced processing of food-related pictures in female external eaters . Appetite 53 , 376 –383 . 10.1016/j.appet.2009.07.022 19647769 Nordgren L. F. Van Harreveld F. Van Der Pligt J. (2009 ). The restraint bias: how the illusion of self-restraint promotes impulsive behavior . Psychol. Sci. 20 , 1523 –1528 . 10.1111/j.1467-9280.2009.02468.x 19883487 Oliveira L. Mocaiber I. David I. A. Erthal F. Volchan E. Pereira M. G. (2013 ). Emotion and attention interaction: a trade-off between stimuli relevance, motivation and individual differences . Front. Hum. Neurosci. 7 :364 . 10.3389/fnhum.2013.00364 23874284 Patton J. H. Stanford M. S. Barratt E. S. (1995 ). Factor structure of the Barratt impulsiveness scale . J. Clin. Psychol. 51 , 768 –774 . 10.1002/1097-4679(199511)51:63.0.CO;2-1 8778124 Rolls E. T. (1999 ). Brain and Emotions . Oxford : Oxford University Press . Seibt B. Häfner M. Deutsch R. (2007 ). Prepared to eat: how immediate affective and motivational responses to food cues are influenced by food deprivation . Eur. J. Soc. Psychol. 37 , 359 –379 . 10.1002/ejsp.365 Steinglass J. E. Figner B. Berkowitz S. Simpson H. B. Weber E. U. Walsh B. T. (2012 ). Increased capacity to delay reward in anorexia nervosa . J. Int. Neuropsychol. Soc. 18 , 773 –780 . 10.1017/S1355617712000446 22591835 Stephens R. Tunney R. J. (2004 ). Role of glucose in chewing gum-related facilitation of cognitive function . Appetite 43 , 211 –213 . 10.1016/j.appet.2004.07.006 15458808 Stice E. Telch C. F. Rizvi S. L. (2000 ). Development and validation of the eating disorder diagnostic scale: a brief self-report measure of anorexia, bulimia, and binge-eating disorder . Psychol. Assess. 12 :123 . 10.1037/1040-3590.12.2.123 10887758 Thamotharan S. Lange K. Zale E. L. Huffhines L. Fields S. (2013 ). The role of impulsivity in pediatric obesity and weight status: a meta-analytic review . Clin. Psychol. Rev. 33 , 253 –262 . 10.1016/j.cpr.2012.12.001 23313762 Thomas J. M. Higgs S. Dourish C. T. Hansen P. C. Harmer C. J. McCabe C. (2015 ). Satiation attenuates BOLD activity in brain regions involved in reward and increases activity in dorsolateral prefrontal cortex: an fMRI study in healthy volunteers . Am. J. Clin. Nutr. 101 , 697 –704 . 10.3945/ajcn.114.097543 25833968 Turton R. Bruidegom K. Cardi V. Hirsch C. R. Treasure J. (2016 ). Novel methods to help develop healthier eating habits for eating and weight disorders: a systematic review and meta-analysis . Neurosci. Biobehav. Rev. 61 , 132 –155 . 10.1016/j.neubiorev.2015.12.008 26695383 Uher R. Brooks S. J. Bartholdy S. Tchanturia K. Campbell I. C. (2014 ). Increasing cognitive load reduces interference from masked appetitive and aversive but not neutral stimuli . PLoS ONE 9 :e94417 . 10.1371/journal.pone.0094417 24709953 Wang X. Dvorak R. D. (2010 ). Sweet future fluctuating blood glucose levels affect future discounting . Psychol. Sci. 21 , 183 –188 . 10.1177/0956797609358096 20424042 Weber E. U. Johnson E. J. Milch K. F. Chang H. Brodscholl J. C. Goldstein D. G. (2007 ). Asymmetric discounting in intertemporal choice: a query-theory account . Psychol. Sci. 18 , 516 –523 . 10.1111/j.1467-9280.2007.01932.x 17576265 Wierenga C. E. Ely A. Bischoff-Grethe A. Bailer U. F. Simmons A. N. Kaye W. H. (2014 ). Are extremes of consumption in eating disorders related to an altered balance between reward and inhibition? Front. Behav. Neurosci. 8 :410 . 10.3389/fnbeh.2014.00410 25538579 Witt A. A. Raggio G. A. Butryn M. L. Lowe M. R. (2014 ). Do hunger and exposure to food affect scores on a measure of hedonic hunger? An experimental study . Appetite 74 , 1 –5 . 10.1016/j.appet.2013.11.010 24269255

PMC005xxxxxx/PMC5002415.txt

==== Front Front NeurolFront NeurolFront. Neurol.Frontiers in Neurology1664-2295Frontiers Media S.A. 10.3389/fneur.2016.00138NeuroscienceCase ReportCreutzfeldt–Jakob Disease: Analysis of Four Cases Al Balushi Ali 1*Meeks Marshall W. 1Hayat Ghazala 1Kafaie Jafar 11Department of Neurology, Saint Louis University School of Medicine, Saint Louis, MO, USAEdited by: Raymond Scott Turner, Georgetown University, USA Reviewed by: Steve M. Gentleman, Imperial College London, UK; Annalisa Pastore, King’s College London, UK *Correspondence: Ali Al Balushi, albalushia@slu.eduSpecialty section: This article was submitted to Neurodegeneration, a section of the journal Frontiers in Neurology 29 8 2016 2016 7 13802 6 2016 10 8 2016 Copyright © 2016 Al Balushi, Meeks, Hayat and Kafaie.2016Al Balushi, Meeks, Hayat and KafaieThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.Background Creutzfeldt–Jakob disease (CJD) is a rare, rapidly progressive neurodegenerative disease that almost always results in death in under a year from onset of symptoms. Here, we report four cases of CJD with different clinical presentations diagnosed at our institution over a 2-year period. Cases The first patient is an 82-year-old woman who presented with depression, cognitive decline, and word-finding difficulty over 4 weeks. The patient deteriorated neurologically to akinetic mutism and death within 6 weeks of presentation. The second patient is a 54-year-old woman with liver cirrhosis who presented with confusion, ataxia, and multiple falls over 4 weeks. She was treated initially for hepatic encephalopathy but continued to progress to mutism, startle myoclonus, and obtundation. Death occurred within 4 weeks of presentation. The third patient is a 58-year-old woman who presented with an 8-week history of confusion, urinary incontinence, Parkinsonism, ataxia, and myoclonus. Death occurred within 2 months from presentation. The fourth patient is a 67-year-old man who presented with a 6-week history of headache, blurred vision, ataxia, and personality change and progressed to confusion, myoclonus, akinetic mutism, and obtundation. Death occurred within 3 weeks from presentation. Conclusion These four cases highlight the varied possible clinical presentations of CJD and demonstrate the importance of considering CJD in patients with atypical presentations of rapidly progressive cognitive decline. To diagnose CJD, brain biopsy remains the gold standard. However, the presence of CSF protein 14-3-3, typical MRI findings and suggestive EEG abnormalities, all support the diagnosis. Creutzfeldt–Jakob diseasedementianeuropathologyprion proteinspongiform encephalopathy ==== Body Introduction Creutzfeldt–Jakob disease (CJD) is a rare neurodegenerative disorder that causes rapidly progressive dementia leading to death. It belongs in a group of diseases known as prion diseases. The central pathological event of CJD is formation of an abnormally folded protein called scrapie prion protein (PrPSc) from the wild-type cellular prion protien PrPC. The PrPSc then acts as a template for more PrPC to be improperly folded into PrPSc in a process that is poorly understood. Unlike PrPC, the PrPSc is insoluble and cannot be degraded by proteinase enzymes. This substance accumulates and results in the pathological changes characteristic of CJD: spongiform changes of gray matter and neuronal loss without inflammation (1). The incidence of CJD in the United States is estimated to be 1–1.5 per million per year (2). The age of onset is usually between 55–75 years, median 68 years, and both genders are affected equally. There are four subtypes of CJD: sporadic, familial, iatrogenic, and variant form (3). Sporadic CJD is the most common form of the disease and constitutes about 85–90% of all CJD cases (4). Clinical presentation is manifested by rapidly progressive cognitive decline and varied associated neuropsychiatric manifestations like myoclonus, cerebellar ataxia, visual symptoms, pyramidal and extrapyramidal signs, and akinetic mutism. The median duration of survival is approximately 4.5 months from onset of symptoms, with 90% of patients surviving less than 1 year (4). Unfortunately, CJD is always fatal. Definite diagnosis of CJD requires brain biopsy, which is not always feasible given the highly invasive nature of the procedure. Other limitations of brain biopsy include concerns of transmission to health care workers and reduced diagnostic yield if obtained tissue is insufficient or performed on an area of the brain without spongiform changes/PrPSc deposition (5–7). These factors can make the diagnosis of CJD challenging, especially if it presents in an atypical way. Here, we present four cases that we diagnosed with sporadic CJD from June 2014 to March 2016 at St. Louis University Hospital in St. Louis, MO, USA. Table 1 summarizes the cases. Table 1 Patient characteristics. Age/sex Clinical presentation Initial diagnosis Time from onset to diagnosis MRI findings EEG findings CSF 14-3-3 Autopsy 82/F Memory impairment, aphasia, and depression Alzheimer’s dementia 7 weeks Small vessel disease Generalized slowing Positive Positive 54/F Confusion, ataxia, and falls Hepatic encephalopathy 6 weeks DWI and FLAIR basal ganglia hyperintensity Periodic sharp wave complexes Positive Positive 58/F Confusion, myoclonus, urinary incontinence, ataxia, and Parkinsonism CJD 10 weeks DWI and FLAIR hyperintensity in basal ganglia and pulvinar nuclei of the thalami Periodic sharp wave complexes Positive Not done 67/M Headache, blurred vision, ataxia, personality change, confusion, and myoclonus Subacute encephalitis 9 weeks DWI and FLAIR hyperintensity in basal ganglia and left frontal cortex Periodic sharp wave complexes and slow background Positive Positive Case Presentations The first patient is an 82-year-old Caucasian woman who presented with a 4-week history of disorientation, word-finding difficulties, and depressed mood. Prior to the onset of these symptoms, she lived alone and was completely independent. Clinically, she demonstrated fluctuating level of orientation and word-finding difficulties. She deteriorated over the course of 2 weeks and developed apathy, urinary incontinence, and akinetic mutism. Laboratory tests revealed normal vitamin B12, folate, TSH, ACE level, ESR, ANA, ANCA, TPO antibodies, and complements levels. HIV 1 and 2 and RPR returned negative. A 24-h EEG showed generalized slowing. Brain MRI revealed chronic small vessel ischemic changes. CSF studies demonstrated normal protein, glucose, and cell counts. PET CT scan was done to rule out occult malignancy and was negative for abnormal uptake. CSF paraneoplastic antibodies returned negative. With more common etiologies of aphasia and rapidly progressing dementia now ruled out, CJD became suspected. CSF 14-3-3 protein and CSF tau protein were both elevated. Brain biopsy was recommended to the family for confirmation of CJD but they declined and instead chose supportive care. The patient passed away 6 weeks later secondary to pneumonia. Autopsy was performed. Her brain was sent to the National Prion Disease Pathology Surveillance Center (NPDPSC) in Cleveland, OH, USA, which confirmed the diagnosis of CJD by detecting PrPSc. The second patient is a 54-year-old Caucasian woman who presented with increasing confusion, ataxia, and multiple falls. Her past medical history was significant for type 2 diabetes mellitus, non-alcoholic steatohepatitis complicated by liver cirrhosis (Child-Pugh A), and depression. Her symptoms started insidiously over 1 month and then significantly progressed. In the 2 weeks prior to admission, she was not able to walk without support and became mute. On clinical exam, she opened her eyes spontaneously but did not follow commands. She localized pain. Occasional startle myoclonus was also noticed. Systemic exam showed no evidence of astrexis, ascites, or organomegaly. Lab tests including ammonia, liver enzymes, and INR were within normal range. Blood count and metabolic panel were normal. She was started on lactulose and intravenous thiamine for suspected hepatic encephalopathy. Continued workup including vitamin B12, TSH, ANA, ANCA, caeruloplasmin, TPO antibodies, and ACE level all came normal. HIV 1 and 2 and RPR returned negative. Heavy metal screen was negative. CSF studies including T. wipplei and ACE level were negative. Autoimmune paraneoplastic antibodies were negative from both serum and CSF. Brain MRI revealed increased signal intensity in the basal ganglia on DWI and FLAIR sequences. Over the subsequent 2 weeks, her mental status continued to worsen. EEG showed generalized periodic sharp wave complexes and slow background activity (Figure 1). Given her MRI and EEG results, CJD became suspected. Her CSF protein 14-3-3 and CSF tau amount were both elevated. Before we could confirm her diagnosis with a brain biopsy, she developed respiratory failure secondary to aspiration pneumonia and passed away. Autopsy was performed. Her brain was sent to NPDPSC, which detected PrPSc and confirmed the diagnosis of CJD. Figure 1 EEG of patient 2 with periodic sharp wave complexes circled. The third patient is a 58-year-old African–American woman who presented with 2-month history of rapidly progressive decline in cognitive function. Two weeks prior to presentation, she developed bladder and bowel incontinence. On exam, she appeared withdrawn and was oriented to herself and place only. She had limb rigidity with bradykinesia and exhibited frequent myoclonic jerks of her upper limbs. Finger to nose test showed mild ataxia. Basic labs including TSH, vitamin B12, folate, and ceruloplasmin levels were within normal. HIV 1 & 2 and RPR came negative. Heavy metal and urine toxicology screens were negative. Brain MRI showed increased signal intensity within the caudate nuclei, putamen, and both pulvinar nuclei of the thalami on DWI and FLAIR sequences (Figure 2). EEG showed periodic sharp wave complexes. Figure 2 MRI brain (1. DWI, 2. FLAIR sequence) of patient 3 that demonstrates basal ganglia hyperintense signal. CSF studies revealed normal protein, glucose, and cell count. CSF ACE level was normal and CSF VDRL came negative. A therapeutic trail of Carbidopa–Levodopa was tried for the rigidity and bradykinesia, but this did not result in any improvement. Paraneoplastic antibody screen came negative; however, CSF 14-3-3 returned positive. Given strong suspicion of CJD, brain biopsy was advised but patient’s family declined it. She passed away 2 months later. Autopsy was not performed due to family wishes. The fourth patient is a 67-year-old Caucasian man who presented with a 6-week history of sharp holocranial headache, blurred vision, vertigo, imbalance, and personality change. On exam, he was lethargic, disoriented, and ataxic. His deep tendon reflexes were diffusely exaggerated with bilateral Babinski sign. During his hospitalization, he became irritable, agitated, and aggressive. He subsequently developed startle myoclonus and worsening confusion. Given his presentation with headaches in addition to mental status change, initial diagnostic consideration included subacute CNS infection. CT head was negative for acute process. Lumbar puncture showed normal CSF cell count and protein. CSF viral, bacterial, and fungal infectious workup was negative. HIV and RPR were negative. Extensive evaluation for encephalopathy yielded; normal metabolic workup, negative systemic infectious process, and negative heavy metal and urine toxicology screens. TSH, vitamin B12, folate, and ceruloplasmin levels were within normal as well as ESR, CRP, and ANA. MRI brain demonstrated mild signal hyperintensity in the basal ganglia and left frontal cortex on DWI and FLAIR sequences. Magnetic resonance angiography and venography were normal. Patient underwent evaluation for occult malignancy with PET CT scan, which did not reveal any focus of abnormal uptake. EEG showed generalized periodic sharp wave complexes and slow background activity. CSF paraneoplastic antibodies came negative. CSF 14-3-3 resulted positive. The patient’s clinical status deteriorated over the subsequent 3 weeks to obtundation and akinetic mutism. Patient’s family opted for withdrawal of care at this stage. Patient passed away. Autopsy was performed and brain tissue was sent to the NPDPSC, which confirmed the diagnosis of CJD. Discussion This case series demonstrates the myriad of possible presentations of CJD. While our third and fourth patients had a very typical presentation, the first two patients presented with clinical syndromes resembling common alternative diagnoses. In the diagnostic evaluation of a patient suspected of having CJD, it is imperative to first rule out common differential diagnoses, some of which are reversible. These include: vascular, toxic, metabolic, infectious, vitamin deficiencies, iatrogenic, autoimmune, and paraneoplastic etiologies (8). Other neurodegenerative disorders, such as frontotemporal dementia and Alzheimer’s disease (AD) must be distinguished from CJD as well. The first patient presented with a rapidly progressive cognitive decline and expressive aphasia at the age of 82 years. Given the patient’s advanced age, AD was initially on the patient’s differential diagnosis. However, the rapid progression of her symptoms over only 4 weeks raised concerns about other possible etiologies such as autoimmune encephalitis and CJD. Sporadic CJD typically affects patients aged 55–75. It is rarer in patients over 80 years of age but has been reported in case reports (9, 10). Alzheimer’s disease can be differentiated clinically from CJD without difficulty, but there are atypical presentations of Alzheimer’s dementia that make this more difficult. Rapidly progressive Alzheimer’s disease has been well described in the literature (11). Van Everbroeck et al. (12) reported the differential diagnoses for 201 patients who underwent evaluation with CSF 14-3-3 for possible CJD, of which 45 patients (22%) had AD as a final diagnosis. In another report from Geschwind et al. (8), 5 out of 67 non-prion diagnoses of 178 patients presenting with rapidly progressive dementia initially suspected of having CJD turned out to be AD. Similarly, in a report from the National Prion Center in Ohio, 352 out of 1,106 brain autopsies performed for evaluation of rapidly progressive dementia were negative for prion disease. Alzheimer’s dementia was diagnosed in 154 patients out of these 352 (43%) (13). Another retrospective study from the Netherlands evaluated brain autopsies of patients with probable or possible CJD over 11 years. This study found only 146 out of 280 patients (52%) had CJD. AD was diagnosed in 40% of the remaining autopsies (14). CJD may present with focal cortical symptoms like aphasia, which is included in the 2007 UCSF criteria for probable CJD (15–17). Our second patient had a background history of liver cirrhosis and, therefore, her initial presentation was suspected to be hepatic encephalopathy, for which she was treated with lactulose. However, her unresponsiveness to treatment, along with her other neurologic deficits such as ataxia prompted a search for an alternative diagnosis. The patient received an EEG to rule out non-convulsive seizure activity, but it showed periodic sharp wave complexes, a finding that hinted to the diagnosis of sCJD. Wieser et al. reported that “the EEG in sCJD shows characteristic changes depending on the stage of the disease, ranging from non-specific findings such as diffuse slowing and frontal rhythmic delta activity in early stages to disease-typical periodic sharp wave complexes in middle and late stages to a reactive coma traces or even alpha coma in preterminal EEG recordings” (18). The diagnostic utility of periodic sharp wave complexes has been demonstrated in multiple studies; Steinhoff et al. (19) found that periodic sharp wave complexes have a high specificity at 91% and a lower sensitivity at 64%. Those figures corresponded to a very high positive predictive value of 95%. Hepatic encephalopathy typically results in EEG findings of generalized slowing and triphasic waves of the frontal lobes (20). However, periodic sharp wave complexes are rather unusual, which helped us to think about the diagnosis of CJD in our second patient. It can be difficult to distinguish hepatic encephalopathy coexisting with CJD presentation (21), but another hint that helped us make the diagnosis of sCJD in this patient was the brain MRI findings. In sCJD, brain MRI may show signal intensity change in various cortical regions and/or deep nuclei. Zerr et al. found the abnormal pattern of increased signal intensity in the basal ganglia and/or ≥2 cortical areas (temporal, parietal, or occipital) on DWI or FLAIR sequences in 83% of sCJD patients (23). Our second and third patients had typical basal ganglia hyperintensities. Contrary to the first two patients, our third and fourth patients had a rather typical clinical presentation of CJD and were diagnosed without difficulty. Both presented with a rapidly progressive dementia and suggestive clinical findings of ataxia, startle myoclonus, and akinetic mutism in addition to visual, pyramidal, and extrapyramidal symptoms, characteristic periodic sharp wave complexes on EEG, basal ganglia hyperintensities on brain MRI, and positive CSF 14-3-3, which are all typical findings of CJD. The MRI-CJD Consortium criteria for diagnosis of sCJD are shown in Table 2 (23). Table 2 MRI-CJD consortium criteria for sporadic Creutzfeldt–Jakob disease. I. Clinical signs Dementia Cerebellar or visual Pyramidal or extrapyramidal Akientic mutism II. Tests Periodic sharp wave complexes in EEG Protein 14-3-3 detection in CSF (in patients with disease duration of less than 2 years) High signal abnormalities in caudate nucleus and putamen or at least two cortical regions (temporal–parietal–occipital) either on DWI or FLAIR sequences Probable CJD: two out of I and at least one out of II. Possible CJD: two out of I and duration less than 2 years. The median survival time after diagnosis for these four patients was 5.3 weeks, which is less than the expected survival time of 20 weeks (22). This is because the families of most patients opted for palliative care after diagnosis was supported by ancillary studies. Another explanation is possibly our patients presented at relatively more advantaged stage of the disease. Their cognitive decline at onset could have been subtle and went unrecognized until more symptoms became obvious. Conclusion Creutzfeldt–Jakob disease is a fatal neurodegenerative disease that presents with rapidly progressive dementia and a wide range of neuropsychiatric manifestations. CJD usually results in death in less than 1 year after onset. It may present in atypical ways, so it is important for physicians to include it in the differential diagnosis of patients presenting with a rapidly progressive dementia after ruling out common etiologies. Author Contributions AA, JK, and GH were directly involved in the patients’ care. AA wrote the initial draft of the manuscript. MM edited the manuscript and inserted the tables and figures. AA, GH, and JK reviewed and approved the manuscript. Conflict of Interest Statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The authors would like to acknowledge the following physicians who participated in medical care of the reported cases: Dr. Pratap Chand, Dr. Erik Krause, Dr. Fredreck Yap, Dr. Alex Befeler, Dr. Elizabeth Strong, and Dr. Jazba Somroo. ==== Refs References 1 Prusiner SB Scott MR Dearmond SJ Cohen FE Prion protein biology . Cell (1998 ) 93 :337 –48 .10.1016/S0092-8674(00)81163-0 9590169 2 Masters CL Harris JO Gajdusek DC Gibbs CJ JrBernoulli C Asher DM . Creutzfeldt-Jakob disease: patterns of worldwide occurrence and the significance of familial and sporadic clustering . Ann Neurol (1979 ) 5 :177 –88 .10.1002/ana.410050212 371520 3 Ladogana A Puopolo M Croes EA Budka H Jarius C Collins S Mortality from Creutzfeldt-Jakob disease and related disorders in Europe, Australia, and Canada . Neurology (2005 ) 64 :1586 –91 .10.1212/01.WNL.0000160117.56690.B2 15883321 4 Bucelli RC Ances BM . Diagnosis and evaluation of a patient with rapidly progressive dementia . Mo Med (2013 ) 110 (5 ):422 –8 .24279195 5 Jara M John B Kreindel S Wilk-Rivard E Demaria A Jr . Diagnostic evaluation for Creutzfeldt-Jakob disease in Massachusetts, 1991-2001 . Clin Infect Dis (2005 ) 41 :829 –33 .10.1086/432723 16107981 6 Warren JD Schott JM Fox NC Thom M Revesz T Holton JL Brain biopsy in dementia . Brain (2005 ) 128 :2016 –25 .10.1093/brain/awh543 15901648 7 Jansen GH Vogelaar CF Elshof SM Distribution of cellular prion protein in normal human cerebral cortex – does it have relevance to Creutzfeldt-Jakob disease? Clin Chem Lab Med (2001 ) 39 :294 –8 .10.1515/CCLM.2001.045 11388651 8 Geschwind MD Shu H Haman A Sejvar JJ Miller BL . Rapidly progressive dementia . Ann Neurol (2008 ) 64 :97 –108 .10.1002/ana.21430 18668637 9 de Silva R Findlay C Awad I Harries-Jones R Knight R Will R . Creutzfeldt-Jakob disease in the elderly . Postgrad Med J (1997 ) 73 :557 –9 .10.1136/pgmj.73.863.557 9373595 10 Johnson RT Gonzalez RG Frosch MP Case records of the Massachusetts General Hospital. Case 27-2005. An 80-year-old man with fatigue, unsteady gait, and confusion . N Engl J Med (2005 ) 353 :1042 –50 .10.1056/NEJMcpc059024 16148290 11 Schmidt C Wolff M Weitz M Bartlau T Korth C Zerr I . Rapidly progressive Alzheimer disease . Arch Neurol (2011 ) 68 :1124 –30 .10.1001/archneurol.2011.189 21911694 12 Van Everbroeck B Dobbeleir I De Waele M De Deyn P Martin JJ Cras P Differential diagnosis of 201 possible Creutzfeldt-Jakob disease patients . J Neurol (2004 ) 251 (3 ):298 –304 .10.1007/s00415-004-0311-9 15015009 13 Chitravas N Jung RS Kofskey DM Blevins JE Gambetti P Leigh RJ Treatable neurological disorders misdiagnosed as Creutzfeldt-Jakob disease . Ann Neurol (2011 ) 70 :437 –44 .10.1002/ana.22454 21674591 14 Jansen C Schuur M Spliet WG Van Gool WA Van Duijn CM Rozemuller AJ . [Eleven years of autopsy on account of Creutzfeldt-Jakob disease in the Netherlands] . Ned Tijdschr Geneeskd (2009 ) 153 :A172 .19785859 15 Shuttleworth EC Yates AJ Paltan-Ortiz JD . Creutzfeldt-Jakob disease presenting as progressive aphasia . J Natl Med Assoc (1985 ) 77 :649 –50 .3900429 16 Kobylecki C Thompson JC Jones M Mills SJ Shaunak S Ironside JW Sporadic Creutzfeldt-Jakob disease presenting as progressive nonfluent aphasia with speech apraxia . Alzheimer Dis Assoc Disord (2013 ) 27 :384 –6 .10.1097/WAD.0b013e318260ab27 22732678 17 Riancho J Delgado-Alvarado M Fernandez-Torre JL Sanchez-Juan P Polo JM Subacute progressive aphasia: a rare presentation of Creutzfeldt-Jakob disease . J Neurol (2016 ) 263 :600 –2 .10.1007/s00415-016-8054-y 26872666 18 Wieser HG Schindler K Zumsteg D EEG in Creutzfeldt-Jakob disease . Clin Neurophysiol (2006 ) 117 :935 –51 .10.1016/j.clinph.2005.12.007 16442343 19 Steinhoff BJ Zerr I Glatting M Schulz-Schaeffer W Poser S Kretzschmar HA . Diagnostic value of periodic complexes in Creutzfeldt-Jakob disease . Ann Neurol (2004 ) 56 :702 –8 .10.1002/ana.20261 15449324 20 Cash WJ Mcconville P Mcdermott E Mccormick PA Callender ME Mcdougall NI . Current concepts in the assessment and treatment of hepatic encephalopathy . QJM (2010 ) 103 :9 –16 .10.1093/qjmed/hcp152 19903725 21 Camacho-Munoz B Hernandez-Ramos FJ Ortega-Martinez De Victoria L . [Creutzfeldt Jakob disease and hepatic encephalopathy: coexistence in the same patient] . Rev Neurol (2001 ) 33 :540 –4 .11727236 22 Pocchiari M Puopolo M Croes EA Budka H Gelpi E Collins S Predictors of survival in sporadic Creutzfeldt-Jakob disease and other human transmissible spongiform encephalopathies . Brain (2004 ) 127 :2348 –59 .10.1093/brain/awh249 15361416 23 Zerr I Kallenberg K Summers DM Romero C Taratuto A Heinemann U Updated clinical diagnostic criteria for sporadic Creutzfeldt-Jakob disease . Brain (2009 ) 132 :2659 –68 .10.1093/brain/awp191 19773352

PMC005xxxxxx/PMC5002416.txt

==== Front Front PsycholFront PsycholFront. Psychol.Frontiers in Psychology1664-1078Frontiers Media S.A. 10.3389/fpsyg.2016.01286PsychologyOriginal ResearchIndividual Differences in Working Memory Capacity Modulates Semantic Negative Priming from Single Prime Words Ortells Juan J. 1*Noguera Carmen 1Álvarez Dolores 1Carmona Encarna 1Houghton George 21Department of Psychology, University of AlmeríaAlmería, Spain2School of Psychology, University of Wales BangorBangor, UKEdited by: Antonino Vallesi, University of Padua, Italy Reviewed by: Roberto Dell’Acqua, University of Padua, Italy; Jason F. Reimer, California State University, San Bernardino, USA; Hsuan-Fu Chao, Chung Yuan Christian University, Taiwan *Correspondence: Juan J. Ortells, jortells@ual.esThis article was submitted to Cognition, a section of the journal Frontiers in Psychology 29 8 2016 2016 7 128612 5 2016 11 8 2016 Copyright © 2016 Ortells, Noguera, Álvarez, Carmona and Houghton.2016Ortells, Noguera, Álvarez, Carmona and HoughtonThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.The present study investigated whether semantic negative priming from single prime words depends on the availability of cognitive control resources. Participants with high vs. low working memory capacity (as assessed by their performance in complex span and attentional control tasks) were instructed to either attend to or ignore a briefly presented single prime word that was followed by either a semantically related or unrelated target word on which participants made a lexical decision. Individual differences in working memory capacity (WMC) mainly affected the processing of the ignored primes, but not the processing of the attended primes: While the latter produced reliable positive semantic priming for both high- and low-WMC participants, the former gave rise to reliable semantic negative priming only for high WMC participants, with low WMC participants showing the opposite positive priming effect. The present results extend previous findings in demonstrating that (a) single negative priming can reliably generalize to semantic associates of the prime words, and (b) a differential availability of cognitive control resources can reliably modulate the negative priming effect at a semantic level of representation. working memory capacitynegative primingindividual differencesattentional controlsemantic primingMinisterio de Economía y Competitividad10.13039/501100003329PSI2014-53856-P ==== Body Introduction Selection of a relevant stimulus from among competing irrelevant stimuli is a core cognitive ability. Without efficient selection, coherent interaction with a dynamic and complex environment becomes impossible. An influential paradigm in cognitive psychology that was originally developed to measure attentional selection is that of Negative Priming (NP). NP is typically observed in selective attention tasks that present target stimuli among distractors in two consecutive displays (the first called the prime display and the second called the probe display). The NP effect (Tipper, 1985) is the finding that observers are slower to respond to a target stimulus that appeared as a distractor stimulus on the prime display compared with a target stimulus that did not appear on the prime display (Dalrymple-Alford and Budayr, 1966; Tipper, 1985). The NP effect has been observed for a variety of stimuli and a variety of populations (for reviews, see, e.g., Fox, 1995; Tipper, 2001; Frings et al., 2015). Unlike some other selection tasks (e.g., Stroop, flanker tasks), the NP task allows investigation of the fate of the representation of previously encountered but ignored stimuli. NP is commonly observed when the to-be-ignored distractor is “selected against” a concurrent attended target. However, further work has also reported reliable NP even in the absence of distractor stimuli on the prime display (i.e., single-NP; e.g., Milliken et al., 1998; Frings and Wentura, 2005; Noguera et al., 2007, 2015; Chao and Yeh, 2008). There are at least two competing explanations for the NP effect. On the one hand, it has been taken to imply that successful selective attention to a target stimulus depends in part on selective inhibition of any distractors present. This forward-acting inhibitory account of NP is based on two fundamental ideas: First, both relevant and irrelevant stimuli are initially processed in parallel. Second, selection is a dual process in which an excitatory mechanism acting to enhance the processing of targeted information is complemented by an inhibitory mechanism acting to suppress (and/or decouple from potential effectors) the activation levels of the distractor’s internal representations. The residual inhibition associated with the distracting item is presumed to produce the delayed responding (NP) to this stimulus when it appears as a target on a subsequent probe display (Houghton et al., 1996). On the other hand, NP may reflect a backward-acting process of episodic retrieval (e.g., Fox and de Fockert, 1998). On this account, NP occurs when the current target triggers retrieval of a previous encounter with the same stimulus, which on that occasion served as an irrelevant distractor, causing a delay in the response selection process (e.g., Neill et al., 1992). Evidence has accumulated on both sides of this debate, and has led to some compelling hybrid models proposing how both forward-acting inhibition and backward-acting retrieval processes could contribute to the effect (e.g., Kane et al., 1997; Tipper, 2001). Yet, irrespective of this debate, there is converging evidence to suggest that the processes that contribute to NP are effortful and resource demanding. Thus, the NP effect depends critically on the availability of cognitive control (working memory) resources, which serve to minimize the processing of distractor information (e.g., Lavie et al., 2004; de Fockert et al., 2010; de Fockert, 2013). Indirect evidence for this position is provided by studies on cognitive aging, which tend to show that elderly participants are disproportionally impaired compared to younger participants at tasks that require active rejection of distracting information. For example, by assessing Stroop interference and NP concurrently in the same procedure, Mayas et al. (2012) found that relative to younger participants, older adults showed not only an increased Stroop interference, but also reduced NP from irrelevant (distracting) stimuli, indicating that they failed to inhibit them. Note, however, that such evidence for a link between cognitive control resources and selective attention is mostly indirect, as a reduction in working memory capacity (WMC) in the older (vs. younger) groups is often assumed rather than directly measured in these studies. More direct evidence that cognitive control functions are necessary for NP to occur comes from research showing that NP is directly modulated by working memory load and/or WMC (for recent reviews see Redick et al., 2007; de Fockert, 2013). By measuring distractor interference in a context of varying working memory load (e.g., high vs. low mental load), several studies have demonstrated that a to-be-ignored distractor gives rise to reliable NP only under conditions of low memory load. Under high memory load, NP is eliminated or is even converted to positive priming (PP; e.g., Engle et al., 1995; Chao and Yeh, 2008; de Fockert et al., 2010; see also Chao, 2011). Other studies seek correlations between WM span and selective attention by comparing the performance of low vs. high WM span groups. WM capacity is typically measured with a complex span task, such as the Operation Span -Ospan- task (e.g., Unsworth et al., 2005). In the Ospan task, participants perform number calculations while adding to a list of words (or letters) they keep in memory; working memory span is the sum of all correctly recalled word (or letter) lists. Consistent individual differences in a NP task as a function of WM capacity have been reported, such that only high WM capacity participants showed reliable NP, whereas low WM capacity participants did not (e.g., Conway et al., 1999; see also Long and Prat, 2002). Together, these findings suggest that having low WM capacity has the same effects on NP as having a high load on working memory. It should be noted, however, that research so far examining a dependence of NP on cognitive control (working memory) resources, has used different versions of the identity NP paradigm, in which the prime stimulus itself is repeated as the target stimulus on the subsequent probe display. It thus remains unclear whether a differential availability of control resources (i.e., participants with high vs. low working memory capacities) could modulate NP not only at a relatively low feature (perceptual) level, but also at a more abstract (semantic) level of representation. This issue was addressed in the present research. Traditionally, a central issue in research on the NP effect concerns the level of representation at which it operates (Damian, 2000; see also Tipper, 1985). The reason for that interest is that NP has usually been considered a relevant finding not only in promoting dual conceptions of selective attention (i.e., excitatory mechanisms would be complemented by inhibitory processes), but also in suggesting that an ignored stimulus may undergo a deep level of processing, as proposed for example by late-selection attention models (e.g., Deutsch and Deutsch, 1963). Consequently, it would be critically important to demonstrate that NP does not depend on the physical identity between the prime and the target, and it can also generalize between semantic associates belonging to the same semantic category (cf. Neill and Mathis, 1998). In fact, Tipper (1985) and Tipper and Driver (1988) demonstrated that NP could generalize to the semantic associates of the prime stimulus (e.g., cat–dog). They explained this semantic NP effect by appealing to a spreading prospective inhibition mechanism that mirrors automatic spreading activation suggested to underlie semantic PP from attended stimuli (e.g., Collins and Loftus, 1975; Neely, 1977). According to Tipper and Driver, when a prime distractor is ignored, inhibition beginning at the central representation of that stimulus would spread to related representations, lowering their activation levels below baseline (see Houghton and Tipper, 1994, for a different explanation of semantic NP in terms of inhibitory processes in selective attention). Note that semantic NP can also be accounted for by episodic retrieval theories (Neill et al., 1992; Neill, 1997) by assuming that during prime selection “not-respond” (or “to-be-ignored”) tags can be placed not only on the distractor stimulus itself, but also on related items activated by the distractor. As suggested by Neill (1997), the current target stimulus would cue the retrieval of past processing episodes involving similar stimuli. Accordingly, when the retrieved episode includes information about the response and/or relevance of that stimulus, semantic NP would occur if a previously encoded item semantically related to the current target had been ignored. Although it is usually accepted that NP can also rely on the semantic similarity between the prime distractor and the probe target, the evidence of semantic NP has been elusive thus far, especially when words are used as prime stimuli (e.g., Tipper and Driver, 1988; see Fox, 1995, for a review). Unlike the NP effects from identity or spatial tasks, semantic NP effects from prime words have often been weak and difficult to replicate, with the observed effects being highly sensitive to minor procedural/methodological differences, such as the strength -and forward vs. backward direction- with which the prime and target words are associated; or the type of probe task. Still another factor that might be critical in explaining differences in semantic NP effects across conditions and studies is the existence of individual differences in attention control and working memory capacities (WMC). As previously noted, it has not been investigated yet whether the semantic NP effect could critically depend on WMC. It remains possible that obtaining contradicting semantic NP findings even under highly similar task conditions could at least partly be due to differences between studies regarding the differential proportion of participants showing high vs. low WMC. Note that in order to obtain a reliable semantic NP effect it is not only necessary that the prime distractor is actively ignored (as it is the case regarding identity NP). It is also critical that the activation of its abstract semantic memory representation spreads to the representations of semantically related items, and that inhibition (or “to-be-ignored” or “not-respond” tags) is also applied on these representations related to the ignored prime. It is well established that prefrontal areas, in particular the dorsolateral prefrontal cortex (PFC), seem to support not only attentional control and working memory functions (e.g., active manipulation of task-relevant information, and interference blocking of competing information), but it also plays a role in semantic processing (e.g., Petersen et al., 1988; Shivde and Thompson-Schill, 2004), particularly under task conditions that encourage a strategic or controlled processing of semantic information (e.g., high proportion of related prime-target pairs; a long prime-target stimulus onset asynchrony -SOA-; see for example, Hutchison, 2007). Based on results of several priming studies in thought-disordered schizophrenic patients, Spitzer and colleagues have suggested that PFC could modulate not only controlled processing, but also automatic semantic processing (e.g., Spitzer et al., 1993; see also Kiefer et al., 2005). Schizophrenic patients frequently exhibited increased semantic priming effects for both directly (hen–egg) and indirectly (lemon [sour]–sweet) related prime–target word pairs, compared with healthy control subjects, particularly under conditions that minimize strategic processes (i.e., at short prime-target SOA; a low relatedness proportion –RP). According to these authors, PFC focuses retrieval of semantic information to meaning aspects related to the context, so that automatic spreading activation in semantic networks reaches only closely related nodes (e.g., lemon– sour). Because of dysfunctional prefrontal information processing in schizophrenic patients, spreading activation during semantic access would be stronger and far-reaching, thus resulting in increased direct and indirect automatic priming effects. Because individuals differ greatly in their PFC functioning (Gazzaniga et al., 1998; Kane and Engle, 2002) and because this difference can underlie individual differences not only in attentional control and working memory tasks, but also in semantic priming tasks, it not implausible that semantic NP could be even more sensitive than identity NP to a differential availability of cognitive (working memory) resources. The present research addresses this issue. Current Study As noted above, research examining semantic NP from words has usually produced smaller effects than those obtained from identical prime-probe stimuli, with reliable semantic NP being observed under some limited conditions. Furthermore, a sizeable portion of previous studies have reported a failure to replicate semantic NP from words (see Fox, 1995, for a review), such that some authors have questioned whether semantic NP actually exists (e.g., MacLeod et al., 2002). It is thus not surprising that semantic NP is only barely mentioned in the recent review by Frings et al. (2015; p. 1578). Based on these considerations, it is important to establish the boundary conditions under which consistent semantic NP may be found. Over the last two decades, evidence has accumulated that obtaining reliable semantic NP from words critically depends on (i) instructing participants to actively ignore the prime distractor (e.g., Ortells and Tudela, 1996; Milliken et al., 1998); (ii) using a prime-probe SOA interval long enough to allow an efficient engagement of controlled attentional processes (e.g., Ortells et al., 2001; Noguera et al., 2007); (iii) presenting related prime–target pairs that are both categorically and associatively related (e.g., cat–dog; nose–mouth; see for example Abad et al., 2003), particularly in the forward direction (e.g., Hutchison, 2002); and (iv) requiring participants to make a relatively demanding forced-choice task, such as lexical decision or semantic categorization (instead of naming) on the probe target (e.g., Richards, 1999; Ortells et al., 2001; Hutchison, 2002; Noguera et al., 2007). In the present research we used a paradigm (Daza et al., 2007; Noguera et al., 2007, 2015) that respects all the above conditions, in which participants are required to make a lexical decision (word/non-word) on a probe target that is centrally presented either alone or along with a letter-string distractor (depending on the experiment). The target display is preceded by a prime display containing a single central word (presented briefly and post-masked), which on related trials (50% of word trials) is associatively and categorically related to the upcoming target (e.g., tiger–lion; face–eyes). On unrelated trials (50% of word trials), the prime–target pairs are unrelated words belonging to different semantic categories (e.g., tiger–face; eyes–lion). Immediately before the prime display onset either YES (in green) or NO (in red) is presented, varying randomly from trial to trial. Participants are instructed that if they see YES they should “attend to and remember” the following (prime) word, as it would be tested in a subsequent memory task. However, if they see NO they should actively ignore the following prime word, as it would otherwise disrupt their memory for the “attend to and remember” words1. Previous work (Noguera et al., 2007, 2015) has shown a consistent interaction between attentional instructions and priming effects. The YES cue gave rise to reliable semantic PP, while the NO cue produced reliable semantic NP, regardless of whether the probe target was presented with or without distractors. Apart from replicating this pattern of results, the main goal of the current work was to explore whether semantic NP from ignored words depends on the availability of cognitive control resources. Accordingly, participants with high vs. low working memory capacities (as assessed by their performance in several attentional control and complex span tasks) performed a semantic NP task similar to that used by Noguera et al. (2007; Experiment 4). To the extent that semantic NP depends on WMC, we expected to find a reliable interaction between ignored priming and WMC. Thus, the prime words that participants were instructed to actively ignore should produce reliable semantic NP only for high-WMC participants, with no NP effect or even an opposite PP effect being found for low-WMC participants. By contrast, no reliable interaction of priming with WMC was expected regarding the attended prime words. Although our prime-target SOA (600 ms) seems to be long enough for participants to consciously generate likely targets, the prime-target relatedness proportion employed was 0.5, which is unlikely to recruit effortful strategic processes (Neely, 1977, 1991). In addition, the non-word trials may be considered unrelated trials, as a non-word target was always preceded by a prime word belonging to a different category. Hence the aggregate probability that a prime word is followed by a semantically related target is 0.33. In this case, the priming effects should mainly reflect automatic spreading activation in semantic networks, as participants could not reliably use the attended prime to develop expectancy for specific related targets during the interval between prime and target onset (e.g., Neely, 1991; Hutchison, 2007; Hutchison et al., 2014). To the extent that the attended priming effects in this task are mainly the result of automatic processing mechanisms (i.e., spreading activation), they should be insensitive to individual differences in WMC. We therefore expect that the attended primes should produce reliable PP effects of a similar size for both low-WMC and high-WMC participants (namely, no reliable interaction between attended priming and WMC). Materials and Methods Participants Screening for Working Memory Capacity and Attention Control A sample of 200 native Spanish speakers with normal or corrected-to-normal vision were prescreened for WMC on the basis of their performance on an automated versions of operation and symmetry complex span tasks (see Unsworth et al., 2005, 2009 for more task details). These complex span tasks have demonstrated good reliability and validity, and are strongly predictive of a person’s complex cognition, fluid intelligence, or attention control abilities. The automated operation span task (AOSPAN; Unsworth et al., 2005) requires participants to solve a series of simple mathematical operations while trying to remember a variable set of unrelated items. For example, the participant is shown an operation such as “(3 × 1) -1 = ” for a period of time; the next screen shows a digit such as 6, and the participant is to click on the “yes” or “no” box on the screen to indicate whether the digit is the correct answer to the arithmetic operation. The participant is then shown a letter for 800 ms. After three to seven such items, the participant views a matrix of letters on the screen, attempting to recall the letters in the order in which they were presented (by clicking them with mouse). Participants are encouraged to put equal emphasis on math performance and on letter recall. The number of operation-letter pairs per series varied from three to seven (three series of each length are performed, and the order of presentation was random, so that the participant could not predict series length). The dependent measure, or global Aospan score, is the sum of letters correctly recalled from series that are recalled perfectly (all letters in their correct order with no intrusions). The total possible score thus ranged from 0 to 75. If accuracy on the mathematical operations was below 85%, the subject was excluded. Subjects who scored 44 or higher were classified as high span and those who scored 24 or lower were classified as low span. These cutoffs reflect the upper and lower quartiles of our 200-subject pool. The structure of the automated symmetry span task (ASYMSPAN; Unsworth et al., 2009) is similar to that of the AOSPAN just described, with the following exceptions. First, instead of remembering letters, subjects are presented with a 4 × 4 matrix of blank squares, with one square colored in red on a given trial. Second, instead of solving mathematical problems, subjects make vertical symmetry decisions about an 8 × 8 figure composed of black and white squares. At recall, the subject must indicate the location of the squares within the matrix that were colored for that trial in the same order as they appeared by clicking on the cells of an empty matrix. Finally, the list length can vary between two and five (with three trials of each list length), for a total of 42 symmetry figures and square locations on the task. The dependent measure, or global Asymspan score, is the number of square locations recalled in the correct sequential order (with no intrusions) across all trials. The total possible score thus ranged from 0 to 42. Subjects who scored either 23 or higher, or 12 or lower, were classified as high low span, respectively. These cutoffs reflect the upper and lower quartiles of our 200-subject pool. In addition to obtaining independent global scores in the two span tasks for each participant, we also calculated a z-score WMC composite. Performance on each task was transformed into a z-score based on our database of over 200 scores. A z-score WMC composite was created by averaging across the two complex span tasks’ z-scores for each participant. Quartiles were then computed from the averaged distribution, with z-scores of +0.575 and -0.58 corresponding, respectively, to the upper and lower quartiles of our 200-subject pool. The participants also performed a version of the antisaccade task (Hutchison, 2007; see also, Kane et al., 2001; Unsworth et al., 2004). Previous work with this task has demonstrated differences in groups thought to differ in WMC, such as older vs. younger adults, schizophrenics or patients with lesions in the PFC vs. healthy controls (e.g., de Jong, 2001; see Everling and Fischer, 1998, for a review). In this task, participants are told that an asterisk will appear to the left or right of fixation, and immediately followed by a target stimulus (O or Q). The latter appears either on the same (prosaccade condition), or on the opposite side of the screen (antisaccade condition), as the asterisk. Participants were informed that their task in the antisaccade trials was to look away from the flashed asterisk in order to identify the target before it disappeared. Trials began with a white fixation (+) presented on a gray background for either 1000 or 2000 ms. Following the fixation, a white asterisk (∗) appeared 3° to the left or right of fixation for 300 ms. Both asterisk location (left or right) and asterisk delay (1000 or 2000 ms) varied randomly on a trial-by-trial basis to prevent participants from anticipating when or where the asterisk would appear. Following the asterisk, the target appeared 3° to either the same (prosaccade block) or the opposite side of fixation (antisaccade block) for 100 ms and was immediately replaced by a backward pattern mask (##). The pattern mask was displayed for 5,000 ms, during which time the participants were to press either the Q or the O key to indicate the identity of the target. The timing of the trials was designed such that if participants accidentally made a saccade toward (as opposed to away from) the asterisk, they would not have time to plan and execute another saccade to the opposite side of the screen and reach the target. Participants completed a total of 128 trials: 64 trials (16 practice) for the antisaccade block, and 64 trials for the prosaccade block, with the order of blocks being counterbalanced across participants. Participants Twenty-four low (19 women) and twenty-four high (19 women) WMC participants, who had scored respectively in the lower (<24) and upper (>44) quartiles in the AOSPAN task of our 200-subject database, were recruited for the lexical decision study. The z-score WMC composite for each participant in the high-WMC and low-WMC groups fell also within the upper (>+0.575) and lower (<-0.58) quartiles compared to our database (see Table 1 below). Participants were between 18 and 51 years old (M = 21.9, SD = 6.7 for lows; M = 21.9, SD = 4.8 for highs), and all of them received credit toward course requirements as compensation. All participants signed a written consent after the nature and the consequences of the experiment had been explained. The experiment was conducted in accordance with the Declaration of Helsinki. Table 1 Summary statistics for performance in the complex span (Aospan, Asymspan, and z-score composite) and attentional control (antisaccade and prosaccade trials) tasks by Low-WMC and High-WMC groups. Low-WMC Group High-WMC Group M SD M SD Complex Span tasks   Aospan score 15.38 6.53 52.83 7.68   Asymspan score 7.13 4.50 23.67 8.30   z-score composite -1.10 0.31 +1.01 0.33 Antisaccade task   Antisaccade trials     RT (ms) 726 165 573 149     AC (%) 0.77 0.12 0.91 0.12   Prosaccade trials     RT (ms) 535 157 471 106     AC (%) 0.96 0.062 0.096 0.098 Stimuli and Apparatus The stimulus set was similar to that used by Noguera et al. (2007; see also Noguera et al., 2015). It consisted of 48 concrete and familiar Spanish nouns of 4–7 letters length (16 per category) belonging to four semantic categories (geographical features and atmospheric phenomena, foods, animals, and body parts; see Appendix A), which were selected from the intra-categorical associative norms published by Callejas et al. (2003). From that 48-word set, 24 items were presented only as primes and the remaining 24 were presented only as targets (a different word set was presented during practice trials). A further set of 24-words (six from each category) of length 4–7 letters was selected, and one letter from each word was changed to produce an orthographically regular, pronounceable non-word. All stimuli were presented on a computer screen at a viewing distance of approximately 60 cm. Stimulus delivery and response recordings were controlled by E-prime software (Psychology Software Tools Inc.2). Each trial consisted of a sequence of nine critical displays (see Figure 1): Blank screen, fixation, instruction, fixation, blank screen, prime, mask, blank screen, and target. The fixation display consisted of a white asterisk (∗) presented at the center of the screen on a dark gray background. The instruction display consisted of either the word “YES” printed in green or the word “NO” printed in red and presented just above fixation. Both the prime and the probe displays consisted of a single uppercase letter string (4, 5, 6, or 7 letters) presented at the center of the screen, and subtending an average visual angle of about 2.21° wide and 0.49° high. The mask display consisted of a series of ampersands (“&&&&&&”) at the center of the screen. Participants responded by pressing either the “m” or “c” keys, indicating whether the target letter string was either a meaningful word or a non-word. Mappings of word/non-word decisions and correct key (m or c) were counterbalanced across participants. FIGURE 1 Example of the temporal sequence of events when the target is a “word” presented for the instruction of Attending to and remembering (“YES” printed in green) the prime word. The word stimuli shown here for related and unrelated trials have been translated from Spanish to English. Stimuli are not drawn to scale. Design and Procedure General task instructions were displayed on the monitor and also orally delivered. The timing of the events was as follows (see Figure 1): (1) Blank screen presented for a random duration between 1000 and 1500 ms; (2) Fixation display (∗), presented for 500 ms; (3) Instruction display, consisting of either the word “SI” (yes) in green or “NO” (no) in red at the center of the screen for 250 ms. The participants were instructed that if they saw “YES” they should “attend to and remember” the following word as it would be tested in a subsequent memory task. However, if they saw “NO” then they should “ignore” the following prime word, as it was a distractor that would disrupt their memory for the “attend to and remember” words. The attentional cue varied randomly from trial to trial; (4) Fixation display (∗) presented for 800ms; (5) Blank screen for 800 ms; (6) Prime display containing a single word displayed at fixation for 50 ms; (7) Masking display presented for 50 ms; (8) Blank screen of 500 ms (thus resulting in a prime–target stimulus onset asynchrony – SOA – of 600 ms); (9) Probe display containing a single target letter string for 300 ms, on which the participants made a lexical decision (word vs. non-word). The participants were told to press the appropriate response key (m or c) as quickly and accurately as possible, with the computer emitting a 500-ms beep if the participants made an error. Participants took part in a single session (lasting about 25 min) consisting of 24 practice trials (16 word and 8 non-word trials) followed by 144 experimental trials (divided into 2 consecutive blocks of 72 trials each), consisting of 48 trials containing a non-word target and 96 trials containing a word target. All the 48 non-word trials could be viewed as “unrelated” trials, because every non-word target was preceded by a prime word belonging to a different category to that from which the non-word had been created3. Of the 96 word trials, the prime and target words belonged to different categories on 48 (unrelated) trials, and they were highly associated category members on 48 (related) trials. On half of both the 48-related and the 48-unrelated word trials, participants were instructed to “attend to and remember” the prime word preceding the target (attended trials), whereas on the other half they were instructed to actively “ignore” the prime word preceding the target (ignored trials). Different prime words were always presented on both “attended” and “ignored” trials, with the prime words assigned to each instruction type being counterbalanced across participants. For each participant, every prime word (e.g., LION) appeared two times on non-word trials (i.e., the prime word was followed by two different unrelated non-words), and four times (either as a to-be-attended or as a to-be-ignored prime) on word trials: Twice being followed by a highly associated target word from the same semantic category (i.e., the first ranked exemplar on forward direction in the norms of Callejas et al., 2003; e.g., LION-TIGER), and twice by an unrelated target word belonging to a different category (e.g., LION-HAND). The main factors manipulated in the experiment were WMC, manipulated between-participants at two levels (High vs. Low capacity), Instructions (Attend to vs. Ignore), and Prime-Target Relatedness (Related vs. Unrelated). The last two factors were manipulated within-participants with a different random order for each individual. Half of the trials were “Attend to and remember” and half were “Ignore.” Within each of these conditions, half of the target words were related to the preceding prime words and the remaining half was unrelated. The breakdown of the trials was the same for the non-word targets but only the trials containing word targets were analyzed. Participants were informed that after completing the experiment, they would carry out a recognition task about the previously attended words. Although participants were told that there was only one recognition test, two tests were actually presented, one about the attended and the other about the ignored items. They had to classify each word as “new” or “old” depending on whether or not the word had appeared in the experiment. A total of 48 words were presented: 24 “old” words presented as either attended or ignored words on the prime display, and 24 “new” items (not previously presented) belonging to the same four semantic categories mentioned above. Each participant received a different random ordering of the old and new items. Results and Discussion Working Memory (Complex Span) and Attention Control (Antisaccade) Tasks Descriptive statistics (means and standard deviations) for performance in both the two complex span tasks (global span and z-composite scores), and the antisaccade task (RTs and accuracy in the prosaccade and antisaccade blocks) for both high-WMC and low-WMC groups are presented in Table 1. As can be seen in Table 1, high-WMC individuals performed significantly better than low-WMC individuals in the two complex span tasks [Aospan: t(46) = 18.1; Asymspan: t(46) = 9.6; z-score composite: t(46) = 17.6; all ps< 0.001]. High-WMC participants also reliably outperformed low-span participants in both response latencies [t(46) = -3.34, p = 0.002] and accuracy [t(46) = 4.52, p < 0.001] for the antisaccade trials. In contrast, high- and low-WMC participants performed virtually identically in the prosaccade trials where fast and accurate target identification would be aided by a relatively automatic orienting response. Thus, the differential performance of the two WMC groups does not simply reflect a generalized performance deficit in low- compared with high-WMC subjects, such as lack of attention or slower processing-speed. These impressions were confirmed by results of further mixed analyses of variance in which WMC (high vs. low) was treated as a between-participants factor, and saccade type (antisaccade vs. prosaccade trials) as the within-participants variable. The ANOVAs yielded reliable main effects of both WMC and saccade type, and more interestingly, a reliable interaction between these two variables in both RTs [F(1,46) = 9.25, p < 0.001, η2 = 0.18] and Accuracy [F(1,46) = 17.9, p < 0.001, η2 = 0.28] in the antisaccade condition. The analyses of these interactions demonstrated that high- and low-WMC individuals showed a fairly similar performance on the relatively automatic prosaccade trials. In clear contrast, they reliably differed in the antisaccade trials, with low-span participants having significantly longer latencies [t(46) = 3.34, p = 0.002] and reduced accuracy [t(46) = 4.52, p < 0.001] compared to high-span participants. The pattern of inter-individual task performance differences was also assessed by a correlation analysis for the entire sample of 48 subjects. Verbal and visuospatial span tasks were highly correlated (r = 0.78, p < 0.001), suggesting that individual differences in WMC reflect predominantly domain-general rather domain-specific resources. In addition, high-WMC was correlated with a better performance in the antisaccade trials for both response latencies [Aospan: r = -0.37, p = 0.009; Asymspan: r = -0.51, p < 0.001; z-composite: r = -0.46, p < 0.001], and accuracy [Aospan: r = 0.49; Asymspan: r = 0.63; z-composite: r = 0.59, all ps < 0.001]. In contrast, WCM measures did not reliably predict either latency or accuracy in the prosaccade trials4. Priming Task Trials containing an incorrect response (2.4% of trials) or those with reaction times (RTs) falling more than 2.5 standard deviations from the overall mean RT (2.1% of trials) were removed from analyses. Mean RTs from correct word trial responses were calculated for the participants in each WM capacity group (High vs. Low) as a function of Instructions (Attend to vs. Ignore), and Prime-Target Relatedness (Related vs. Unrelated). Two analyses of variance (ANOVA) were conducted, one with the participants (F1) and other with words (F2) as the random factor. Mean RTs and mean error rates as a function of WM Capacity, Instructions and Relatedness are shown in Table 2. Table 2 Mean (SD) reaction times (in ms), and error percentages (in %) as a function of Working Memory Capacity (High vs. Low capacity), Instructions (Attend to vs. Ignore the prime), and Prime-target Relatedness (Related vs. Unrelated). Instructions Attend to Ignore Low-WMC Group    Unrelated 515 (108.1) 510 (153.7) 3.5 (0.06) 4.9 (0.07)    Related 459 (122.5) 474 (164.2) 1.6 (0.03) 2.1 (0.03) High-WMC Group    Unrelated 574 (211.2) 523 (162.3) 1.7 (0.04) 2.4 (0.05)    Related 504 (215.1) 553 (183.1) 1.0 (0.01) 2.5 (0.05) The error rate analysis revealed only a significant main effect of Relatedness [Related = 1.6%; Unrelated = 3.3%; (F1(1,46) = 9.71, p = 0.003, η2 = 0.17; F2(1,46) = 5.39, p = 0.025, η2 = 0.11]. The analysis of RTs showed a significant main effect for Relatedness [Related = 498 ms; Unrelated = 531 ms; F1(1,46) = 22.98, p < 0.001, η2 = 0.23; F2(1,46) = 14.34, p < 0.001, η2 = 0.24], which interacted with Instructions [F1(1,46) = 25.53, p < 0.001, η2 = 0.36; F2(1,46) = 12.63, p < 0.001, η2 = 0.22]. More interestingly, there was a reliable three-way interaction between Instructions, Relatedness, and WMC [F1(1,46) = 11.54, p = 0.001, η2 = 0.201; F2(1,46) = 7.01, p = 0.011, η2 = 0.13]. Further analyses of this interaction revealed a differential priming pattern from attended vs. ignored primes as a function of WMC (see Figure 2). FIGURE 2 Semantic priming effects (unrelated minus related) for Attended and Ignored primes for High-WMC and Low-WMC participants. The vertical lines depict the standard error of priming scores for each condition. Significant contrasts are highlighted by asterisks (∗p < 0.05; ∗∗p < 0.01). The attended prime gave rise to reliable PP effects for both High-WMC [+70 ms; F1(1,23) = 28.97, p < 0.001, η2 = 0.56; F2(1,23) = 16.13, p < 0.001, η2 = 0.41] and Low-WMC participants [+50 ms; F1(1,23) = 16.5, p < 0.001, η2 = 0.42; F2(1,23) = 11.72, p = 0.002, η2 = 0.34]. This pattern was as predicted. With a relatedness proportion below 0.50, the semantic priming effects would be mainly driven by automatic spreading activation, rather than controlled expectancy generation, regardless of whether a short or longer prime-target SOA is used (Neely, 1991; Kiefer et al., 2005; Hutchison et al., 2014). This view is supported by the absence of a reliable correlation between the attended priming effects and any WMC measure (Aospan, Asymspan, z-composite). Regarding the ignored primes, a differential (opposite) priming pattern as a function of WMC was found, as revealed by a reliable interaction between Ignored Priming and WMC [F1(1,46) = 11.54, p = 0.001, η2 = 0.201; F2(1,46) = 9.89, p = 0.003, η2 = 0.18]. Further analyses of this interaction revealed that the ignored primes produced reliable PP in Low-WMC individuals [+36 ms; F1(1,23) = 7.95, p = 0.01, η2 = 0.26; F2(1,23) = 5.54, p = 0.027, η2 = 0.19], but an opposite NP effect in High-WMC individuals [-30 ms; F1(1,23) = 6.69, p = 0.017, η2 = 0.225; F2(1,23) = 4.42, p = 0.047, η2 = 0.16]. Given the finding of 36 ms facilitation from ignored primes in the low WMC group, we directly compared this to the 50 ms facilitation from the attended primes in the same group, and found it not to reach significance [t(23) = 1.36, p > 0.19]. Thus the low WMC group showed only an unreliable reduction in positive semantic priming following the ignore cue, compared to the very reliable semantic NP shown by the high WMC group. Further evidence in support of a dependence of semantic NP on WMC is the finding that the ignored priming effects (unlike the attended priming effects) did reliably correlate with all the WMC measures in our study [Aospan: r = -0.40, p = 0.005; Asymspan: r = -0.41, p = 0.004; z-composite: r = -0.33, p = 0.02]. These results replicate and extend some previous findings (e.g., Engle et al., 1995; Conway et al., 1999; Chao and Yeh, 2008; de Fockert et al., 2010) in showing that a differential availability of cognitive control (WM) resources reliably modulates NP not only at the feature (perceptual) level but also at a more abstract (semantic) level of representation. An inspection of Table 2 shows that whereas in the low WMC group RTs to unrelated targets were fairly similar for attend and ignore conditions, the same did not occur in the high WMC group. In this latter case, response latencies to unrelated targets in the ignore condition were reliably slower [F(1,23) = 5.8, p = 0.025, η2 = 0.19] than those in the attend condition. One could argue that this differential response pattern in the unrelated condition could be at least partially responsible for the reliable priming by instructions interaction that was found in the High-WMC group. While the source these differences remains unclear, it is very likely they are mainly due to non-systematic (random) between-participants variability. Note that in some prior single-prime NP studies (e.g., Ortells et al., 2003; Noguera et al., 2007; see also Daza et al., 2007) a fairly similar interaction between Instruction and Relatedness (and ignored NP effects of a similar size across experiments) is found regardless of whether RTs to unrelated targets are either different or fairly similar between attend and ignore instructions (see for example Ortells et al., 2003; Experiments 1 vs. 2; see also Noguera et al., 2007; Experiments 2 vs. 5). Second and more importantly, a re-examination of our data in the High-WMC group revealed that in 3 of the 24 participants RTs to the unrelated targets were much slower for the attended than for the ignored condition (2 of the 3 showed NP effects from the ignored primes). We then conducted further ANOVAs in which the data from these three participants were removed. The results of these new analyses again revealed a reliable three-way interaction between Instructions, Relatedness, and WMC [F1(1,43) = 10.7, p = 0.002, η2 = 0.20; F2(1,46) = 5.97, p = 0.018, η2 = 0.12]. Even more importantly, in the High-WMC group the Instructions by Relatedness interaction was again significant [F1(1,20) = 29.64, p = 0.001, η2 = 0.201; F2(1,23) = 17.65, p < 0.001, η2 = 0.43], with the attended prime producing reliable PP [+60 ms; F1(1,20) = 19.9, p < 0.001, η2 = 0.50; F2(1,23) = 7.82, p = 0.010, η2 = 0.25] and the ignored primes reliable NP [-36 ms; F1(1,20) = 8.8, p = 0.008, η2 = 0.31; F2(1,23) = 4.43, p = 0.046, η2 = 0.16]. But interestingly, whereas RT differences between instruction types for related targets were very similar to those found with 24 participants [Attend = 477 ms; Ignore = 551 ms; F1(1,20) = 17.93, p < 0.001, η2 = 0.47; F2(1,23) = 10.34, p = 0.004, η2 = 0.31], the RT differences between attend and ignore instructions for the unrelated targets now failed to reach significance [Attend = 537 ms; Ignore = 515 ms; F1(1,20) = 1.9, p > 0.18; F2 < 1]. In addition, the correlation analysis for the reduced sample of 45 participants again showed that the ignored priming effects (but not the attended priming effects) did reliably correlate with all the WMC measures [Aospan: r = -0.50, Asymspan: r = -0.48; z-composite: r = -0.51, all ps < 0.001]. On this basis we would argue that the reliable interaction between Instructions and Relatedness does mainly depend on differences in WMC. On the other hand, note that whereas Low-WMC individuals were overall faster, but less accurate in the priming task (RT = 489 ms; error rate = 3.2) than the high-WMC group (RT = 539 ms; error rate 1.7), the main effect of WMC was not statistically significant either in RTs or in error rates analyses. These results provide further evidence that general processing speed is not the mechanism responsible for the differential performance of the two WMC groups in our study. Prime Recognition Task The recognition scores for both the attended and ignored primes deviated significantly (p < 0.05) from zero. The analysis of recognition scores showed a reliable main effect for Instructions [F(1,46) = 44.9, p < 0.001, η2 = 0.49], such that they were higher for the attended (d′ = 3.16) than for the ignored primes (d′ = 1.78), thus providing further evidence that the attentional instructions did indeed influence the processing of the primes. The main effect of WM Capacity was also significant [F(1,46) = 5.82, p = 0.02, η2 = 0.12], such that the recognition scores were reliably higher for High-WMC (d′ = 2.94) than for Low-WMC participants (d′ = 2.011). Moreover, the interaction between Instructions and WM Capacity was also significant [F(1,46) = 4.39, p = 0.043, η2 = 0.043]. This was due to the fact that the differences between high- and low-WMC participants on prime recognition was significant for the attended [High-WMC = 3.84; Low-WMC = 2.48; t(46) = 3.68, p = 0.001], but not for the ignored primes [High-WMC = 2.04; Low-WMC = 1.54; t(46) = 1.01, p > 0.31]. Consistent with the above findings are the results of the correlation analysis, which showed that the recognition scores for only the attended primes reliably correlated with all the WMC span scores [Aospan: r = 0.43, p = 0.002; Asymspan: r = 0.41, p = 0.004; z-composite: r = 0.44, p = 0.02]. These results suggests that a higher WMC was associated with an increased recognition of primes words that participants were instructed to attend and remember. On the other hand, despite the reliable interaction between Instructions and WM Capacity, we also found that both high-WMC and low-WMC individuals performed reliably better for the attended than for the ignored primes [High-WMC: Attended = 3.84; Ignored = 2.04; t(46) = 5.5, p < 0.001; Low-WMC: Attended = 2.48; Ignored = 1.54; t(46) = 3.8, p = 0.001]. This is a relevant finding in demonstrating that the fairly similar PP effects from the attended and ignored primes that were observed in the low-WMC group, cannot be explained in terms of a lesser ability to follow the different instructions and/or to switch attention between them during the priming task. General Discussion The current study had two main goals. The first goal was to replicate the differential semantic priming pattern (positive vs. negative) that was found by Noguera et al. (2007, 2015). As noted in the Introduction, the evidence of semantic NP from prime words has been elusive thus far, such that some authors have questioned the existence of the effect (e.g., MacLeod et al., 2002; see also Frings et al., 2015). Therefore, we consider it important to demonstrate that a consistent semantic NP from words can be found as long as several boundary conditions are fulfilled. The second and more important goal was to investigate whether semantic NP from single prime words could critically depend on the availability of cognitive control (working memory) resources. To this end, individuals high and low in WMC (as assessed by their performance in complex span and attention control tasks) performed a semantic NP task similar to that of Noguera et al. (2007; Experiment 4). They were instructed to either attend to or ignore a briefly presented single prime word that was followed by either a semantically related or unrelated target word on which participants made a lexical decision task. Given the 50% prime-target relatedness proportion on word trials, it is highly unlikely that participants could strategically use the attended prime to develop expectancy for a specific related target word (Neely, 1991; Hutchison, 2007; Hutchison et al., 2014). Consequently, we predicted that the attended priming effects would not greatly affected by individual differences in WMC (i.e., no reliable interaction between attended priming and WMC). In contrast, to the extent that semantic NP depends on WMC, we expected that the ignored primes would give rise to reliable NP only in high-WMC but not in low-WMC participants (i.e., a reliable interaction between ignored priming and WMC). As predicted, we found that individual differences in WMC mainly affected the processing of the ignored (irrelevant) primes, but not the processing of the attended (relevant) primes: Whereas positive semantic priming was found in both participant groups, reliable semantic NP was shown only by the high WMC group. In stark contrast, the low WMC participants showed a PP effect from ignored primes. Indeed, this facilitatory effect was only slightly reduced compared to that generated by the attended primes, and the difference (14 ms.) was not significant. However, on the prime recognition task the low WMC group did show a reliable difference between attended and ignored primes, and overall, only the former condition showed a correlation with WMC. This supports the view that encoding of attended items into memory for a later recognition test is an active process which depends on WMC. At least for the low WMC group, ignoring an item for memory encoding appears to be more of a passive process, a decision not to engage the (effortful) encoding mechanism on a stimulus that is otherwise fully processed (leading to semantic facilitation). On the basis of the semantic NP results, the high WMC group appears to have the capacity to engage a more active, inhibitory, processing of to-be-ignored primes, suppressing the spread of activation through the semantic network. As noted in the Introduction, the NP effect has been examined in great detail because it provides a window into the processes involved in selective attention. Note that the task requirement of selecting a target against distractors is retained in the standard NP procedure, in which participants are presented with couplets of prime and probe displays containing at least two stimuli, the to-be-responded target and the to-be-ignored distractor. It thus makes sense that attentional selection of a target against its distractor is a critical aspect in theories of NP, including the inhibition account (Tipper, 1985, 2001; Houghton and Tipper, 1994; Houghton et al., 1996) and the retrieval account (Neill et al., 1992; Neill, 1997; Neill and Mathis, 1998). According to the inhibition account of NP, the principal factor underlying NP is the suppression of the competing distractor by an inhibitory mechanism in the selection process of the prime display. According to the retrieval account, response to a probe target is hampered when its presentation retrieves a tag incompatible with the current behavioral goal, such as “do not respond”. In both accounts, selection in prime trials would lead to either the inhibition of the distractor or a tag associated with the distractor. At first sight, obtaining NP from a single ignored prime (single-prime NP) would run counter to the traditionally accepted assumption that such an effect occurs because of a reaction to the distractor interference in selective attention situations. As pointed out by Milliken et al. (1998), because selection was not required during the presentation of a single prime stimulus (i.e., there was no target to which participants responded during prime presentation), the prime was not selected against, and hence not inhibited. It could, however, be argued that in an experimental procedure requiring participants to “ignore” the first (i.e., the prime) of the two stimuli presented in a rapid temporal sequence, the act of ignoring a single prime might involve the same inhibitory (or “do-not-respond” tagging) mechanisms that can actively suppress distracting items presented in irrelevant spatial locations on more conventional NP tasks. In this view, single-prime NP could also be explained in terms of selective attention processes, if the assumption is made that selective inhibition (or do-not-respond tagging) acting on pre-activated internal (abstract) representations of a prime word can operate not only under “spatial” co-ordinates (i.e., where to attend vs. ignore), but also under a “temporal” (i.e., when to attend vs. ignore) dimension (see Tipper, 2001, for a similar line of argument). On the other hand, the dependence of NP effects on the presence of distractor stimuli in the probe display has been usually found only for tasks such as letter identification (e.g., Moore, 1994) or word (or color) naming (e.g., Milliken et al., 1998; Neill and Kahan, 1999). But when a (perhaps much more demanding) forced-choice binary task, such as a lexical decision or a semantic categorization is used on the probe trials, there are numerous previous reports of reliable NP even if probe distractors are absent (e.g., Yee, 1991; Ortells and Tudela, 1996; Richards, 1999; Ortells et al., 2001; Abad et al., 2003; Daza et al., 2007; Noguera et al., 2007, 2015). While the current study was not mainly intended to investigate the underlying mechanism of NP, our findings of reliable semantic NP for high but not for low WMC individuals can be well accommodated by inhibitory accounts of NP, which assume that forward-acting attentional inhibition is resource demanding. Thus, a lower WMC could be associated to a lesser ability of otherwise-engaged control mechanisms to effectively inhibit and reject the processing of the to-be- ignored primes, thus explaining the lack of NP in the low-WMC group. This latter finding does not necessarily imply an absence of inhibition (or attention control) abilities in low-WMC individuals. Recent working memory research on normal aging (e.g., Gazzaley et al., 2008; Jost et al., 2011) have reported evidence that the selective deficit in suppressing task-irrelevant information during working memory encoding in older adults, would be slowed or delayed rather than generally impaired. Whether low-WMC individuals could show semantic NP from ignored primes (or even perform equivalently to high-WMC subjects) if a longer prime-target SOA had been used in our task remains an interesting issue for future research. The dependence of semantic NP on WMC, as well as the reliable correlation between WMC scores and the ignored priming effects that we observed in the present research would also be consistent with both the inhibition (Hasher et al., 1999, 2007) and the executive attention (Engle and Kane, 2004) theories of working memory. The former account states that older adults and individuals low in WMC primarily have impairment in the ability to reduce (inhibit) interference from task-irrelevant information. The second account assumes that individual differences in WMC would mainly reflect variation in a domain-general attention control ability, needed to actively maintain task relevant representations in the face of distraction (e.g., sustain the task goal and constrain the focus of attention to relevant target items). Our findings that the ignored primes produced semantic NP only in the high-WMC group could thus be attributed to their better ability to either inhibit the ignored (irrelevant) information (Hasher et al., 2007), or maintain task relevant information in an active state and to block and inhibit irrelevant representations from gaining access to WM. There are some recent demonstrations that individuals high in WMC also exhibit improved performance on several memory tasks requiring controlled search abilities, such a free recall, cued recall, or item recognition when recollection and not familiarity is needed. Unsworth and Engle (2007), Unsworth and Spillers (2010) have proposed a dual-component (maintenance/retrieval) model, which can be viewed as an outgrowth of the executive attention account. According to this view, WMC would be composed of both attention control abilities (e.g., active maintenance in the face of distraction) and secondary memory abilities (controlled search), which are necessary to transfer information in and out of working memory. Although our prime recognition task mainly aimed to check the effectiveness of our instruction manipulation, the finding that the attended (and to be remembered) primes were better recognized by high-WMC than by low-WMC participants would also be consistent with this maintenance/retrieval theory. Further evidence in support of both inhibitory and attention control theories of WMC is provided by the results in the antisaccade task. In our study, both verbal and visuospatial span scores showed reliable correlations with response latency and accuracy in the antisaccade, but not in the prosaccade trials. In fact, high- and low-span individuals showed a fairly similar performance on the prosaccade trials, but not on the antisaccade trials, with the low-span group showing longer latencies (and a lesser accuracy) than the high-span group (see Table 1). This pattern replicates that obtained in some previous studies, at least under experimental conditions that make minimal demands on cognitive control, as is the case when the prosaccade and the antisaccade trials are presented across different blocks (Kane et al., 2001; Unsworth et al., 2004; the present study). Under these conditions, it has been suggested that individual differences in working memory span would mainly reflect differences in suppression (inhibition) of prepotent responses (reflexive saccade), rather than in directing the focus of attention (looking forward the flashing cue), which is thought to rely on an exogenous, automatic attentional-capture that does not require the recruitment of executive control. Note also that the absence of significant differences in overall reaction times between the two WMC groups for both the antisaccade trials and the priming task, clearly demonstrates that the differential performance of the two WMC groups in our study cannot be attributed to a generalized performance deficit in low- compared with high-WMC subjects, such as slower responding. Whereas the findings of reliable semantic NP for high-WMC but not for low-WMC individuals are consistent with inhibitory (and attention control) accounts, they are more difficult to explain by a strict episodic retrieval account of NP. To the extent that the retrieval of prior episodic traces, which contributes to the NP effect, is usually assumed to be automatic (Logan, 1988), it remains unclear why a differential WMC should modulate an automatic backward-acting retrieval process. It could be argued that differences in the availability of cognitive control resources might affect the processes involved in labeling or tagging the prime stimulus as more or less relevant. If a lower WMC resulted in the ignored prime being less clearly labeled as irrelevant on the prime display, then retrieval of episodic traces associated with these primes would be less likely to interfere with responses to the probe target, and thus these primes would produce less NP (or even PP) compared with ignored primes that were more successfully labeled as to-be-rejected distractors by participants showing higher working memory capacities. Note, however, that if low-WMC individuals were less able to mark the to-be-ignored primes as irrelevant, then they should also produce recognition scores for these stimuli that are fairly similar to those for the attended primes. But this was clearly not the case in the present research. We found instead that the recognition of ignored primes was reliably impaired for both high- and low-span participants. This latter finding also argues against the possibility that low-WMC participants had more difficulty than high-WMC participants to understand and/or follow the different attention instructions (attend vs. ignore) in our study. The distinction between forward-acting attentional inhibition and backward-acting episodic retrieval theories of NP can be viewed as being somewhat similar to the distinction between proactive and reactive control, which was originally proposed by Braver and colleagues to account for impaired cognitive control exhibited in schizophrenia patients and older adults (e.g., Braver et al., 2001). Proactive control involves maintaining goal information in an accessible state so as to direct attention toward goal-relevant stimuli and away from potential internal and external distractions. This form of cognitive control is effortful and preparatory in nature, as uses predictive cues to prepare for a response to a specific upcoming target. In contrast, reactive control is a backward-acting process that is automatically triggered by target onset and involves retrieving prior contextual (e.g., goal) information from long-term memory. In contrast to proactive control, the reactive form of control does not require continuous effort or monitoring of the environment, but instead involves using a target stimulus to retrieve appropriate actions from long-term memory. The task frequently used to assess this dual-control model is the AX-Continuous Performance Test (AX-CPT), which requires a response to a specific probe target (X letter) only if it follows a specific cue (A letter), and to withhold responses otherwise (e.g., an X letter preceded by a B letter). AX targets occur on 70% of all letter sequences, so a strong expectancy to make a target response is created when the letter A is presented. Proactive control in this paradigm involves maintaining the A cue in working memory and using it to maintain an expectancy to make an X response to the following stimulus. In contrast, reactive control involves using the probe (X) to retrieve the preceding cue from memory and is demonstrated by longer reaction times (or increased errors) in rejecting X targets that were preceded by other cues. Across numerous studies, Braver et al. have provided evidence for reduced proactive control among older adults and individuals with schizophrenia, relative to healthy young adults (see Braver et al., 2007; Braver, 2012, for reviews). Some recent AX-CPT studies have also demonstrated that healthy young adults high in WMC are more likely to efficiently use proactive strategies than low-WMC individuals (e.g., Redick and Engle, 2011; Hutchison et al., 2014; Redick, 2014). The assumption that individuals vary greatly in their ability to maintain context (e.g., task instructions, previous cues) to guide future behavior, would be similar to the role that goal maintenance plays in the executive-attention theory by Engle and Kane (2004). One could argue that the high- and low-WMC participants in our study would differ in their ability to represent and maintain the different instructions (attend vs. ignore) within working memory. The lack of NP in the low-WMC group could be attributed to a greater difficulty in these individuals to effectively represent and/or continuously update the attention instructions in working memory until the prime appeared. While we cannot rule out that individual differences in WMC might reflect a differential use of proactive vs. reactive control strategies, several observations are pertinent here. First, as above noted, the finding that the low-WMC group showed reliably better recognition for the attended than for the ignored primes indicates that they were able to maintain the attend/ignore instructions in working memory. Second, even assuming that high-WMC individuals make a more efficient use of proactive strategies, it remains unclear why this should result in negative, instead of reduced positive, priming from the ignored (irrelevant) primes in our task. Unlike the inhibitory accounts of NP (e.g., Houghton and Tipper, 1994; Tipper, 2001; see also Hasher et al., 2007), the context-processing view is noncommittal on whether a to-be-ignored single-prime word should produce either NP, or reduced PP, relative to a to-be-attended prime. It is now widely agreed that working memory should be viewed as a multifaceted construct, as multiple mechanisms seem to be needed to explain individual differences in WMC (e.g., Shipstead et al., 2012, 2015; Unsworth et al., 2015). It has been recently suggested that variation in WMC and the relation between WMC and higher-order cognitive functions, reflects not only variation in domain-general attention control abilities, or variation in the ability to strategically retrieve information from secondary memory, but also variation in the size (or scope of attention) or capacity of primary memory. By using WM tasks, such as change-detection/location (or visual arrays), which provide a relatively pure measure of storage capacity in a short-term buffer, a variety of recent studies have reported evidence that both behavioral (i.e., the k-index; see Cowan et al., 2005) and electrophysiological (i.e., Contralateral Delay Activity-CDA- amplitude; Vogel and Machizawa, 2004) estimates of a person’s storage capacity, do reliably correlate with complex span performance and also with measures of broad cognitive functions (e.g., fluid intelligence, or gF; see for example Cowan et al., 2005; Fukuda et al., 2010; Shipstead et al., 2014, 2015; Unsworth et al., 2015). Despite such evidence, it is unclear why individuals with larger storage capacities should also be more likely to show NP than individuals with smaller storage abilities. Some recent work suggests that visual arrays performance, just as complex span measures, is not strictly driven by a limited-capacity storage system, but it may also rely on controlled attention abilities (Cowan et al., 2006; Fukuda and Vogel, 2011; Shipstead et al., 2015; Unsworth et al., 2015). Whereas the relationship between attention control and storage capacity remains a lively debated issue (see for example the different conclusions reached by Shipstead et al., 2012, vs. Shipstead et al., 2014, 2015), for future research addressing the dependence of semantic NP on working memory resources, we recommend the use of visual array tasks as estimates of WMC. This would allow more directly investigate whether individual differences in primary memory storage capacity could result critical to obtain semantic NP. Conclusion As predicted, we found that individual differences in WMC mainly affected the processing of the ignored (irrelevant) primes, but not the processing of the attended (relevant) primes: Whereas positive semantic priming was found in both participant groups, reliable semantic NP was shown only by the high WMC group. In stark contrast, the low WMC participants showed a PP effect from ignored primes. Indeed, this facilitatory effect was only slightly reduced compared to that generated by the attended primes, and the difference (14 ms.) was not significant. However, on the prime recognition task the low WMC group did show a reliable difference between attended and ignored primes, and overall, only the former condition showed a correlation with WMC. This supports the view that encoding of attended items into memory for a later recognition test is an active process which depends on WMC. At least for the low WMC group, ignoring an item for memory encoding appears to be more of a passive process, a decision not to engage the (effortful) encoding mechanism on a stimulus that is otherwise fully processed (leading to semantic facilitation). On the basis of the semantic NP results, the high WMC group appears to have the capacity to engage a more active, inhibitory, processing of to-be-ignored primes, suppressing the spread of activation through the semantic network. Author Contributions All the authors contributed equally to: (1) The conception and design of the work as well as analysis and interpretation of data (2) Drafting the work or revising it critically for important intellectual content (3) Final approval of the version to be published and (4) All of them are agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Conflict of Interest Statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer RD and the handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review. This research was financially supported by the Spanish Ministerio de Economía y Competitividad with a research grant to JO (PSI2014-53856-P). 1 While other NP studies have examined the impact of attentional instructions on prime processing, the different instructions have been associated with either different blocks or groups (e.g., Ortells and Tudela, 1996; Milliken et al., 1998). Under these conditions, a differential priming pattern for the attended and ignored primes might reflect some strategic difference between participants other than the effects of the instructions per se. To avoid this possible confound, the instruction either to attend or to ignore the single prime varied randomly from trial to trial (Noguera et al., 2007, 2015). 2 www.pstnet.com/eprime 3 We presented fewer non-word (48) than word trials (96) in order to reduce the “non-word ratio” (i.e., the ratio of trials on which a target is a non-word given that the target is unrelated to its word prime), which would in turn minimize the involvement of retrospective semantic-matching strategies in our priming task (see Neely, 1991 for a more detailed discussion on this issue; see also Ortells et al., 2003; Noguera et al., 2007, for a similar task procedure). Thus, given an unrelated prime-target trial (48 word trials + 48 non-word trials = 96 trials of the total), there was the same probability for the target to be either a “non-word” (48/96) or a “word” (48/96). 4 Several researchers have recently recommended the use of the partial instead of the absolute (global)-scoring method to analyze performance on complex span tasks (e.g., Redick et al., 2012). A partial score reflects the sum of items recalled in the correct serial position, regardless of whether an entire trial was recalled correctly. It has been suggested that relative to absolute-scoring, the partial-scoring method shows higher test-retest correlations, higher internal consistencies and correlations among the different complex span task. Accordingly, partial scores in ospan and symmetry span tasks we also computed for each participant. The overall result pattern was very similar to that found with absolute or global span scores. Namely, high-WMC participants performed significantly better than low-WMC individuals in both span tasks [Aospan: t(46) = 11.5; Asymspan: t(46) = 10; z-partial composite: t(46) = 14.7; all ps < 0.001]. The partial span scores of the two tasks were again highly correlated among them (r = 0.79, p < 0.001). All the partial WMC scores reliably correlated with performance in the antisaccade condition for both response latencies [Aospan: r = -0.35, p = 0.002; Asymspan: r = -0.49, p < 0.001; z-partial composite: r = -0.44, p = 0.002], and accuracy [Aospan: r = 0.50; Asymspan: r = 0.61; z-partial composite: r = 0.60, all ps < 0.001]. In contrast, no reliable correlation was found between any of partial WCM scores and either response latency or accuracy in the prosaccade block. Given that most prior work addressing the relationship between WMC and selective attention tasks has relied on absolute-scoring scores, all analyses reported in the current paper are based on the absolute scoring rather than on partial scoring method in order to remain consistent with prior work. Supplementary Material The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpsyg.2016.01286 Click here for additional data file. ==== Refs References Abad M. J. Noguera C. Ortells J. J. (2003 ). Influence of prime–target relationship on semantic priming effects from words in a lexical-decision task. Acta Psychol. 113 283 –295 . 10.1016/S0001-6918(03)00034-9 Braver T. S. (2012 ). The variable nature of cognitive control: a dual mechanisms framework. Trends Cogn. Sci. 16 106 –113 . 10.1016/j.tics.2011.12.010 22245618 Braver T. S. Barch D. M. Keys B. A. Carter C. S. Cohen J. D. Kaye J. A. (2001 ). Context processing in older adults: evidence for a theory relating cognitive control to neurobiology in healthy aging. J. Exp. Psychol. Gen. 130 746 –763 . 10.1037/0096-3445.130.4.746 11757878 Braver T. S. Gray J. R. Burgess G. C. (2007 ). “Explaining the many varieties of working memory variation: dual mechanisms of cognitive control ,” in Variation in Working Memory , eds Conway A. Jarrold C. Kane M. J. Miyake A. Towse J. N. (New York, NY : Oxford University Press ), 76 –106 . Callejas A. Correa A. Lupiáñez J. Tudela P. (2003 ). Normas asociativas intracategoriales para 612 palabras de seis categorías semánticas en español. Psicológica 24 185 –214 . Chao H. (2011 ). Dissociations between identity and location negative priming. Acta Psychol. 136 81 –89 . 10.1016/j.actpsy.2010.10.005 Chao H. Yeh Y. (2008 ). Controlled processing in single-prime negative priming. Exp. Psychol. 55 402 –408 . 10.1027/1618-3169.55.6.402 19130766 Collins A. Loftus E. (1975 ). A spreading-activation theory of semantic processing. Psychol. Rev. 82 407 –428 . 10.1037/0033-295X.82.6.407 Conway A. R. Tuholski S. W. Shisler R. J. Engle R. W. (1999 ). The effect of memory load on negative priming: an individual differences investigation. Mem. Cogn. 27 1042 –1050 . 10.3758/BF03201233 Cowan N. Elliot E. M. Saults J. S. Morey C. C. Mattox S. Hismjatullina A. (2005 ). On the capacity of attention: its estimation and its role in working memory and cognitive aptitudes. Cogn. Psychol. 51 42 –100 . 10.1016/j.cogpsych.2004.12.001 16039935 Cowan N. Fristoe N. M. Elliott E. M. Brunner R. P. Saults J. S. (2006 ). Scope of attention, control of attention, and intelligence in children and adults. Mem. Cogni. 34 1754 –1768 . 10.3758/BF03195936 Dalrymple-Alford E. C. Budayr B. (1966 ). Examination of some aspects of the Stroop color-word Test. Percept. Mot. Skills 23 1211 –1214 . 10.2466/pms.1966.23.3f.1211 5972923 Damian M. F. (2000 ). Semantic negative priming in picture categorization and naming. Cognition 76 45 –55 . 10.1016/S0010-0277(00)00080-9 Daza M. T. Ortells J. J. Noguera C. (2007 ). Negative semantic priming from consciously vs. unconsciously perceived single words. Psicológica 28 105 –127 . de Fockert J. W. (2013 ). Beyond perceptual load and dilution: a review of the role of working memory in selective attention. Front. Psychol. 4 :287 10.3389/fpsyg.2013.00287 de Fockert J. W. Mizon G. A. D’Ubaldo M. (2010 ). No negative priming without cognitive control. J. Exp. Psychol. Hum. 36 1333 –1341 . 10.1037/a0020404 de Jong R. (2001 ). Adult age differences in goal activation and goal maintenance. Eur. J. Cogn. Psychol. 13 71 –89 . 10.1080/09541440042000223 Deutsch J. A. Deutsch D. (1963 ). Attention: some theoretical considerations. Psychol. Rev. 70 80 –90 . 10.1037/h0039515 14027390 Engle R. W. Conway A. R. Tuholski S. W. Shisler R. J. (1995 ). A resource account of inhibition. Psychol. Sci. 6 122 –125 . 10.1111/j.1467-9280.1995.tb00318.x Engle R. W. Kane M. J. (2004 ). “Executive attention, working memory capacity, and a two-factor theory of cognitive control ,” in The Psychology of Learning and Motivation , ed. Ross B. (New York, NY : Elsevier ), 145 –199 . Everling S. Fischer B. (1998 ). The antisaccade: a review of basic research and clinical studies. Neuropsychologia 36 885 –899 . 10.1016/S0028-3932(98)00020-7 9740362 Fox E. (1995 ). Negative priming from ignored distractors in visual selection: a review. Psychon. Bull. Rev. 2 145 –173 . 10.3758/BF03210958 24203653 Fox E. de Fockert J. W. (1998 ). Negative priming depends on prime-probe similarity: evidence for episodic retrieval. Psychon. Bull. Rev. 5 107 –113 . 10.3758/BF03209464 Frings C. Schneider K. K. Fox E. (2015 ). The negative priming paradigm: an update and implications for selective attention. Psychon. Bull. Rev. 22 1577 –1597 . 10.3758/s13423-015-0841-4 25917144 Frings C. Wentura D. (2005 ). Negative priming with masked distractor-only prime trials: awareness moderates negative priming. Exp. Psychol. 52 131 –139 . 10.1027/1618-3169.52.2.131 15850160 Fukuda K. Vogel E. Mayr U. Awh E. (2010 ). Quantity, not quality: the relationship between fluid intelligence and working memory capacity. Psychon. Bull. Rev. 17 673 –679 . 10.3758/17.5.673 21037165 Fukuda K. Vogel E. K. (2011 ). Individual differences in recovery time from attentional capture. Psychol. Sci. 22 361 –368 . 10.1177/0956797611398493 21310945 Gazzaley A. Clapp W. Kelley J. McEvoy K. Knight R. D’Esposito M. (2008 ). Age-related top-down suppression deficit in the early stages of cortical visual memory processing. Proc. Natl. Acad. Sci. U.S.A. 105 13122 –13126 . 10.1073/pnas.0806074105 18765818 Gazzaniga M. S. Ivry R. B. Mangun G. R. (1998 ). Cognitive Neuroscience: The Biology of the Mind. New York, NY : W.W. Norton . Hasher L. Lustig C. Zacks R. T. (2007 ). “Inhibitory mechanisms and the control of attention ,” in Variation in Working Memory , eds Conway A. R. A. Jarrold C. Kane M. J. Miyake A. Towse J. N. (New York, NY : Oxford University Press ), 227 –249 . Hasher L. Zacks R. T. May C. P. (1999 ). “Inhibitory control, circadian arousal, and age ,” in Attention and Performance XVII: Cognitive Regulation of Performance: Interaction of Theory and Application , eds Gopher D. Koriat A. (Cambridge, MA : MIT Press ), 653 –675 . Houghton G. Tipper S. P. (1994 ). “A dynamic model of selective attention ,” in Inhibitory Mechanism in Attention, Memory and Language , eds Dagenbach D. Carr T. (Orlando, FL : Academic Press ). Houghton G. Tipper S. P. Weaver B. Shore D. I. (1996 ). Inhibition and interference in selective attention: some tests of a neural network model. Vis. Cogn. 3 119 –164 . 10.1080/713756733 Hutchison K. A. (2002 ). The effect of asymmetrical association on positive and negative semantic priming. Mem. Cognit. 30 1263 –1276 . Hutchison K. A. (2007 ). Attentional control and the relatedness proportion effect in semantic priming. J. Exp. Psychol. Learn. Mem. Cogn. 33 645 –662 . 10.1037/0278-7393.33.4.645 17576145 Hutchison K. A. Heap S. J. Neely J. H. Thomas M. A. (2014 ). Attentional control and asymmetric associative priming. J. Exp. Psychol. Learn. Mem. Cogn . 40 844 –856 . 10.1037/a0035781 24548327 Jost K. Bryck R. L. Vogel E. K. Mayr U. (2011 ). Are old adults just like low working memory young adults? Filtering efficiency and age differences in visual working memory. Cereb. Cortex 21 1147 –1154 . 10.1093/cercor/bhq185 20884722 Kane M. J. Bleckley M. K. Conway A. R. Engle R. W. (2001 ). A controlled-attention view of working-memory capacity. J. Exp. Psychol. Gen. 130 169 –183 . 10.1037/0096-3445.130.2.169 11409097 Kane M. J. Engle R. W. (2002 ). The role of prefrontal cortex in working-memory capacity, executive attention, and general fluid intelligence: an individual-differences perspective. Psychon. Bull. Rev. 9 637 –671 . 10.3758/BF03196323 12613671 Kane M. J. May C. P. Hasher L. Rahhal T. Stoltzfus E. R. (1997 ). Dual mechanisms of negative priming. J. Exp. Psychol. Hum. 23 632 –650 . 10.1037/0096-1523.23.3.632 Kiefer M. Ahlegian M. Spitzer M. (2005 ). Working memory capacity, indirect semantic priming and Stroop interference: pattern of interindividual prefrontal performance differences in healthy volunteers. Neuropsychology 19 332 –344 . 10.1037/0894-4105.19.3.332 15910119 Lavie N. Hirst A. de Fockert J. W. Viding E. (2004 ). Load theory of selective attention and cognitive control. J. Exp. Psychol. Gen. 133 339 –354 . 10.1037/0096-3445.133.3.339 15355143 Logan G. D. (1988 ). Toward an instance theory of automatization. Psychol. Rev. 95 492 –527 . 10.1037/0033-295X.95.4.492 Long D. L. Prat C. S. (2002 ). Working memory and Stroop interference: an individual differences investigation. Mem. Cogn. 30 294 –301 . 10.3758/BF03195290 MacLeod C. M. Chiappe D. L. Fox E. (2002 ). The crucial roles of stimulus matching and stimulus identity in negative priming. Psychon. Bull. Rev. 9 521 –528 . 10.3758/BF03196308 12412892 Mayas J. Fuentes L. J. Ballesteros S. (2012 ). Stroop interference and negative priming (NP) suppression in normal aging. Arch. Gerontol. Geriatr. 54 333 –338 . 10.1016/j.archger.2010.12.012 21215468 Milliken B. Joordens S. Merikle P. M. Seiffert A. E. (1998 ). Selective attention: a reevaluation of the implications of negative priming. Psychol. Rev. 105 203 –229 . 10.1037/0033-295X.105.2.203 9577237 Moore C. M. (1994 ). Negative priming depends on probe-trial conflict: where has all the inhibition gone? Percept. Psychophys. 56 133 –147 . 10.3758/BF03208365 7971114 Neely J. H. (1977 ). Semantic priming and retrieval from lexical memory: Roles of inhibitionless spreading activation and limited-capacity attention. J. Exp. Psychol. Gen. 106 226 –254 . 10.1037/0096-3445.106.3.226 Neely J. H. (1991 ). “Semantic priming effects in visual word recognition: a selective review of current findings and theories ,” in Basic Processes in Reading: Visual Word Recognition , eds Besner D. Humphreys G. W. (Hillsdale, NJ : Lawrence Erlbaum Associates ), 264 –336 . Neill W. T. (1997 ). Episodic retrieval in negative priming and repetition priming. J. Exp. Psychol. Learn. 23 1291 –1305 . 10.1037/0278-7393.23.6.1291 Neill W. T. Kahan T. A. (1999 ). Response conflict reverses priming: a replication. Psychon. Bull. Rev. 6 304 –308 . 10.3758/BF03212334 12199215 Neill W. T. Mathis K. M. (1998 ). “Negative priming and related phenomena ,” in Psychology of Learning and Motivation: Advances in Research and Theory , ed. Medin D. L. (San Diego, CA : Academic Press ), 1 –44 . Neill W. T. Valdes L. A. Terry K. M. Gorfein D. S. (1992 ). Persistence of negative priming: II. Evidence for episodic trace retrieval. J. Exp. Psychol. Learn. Mem. Cogn . 18 993 –1000 . 10.1037/0278-7393.18.5.993 1402720 Noguera C. Álvarez D. Carmona E. Parra L. (2015 ). Temperament and negative semantic priming in children 7 to 12 years old. Child Neuropsychol. 21 302 –313 . 10.1080/09297049.2014.913558 24835664 Noguera C. Ortells J. J. Abad M. J. F. Carmona E. Daza M. T. (2007 ). Semantic priming effects from single words in a lexical decision task. Acta Psychol. 125 175 –202 . 10.1016/j.actpsy.2006.07.007 Ortells J. J. Abad M. J. Noguera C. Lupiáñez J. (2001 ). Influence of prime–probe stimulus onset asynchrony and prime precuing manipulations on semantic priming effects with words in a lexical-decision task. J. Exp. Psychol. Hum. 27 75 –91 . 10.1037/0096-1523.27.1.75 Ortells J. J. Fox E. Noguera C. Abad M. J. (2003 ). Repetition priming effects from attended vs. ignored single words in a semantic categorization task. Acta Psychol. 114 185 –210 . 10.1016/j.actpsy.2003.08.002 Ortells J. J. Tudela P. (1996 ). Positive and negative semantic priming of attended and unattended parafoveal words in a lexical decision task. Acta Psychol. 94 209 –226 . 10.1016/0001-6918(95)00045-3 Petersen S. E. Fox P. T. Posner M. I. Mintun M. Raichle M. E. (1988 ). Positron emission tomographic studies of the cortical anatomy of single-word processing. Nature 331 585 –589 . 10.1038/331585a0 3277066 Redick T. S. (2014 ). Cognitive control in context: working memory capacity and proactive control. Acta Psychol. 145 1 –9 . 10.1016/j.actpsy.2013.10.010 Redick T. S. Engle R. W. (2011 ). Integrating working memory capacity and context-processing views of cognitive control. Q. J. Exp. Psychol. 64 1048 –1055 . 10.1080/17470218.2011.577226 Redick T. S. Heitz R. P. Engle R. W. (2007 ). “Working memory capacity and inhibition: cognitive and social consequences ,” in Inhibition in Cognition , eds Gorfein D. S. MacLeod C. M. (Washington, DC : American Psychological Association ), 125 –142 . Redick T. S. Broadway J. M. Meier M. E. Kuriakose P. S. Unsworth N. Kane M. J. (2012 ). Measuring working memory capacity with automated complex span tasks. Eur. J. Psychol. Assess. 28 164 –171 . 10.1027/1015-5759/a000123 Richards A. (1999 ). Implicit memory and anxiety: perceptual identification of emotional stimuli. Eur. J. Cogn. Psychol. 11 67 –86 . 10.1080/713752300 Shipstead Z. Harrison T. L. Engle R. W. (2015 ). Working memory capacity and the scope and control of attention. Atten. Percept. Psychophys. 77 1863 –1880 . 10.3758/s13414-015-0899-0 25911154 Shipstead Z. Lindsey D. R. B. Marshall R. L. Engle R. W. (2014 ). The mechanisms of working memory capacity: primary memory, secondary memory, and attention control. J. Mem. Lang. 72 116 –141 . 10.1016/j.jml.2014.01.004 Shipstead Z. Redick T. S. Hicks K. L. Engle R. W. (2012 ). The scope and control of attention as separate aspects of working memory. Memory 20 608 –628 . 10.1080/09658211.2012.691519 22734653 Shivde G. S. Thompson-Schill S. L. (2004 ). Dissociating semantic and phonological maintenance using fMRI. Cogn. Affect. Behav. Neurosci. 4 10 –19 . 10.3758/CABN.4.1.10 15259886 Spitzer M. Braun U. Hermle L. Maier S. (1993 ). Associative semantic network dysfunction in thought-disordered schizophrenic patients: direct evidence from indirect semantic priming. Biol. Psychiatry 34 864 –877 . 10.1016/0006-3223(93)90054-H 8110913 Tipper S. P. (1985 ). The negative priming effect: inhibitory priming by ignored objects. Q. J. Exp. Psychol. A 37 571 –590 . 10.1080/14640748508400920 4081101 Tipper S. P. (2001 ). Does negative priming reflect inhibitory mechanisms? A review and integration of conflicting views. Q. J. Exp. Psychol. 54 321 –343 . 10.1080/02724980042000183 Tipper S. P. Driver J. (1988 ). Negative priming between pictures and words in a selective attention task: evidence for semantic processing of ignored stimuli. Mem. Cogn. 16 64 –70 . 10.3758/BF03197746 Unsworth N. Engle R. W. (2007 ). The nature of individual differences in working memory capacity: active maintenance in primary memory and controlled search from secondary memory. Psychol. Rev. 114 104 –132 . 10.1037/0033-295X.114.1.104 17227183 Unsworth N. Fukuda K. Awh E. Vogel E. K. (2015 ). Working memory delay activity predicts individual differences in cognitive abilities. J. Cog. Neurosci. 27 853 –865 . 10.1162/jocn_a_00765 Unsworth N. Heitz R. P. Schrock J. C. Engle R. W. (2005 ). An automated version of the operation span task. Behav. Res. Methods 37 498 –505 . 10.3758/BF03192720 16405146 Unsworth N. Redick T. S. Heitz R. P. Broadway J. M. Engle R. W. (2009 ). Complex working memory span tasks and higher-order cognition: a latent-variable analysis of the relationship between processing and storage. Memory 17 635 –654 . 10.1080/09658210902998047 19536691 Unsworth N. Schrock J. C. Engle R. W. (2004 ). Working memory capacity and the antisaccade task: individual differences in voluntary saccade control. J. Exp. Psychol. Learn. Mem. Cogn . 30 1302 –1321 . 10.1037/0278-7393.30.6.1302 15521806 Unsworth N. Spillers G. J. (2010 ). Variation in working memory capacity and episodic recall: the contributions of strategic encoding and contextual-retrieval. Psychon. Bull. Rev. 17 200 –205 . 10.3758/PBR.17.2.200 20382920 Vogel E. K. Machizawa M. G. (2004 ). Neural activity predicts individual differences in visual working memory capacity. Nature 428 748 –751 . 10.1038/nature02447 15085132 Yee P. L. (1991 ). Semantic inhibition of ignored words during a figure classification task. Q. J. Exp. Psychol. A 43 127 –153 . 10.1080/14640749108401002 2017571

PMC005xxxxxx/PMC5002417.txt

==== Front Front PsycholFront PsycholFront. Psychol.Frontiers in Psychology1664-1078Frontiers Media S.A. 10.3389/fpsyg.2016.01289PsychologyEditorialEditorial: Non-pharmacological Interventions for Schizophrenia: How Much Can Be Achieved and How? Andreou Christina 12*Moritz Steffen 11Department of Psychiatry and Psychotherapy, University Medical Center Hamburg-EppendorfHamburg, Germany2Center for Gender Research and Early Detection, University of Basel Psychiatric ClinicsBasel, SwitzerlandEdited and reviewed by: Gianluca Castelnuovo, Catholic University of the Sacred Heart, Italy *Correspondence: Christina Andreou christina.andreou@upkbs.chThis article was submitted to Psychology for Clinical Settings, a section of the journal Frontiers in Psychology 29 8 2016 2016 7 128918 2 2016 12 8 2016 Copyright © 2016 Andreou and Moritz.2016Andreou and MoritzThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.The Editorial on the Research Topic Non-pharmacological Interventions for Schizophrenia: How Much Can Be Achieved and How? psychosisschizophreniapsychotherapeutic processesmetacognitionneurocognition ==== Body For the greatest part of the twentieth century, symptoms of schizophrenia such as delusional beliefs were considered to be “non-understandable,” and attempts to explain and treat these symptoms were predominantly influenced by biological conceptualizations (Mander and Kingdon, 2015). However, insights from behavioral, cognitive and social research as well as societal influences (Mueser et al., 2013; Mander and Kingdon, 2015) have contributed to an increasing appreciation of the importance of cognitive and psychological factors in understanding and treating psychotic symptoms. At the same time, there has been growing discontent with the outcomes achieved through antipsychotic medication alone, especially in terms of functional recovery (Leucht et al., 2009; Jääskeläinen et al., 2013). This, combined with the high reported rates of medication non-adherence (Lieberman et al., 2005), has led to a major boost in the development of non-pharmacological interventions for schizophrenia. Despite promising results, there is still much controversy regarding the usefulness and applicability of psychological interventions in clinical practice, and there is still little evidence regarding their mechanisms of action. The present Research Topic addresses these issues. Naturally, an issue dealing with psychological interventions in schizophrenia could not do without the “heavy artillery,” cognitive behavioral therapy. CBT has been one of the first non-pharmacological interventions to be included in treatment guidelines. However, there is still an ongoing debate about its efficacy (McKenna and Kingdon, 2014). Two articles in the present Research Topic contribute to this debate. Peters et al. provide evidence in favor of CBT effectiveness under routine service delivery conditions in a large sample of patients from a challenging catchment area—a very relevant finding for clinical purposes, since everyday clinical practice may differ from clinical studies in many aspects (e.g., patients with comorbidities, variability in therapist availability and/or experience). On the other hand, Mehl et al. deal with the efficacy of clinical studies on CBT for psychosis. The results of their meta-analysis indicate that CBT has a long-lasting positive effect on delusions compared to standard care, but that this effect might be significantly reduced when CBT is compared to other “active” psychological treatments. However, the authors also provide tentative evidence that theory-driven interventions according to an interventionist-causal approach may lead to improved outcomes compared to standard CBT. Thus, treatment outcomes may be improved using more focused interventions based on knowledge of the factors contributing to psychotic symptoms. Other papers in this issue take up this latter point as well. Three papers focus on variations of metacognitive training (MCT), one of the first interventions to address not delusions per se, but rather reasoning biases associated with their emergence and maintenance. Moritz et al. show that an online metacognitive intervention in the context of a cognitive training program can lead to significant changes of the most prominent biases associated with delusions. So et al. provide evidence that an very brief course of metacognitive training can have beneficial effects in patients with psychosis, and that changes in belief flexibility mediate improvement in delusions. Finally, Balzan and Galletly report on two patients refusing antipsychotic medication, in whom individualized metacognitive therapy led to symptom improvement, confirming that psychological interventions may be a viable option in this patient population (cf. Morrison et al., 2014). An interesting aspect of all three above studies is that they describe short, low-cost interventions that are well suited to address problems such as limited resources and cost considerations, which may hamper the dissemination of psychotherapy interventions (Shafran et al., 2009). Two papers address the processes of improvement rather than determinants of symptoms: Westermann et al. deal with the therapy process and propose that a structured focus on patient motives can improve outcomes of both psychological and pharmacological interventions in patients with psychosis. Menon et al. discuss factors that may affect outcome in group CBT for psychosis and identify an important issue: Despite the wealth of clinical efficacy studies, there is still very little evidence regarding individual factors that may affect treatment success. The authors acknowledge sample size limitations as a cause for this problem and suggest possible solutions. In reading the above articles, one could think that the factors implicated in symptoms and their improvement act independently and/or in an additive manner. However, the reader should keep in mind that different factors may dynamically interact with one another, leading to complex associations with symptoms. In a very interesting analysis, Hesse et al. confirm that self-concept is important for the development of paranoid delusions, but also show that self-concept in itself may be affected by neurocognitive deficits. Hence, cognitive remediation training might contribute to the stability of long-term symptom outcome, even though it is not thought to have a direct effect on delusions (Wykes et al., 2011). However, cognitive remediation programs themselves are being influenced by the above dynamic interaction concept, moving away from the simple ‘drill-and-practice’ approach: In their opinion paper, Cella et al. summarize evidence suggesting that a metacognitive focus, i.e., promoting awareness of cognitive strengths and weaknesses, boosts the efficacy of cognitive remediation by helping patients develop strategies to overcome neurocognitive deficits. Interestingly, metacognitive awareness itself may be affected by high self-esteem (Cella et al., 2014). This stresses the importance of keeping account of multiple patient characteristics during therapy, however “simple” its actual focus may be. Several issues remain open: Many authors in this issue highlight the need to consider outcomes other than psychotic symptoms such as depression or well-being. The reader is also reminded that this Research Topic represents only a small snapshot of a fertile research field that includes a number of alternative approaches (to name but a few, see Kurtz and Richardson, 2012 for social cognitive training, Khoury et al., 2013 for mindfulness interventions, and Grácio et al., 2016 for family interventions). The optimistic take-home message is that, although there is still much work to be done in terms of achieving mainstream status, psychological interventions are not only gradually establishing themselves as effective treatments for psychotic symptoms, but are also furthering our understanding of how these symptoms occur. Author contributions CA, SM carried out literature reviews. CA wrote the first draft of this manuscript. SM critically reviewed the manuscript. Both authors have read and approved the final version of the manuscript. Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. ==== Refs References Cella M. Swan S. Medin E. Reeder C. Wykes T. (2014 ). Metacognitive awareness of cognitive problems in schizophrenia: exploring the role of symptoms and self-esteem . Psychol. Med. 44 , 469 –476 . 10.1017/S0033291713001189 23734941 Grácio J. Gonçalves-Pereira M. Leff J. (2016 ). What do we know about family interventions for psychosis at the process level? A systematic review . Fam. Process 55 , 79 –90 . 10.1111/famp.12155 25900627 Jääskeläinen E. Juola P. Hirvonen N. Mcgrath J. J. Saha S. Isohanni M. . (2013 ). A systematic review and meta-analysis of recovery in schizophrenia . Schizophr. Bull. 39 , 1296 –1306 . 10.1093/schbul/sbs130 23172003 Khoury B. Lecomte T. Gaudiano B. A. Paquin K. (2013 ). Mindfulness interventions for psychosis: a meta-analysis . Schizophr. Res. 150 , 176 –184 . 10.1016/j.schres.2013.07.055 23954146 Kurtz M. M. Richardson C. L. (2012 ). Social cognitive training for schizophrenia: a meta-analytic investigation of controlled research . Schizophr. Bull. 38 , 1092 –1104 . 10.1093/schbul/sbr036 21525166 Leucht S. Arbter D. Engel R. R. Kissling W. Davis J. M. (2009 ). How effective are second-generation antipsychotic drugs? A meta-analysis of placebo-controlled trials . Mol. Psychiatry 14 , 429 –447 . 10.1038/sj.mp.4002136 18180760 Lieberman J. A. Stroup T. S. Mcevoy J. P. Swartz M. S. Rosenheck R. A. Perkins D. O. . (2005 ). Effectiveness of antipsychotic drugs in patients with chronic schizophrenia . N. Engl. J. Med. 353 , 1209 –1223 . 10.1056/NEJMoa051688 16172203 Mander H. Kingdon D. (2015 ). The evolution of cognitive-behavioral therapy for psychosis . Psychol. Res. Behav. Manag. 8 , 63 –69 . 10.2147/PRBM.S52267 25733937 McKenna P. Kingdon D. (2014 ). Has cognitive behavioural therapy for psychosis been oversold? BMJ 348 :g2295 . 10.1136/bmj.g2295 24668904 Morrison A. P. Turkington D. Pyle M. Spencer H. Brabban A. Dunn G. . (2014 ). Cognitive therapy for people with schizophrenia spectrum disorders not taking antipsychotic drugs: a single-blind randomised controlled trial . Lancet 383 , 1395 –1403 . 10.1016/S0140-6736(13)62246-1 24508320 Mueser K. T. Deavers F. Penn D. L. Cassisi J. E. (2013 ). Psychosocial treatments for schizophrenia . Annu. Rev. Clin. Psychol. 9 , 465 –497 . 10.2147/NDT.S49263 23330939 Shafran R. Clark D. M. Fairburn C. G. Arntz A. Barlow D. H. Ehlers A. . (2009 ). Mind the gap: Improving the dissemination of CBT . Behav. Res. Ther. 47 , 902 –909 . 10.1016/j.brat.2009.07.003 19664756 Wykes T. Huddy V. Cellard C. Mcgurk S. R. Czobor P. (2011 ). A meta-analysis of cognitive remediation for schizophrenia: methodology and effect sizes . Am. J. Psychiatry 168 , 472 –485 . 10.1176/appi.ajp.2010.10060855 21406461

PMC005xxxxxx/PMC5002418.txt

==== Front Front MicrobiolFront MicrobiolFront. Microbiol.Frontiers in Microbiology1664-302XFrontiers Media S.A. 10.3389/fmicb.2016.01334MicrobiologyOriginal ResearchScreening of Rhizospheric Actinomycetes for Various In-vitro and In-vivo Plant Growth Promoting (PGP) Traits and for Agroactive Compounds Anwar Sumaira Ali Basharat Sajid Imran *Department of Microbiology and Molecular Genetics, University of the PunjabLahore, PakistanEdited by: Anton Hartmann, Helmholtz Zentrum München, Germany Reviewed by: Lei Zhang, Washington State University, USA; Zakira Naureen, University of Nizwa, Oman *Correspondence: Imran Sajid imran.mmg@pu.edu.pkThis article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology 29 8 2016 2016 7 133414 4 2016 12 8 2016 Copyright © 2016 Anwar, Ali and Sajid.2016Anwar, Ali and SajidThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.In this study 98 rhizospheric actinomycetes were isolated from different wheat and tomato fields, Punjab, Pakistan. The isolates were characterized morphologically, biochemically, and genetically and were subjected to a comprehensive in vitro screening for various plant growth promoting (PGP) traits. About 30% of the isolates screened were found to be the promising PGP rhizobacteria (PGPRs), which exhibited maximum genetic similarity (up to 98–99%) with different species of the genus Streptomyces by using16S rRNA gene sequencing. The most active indole acetic acid (IAA) producer Streptomyces nobilis WA-3, Streptomyces Kunmingenesis WC-3, and Streptomyces enissocaesilis TA-3 produce 79.5, 79.23, and 69.26 μg/ml IAA respectively at 500 μg/ml L-tryptophan. The highest concentration of soluble phosphate was produced by Streptomyces sp. WA-1 (72.13 mg/100 ml) and S. djakartensis TB-4 (70.36 mg/100 ml). All rhizobacterial isolates were positive for siderophore, ammonia, and hydrogen cyanide production. Strain S. mutabilis WD-3 showed highest concentration of ACC-deaminase (1.9 mmol /l). For in-vivo screening, seed germination, and plant growth experiment were conducted by inoculating wheat (Triticum aestivum) seeds with the six selected isolates. Significant increases in shoot length was observed with S. nobilis WA-3 (65%), increased root length was recorded in case of S. nobilis WA-3 (81%) as compared to water treated control plants. Maximum increases in plant fresh weight were recorded with S. nobilis WA-3 (84%), increased plant dry weight was recorded in case of S. nobilis WA-3 (85%) as compared to water treated control plants. In case of number of leaves, significant increase was recorded with S. nobilis WA-3 (27%) and significant increase in case of number of roots were recorded in case of strain S. nobilis WA-3 (30%) as compared to control plants. Over all the study revealed that these rhizospheric PGP Streptomyces are good candidates to be developed as bioferlizers for growth promotion and yield enhancement in wheat crop and can be exploited for the commercial production of different agro-active compounds. plant growth promoting Streptomycesindole acetic acid (IAA)wheat16S rRNA gene sequencingbiofertilizersagro-active compounds ==== Body Introduction Effective farming practices, now-a-days, rely on extensive use of chemical fertilizers in order to enhance plant growth and yield. However, the cost, environmental concerns and the resulting human health hazards due to the inclusion of these chemical fertilizers in food chain are the major limiting factors. Microorganisms have been considered an important source of natural compounds of agro active importance. Use of microbial consortia in the form of bio fertilizers for reduction in the application of chemical fertilizers, pesticides and related agrochemicals, without compromising the plant yield is currently a significant research area in the field of agriculture, microbiology, and biotechnology (Ahmad et al., 2008). Plant growth promoting rhizobacteria (PGPR) is a group of naturally occurring, free living rhizosphere colonizing bacteria that improve plant growth, increase yield, enhance soil fertility, and reduce pathogens as well as biotic or abiotic stresses (Vessey, 2003; Kumar et al., 2014). PGPR help the plants by producing plant growth phytohormones such as indole acetic acid (IAA), cytokinins, and gibberellins (Marques et al., 2010), solubilization of inorganic phosphate (Jeon et al., 2003), asymbiotic nitrogen fixation (Khan, 2005), antagonistic effect against phytopathogenic microorganisms by producing siderophore, antibiotics, and fungicidal compounds (Lucy et al., 2004; Barriuso et al., 2008; Majeed et al., 2015). Actinomycetes are present extensively in the plant rhizosphere and produce various agroactive compounds. In the last few years, this group of bacteria, due to its strong antimicrobial potential, and soil dominant saprophytic nature, gained much attention as plant growth promoters (PGP; Franco-Correa et al., 2010). Actinobacteria can actively colonize plant root systems, can degrade a wide range of biopolymers by secreting several hydrolytic enzymes and tolerate hostile conditions by forming spores (Alexander, 1977). Actinobacteria, especially Streptomyces, also exhibit immense biocontrol action against a range of phytopathogens (Wang et al., 2013). Actinobacteria can produce phytohormones (IAA) and siderophore as well as solubilize phosphate and promote plant growth (Jeon et al., 2003). Actinomycetes have been mainly exploited in pharmaceutical industry since 1940s, Whereas, only a few have been developed as commercial products for plant application in agriculture (Minuto et al., 2006). Streptomycetes have been long considered simply as free-living soil inhabitants, but recently the importance of their complex interactions with plants, and other organisms is being uncovered (Seipke et al., 2011). Interest in the beneficial rhizobacteria associated with cereals has increased recently and several studies clearly demonstrated the positive and beneficial effects of PGPR on growth and yield of different crops especially wheat at different environments under variable ecological conditions (Marques et al., 2010; Mehnaz et al., 2010; Zhang et al., 2012). Wheat (Triticum aestivum) ranked third most abundant cereal crop after maize and rice and covers approximately 30% of the total cereal products worldwide (Fageria and Baligar, 1997). Pakistan is the 18 largest wheat producing country in the world (Tunio et al., 2006) with an average of 20–24 million tons of wheat grown a year. Wheat is grown on major areas in Pakistan but its average yield per hectare is far less than the actual potential (Akhtar et al., 2009). Understanding the diversity and distribution of indigenous actinobacteria in the rhizosphere of particular crops is depended on the knowledge of native actinobacterial populations, their isolation, identification, and characterization. It is therefore mandatory to explore region specific actinobacterial strains that can be used as growth promoters to achieve desired crop production (Deepa et al., 2010). However, in spite of actinobacterial high soil population, secondary metabolite production and capability to endure hostile environments, Streptomyces, and other actinobacteria are unexpectedly under explored for plant-growth promotion, as compared to Pseudomonas or Bacillus spp. (Doumbou et al., 2002). The main objectives of the present study included: to isolate indigenous actinobacteria from the wheat (Triticum aestivum) and tomato (Solanum lycopersicum) rhizosphere, characterize these isolates on the basis of morphological, and physiological characteristics as well as by 16S rRNA gene sequence analysis, to screen actinobacteria for various plant growth promoting activities (PGPAs), such as IAA production, phosphate solubilization, siderophore production, and in-vitro 1-aminocyclopropane-1-carboxylate (ACC) deaminase activity. The PGP potential of selected isolates was also studied in-vivo under axenic conditions and their effect on wheat growth was investigated. Keeping in mind, there is a lack of data on the Plant growth promotion (PGP) potential of actinomycetes isolated from Pakistan, Therefore, this study provides new and novel information. Materials and methods Collection and enrichment of soil samples The soil samples for the isolation of actinobacteria were collected from the rhizosphere of wheat (Triticum aestivum) and tomato (Solanum lycopersicum) plants cultivated in the Punjab province (District: Lahore, Gujranwala, Sheikhupura) Pakistan, during the months of December, 2013–March, 2014. Samples were collected in properly labeled sterile polythene sampling bags. The samples were subjected to physical treatment (heating, 55°C/60 min) according to the method of Seong et al. (2001) and chemical treatment; CaCO3: soil (1:10 w/w; Oskay, 2009), for the enrichment of actinobateria. Isolation and preservation of actinobacteria One Gram of soil sample was mixed in 9 ml of autoclaved water and serially diluted to a final dilution of 10−3, 10−4, and 10−5. The 0.1 ml of each dilution was spreaded on ISP-4 medium (Soluble starch 10 g, CaCO3 2 g, (NH4)2SO4 2 g, K2HPO4 1.0 g, MgSO4.7H2O 1.0 g, NaCl 1.0 g, FeSO4.7H2O 1.0 mg, MnCl2.7H2O 1.0 mg, ZnSO4.7H2O 1.0 mg, Agar 15 g, distilled water 1 L, pH = 7.5 ± 0.2, 28°C) supplemented with 25 μg/ml nalidixic acid and 50 g/ml nystatin as antibacterial and antifungal agents (Taechowisan et al., 2003). Isolated actinobacteria were sub-cultured on ISP-2 medium (Yeast extract 4 g, Glucose 4 g, Malt extract 10 g, Agar15 g, distilled water 1 L, pH = 7.5) and the plates were incubated at 28°C for 7–12 days. All the isolated strains were preserved in 25% glycerol and kept at −80°C. After incubation, 98 colonies of actinobacteria were selected on the basis of morphological characters such as distinct color and colony shape. Among them, 30 strains were detected to be exhibiting Plant growth promoting activities in initial in-vitro screening. Finally, six isolates that included WA-1, WA-3, WC-3, WD-3, TA-3, and TB-4 which exhibited most significant PGP traits were selected. Taxonomic studies Biochemical and physiological characterization A comprehensive morphological, biochemical, and physiological characterization scheme was adopted to determine the taxonomic status of the selected actinobacterial strains. Among the different biochemical characteristics studied, melanin pigment production was determined by following the methods described by Pridham and Lyons (1969) and positive results were indicated by blackening of the media. Carbohydrate utilization was performed by using ISP-9 medium supplemented with different sugars. Decomposition of oxalic acid and other organic acids was performed by using the method of Nitsch and Kutzner (1969). Esculin degradation was detected by blackening of the media after 12 days of incubation was recorded as positive result. Trysoine agar was used for determining tyrosine hydrolysis by actinomycetes (Gordon and Smith, 1953). Xanthine and hypoxanthine utilization is determined by clear zone formation around the colony after 12 days of incubation (Gordon and Smith, 1953). Starch hydrolysis was performed as described by Cowan (1974). After 12 days of incubation, plates were flooded with lugol's iodine and clear zone around colonies were recorded as positive. Urease releases ammonia from urea and the increase in the pH was detected by the indicator phenol red changing from yellow to pink (Gordon et al., 1974), Cell wall type was determined based on the isomers of diaminopimelic acid (DAP) by using the method of Becker et al. (1964). Cultural characteristics such as color of aerial and substrate mycelium and pigmentation of the selected actinobacteria was recorded on ISP-2 medium according to the method of Shirling and Gottlieb (1966). Growth at different pH 5, 6, 7, 8 was noted after 14 days of incubation on the agar plates. Tolerance to temperature was tested at 10, 28, 37, 45°C and visible growth was recorded as positive result. Agar supplemented with different NaCl Concentration 0, 4, 7, 10, 13% was used for determining NaCl tolerance of actinomycetes. In-vitro screening of actinobacteria for their plant growth promoting activities Colorimetric analysis of indole acetic acid (IAA) production Auxin form different strains of actinomycetes was quantified using the method of Tang and Borner (1979) as described by Ali et al. (2009). All Actinobacteria were grown at 28°C in ISP-2 liquid medium in triplicates for 7–12 days at 120 rpm on an orbital shaker. Media treatments were supplemented with different concentrations of L-tryptophan (0, 100, 200, 300, 400, and 500 μg/ml). Cells were removed from culture medium by centrifugation at 14,000 rpm for 15 min (Sigma 2–5, sigma Laborzentrifugen, Osterode, Germany). The supernatant (1 ml) was mixed with 2 ml of Salkowski's reagent (50 ml, 35% perchloric acid, 1 ml of 0.5 M FeCl3 solution) and was incubated at room temperature for 30 min in dark. Development of pink or red color indicates IAA production. Optical density was taken at 535 nm by using spectrophotometer (S-300D; R and M Marketing, Hounslow, UK). Standard curve of IAA was used to measure the concentration of IAA produced by the actinobacteria. Siderophore production The strains were assayed for the siderophore production on the Chrome Azurol S agar according to the method of Alexander and Zuberer (1991). CAS agar plates were prepared and spot inoculated with actinobacterial strains and incubated at 28°C for 7 days. The colonies producing yellow to orange halos were considered positive for siderophore production. Solubilization of phosphate Pikovskaya's agar plates were used for qualitative screening of all actinobacterial isolates (Gaur, 1990). Solubilization index (SI) was calculated by using the formula of Edi Premono et al. (1996). King (1932) method was used for quantitative analysis of solubilization of tri-calcium phosphate in liquid medium. Ammonia production Freshly grown actinobacterial cultures were inoculated into 1 ml of peptone water and incubated at 28°C for 7–12 days with shaking at 120 rpm. After incubation, 0.5 ml of Nessler's reagent was added in each culture tube. Development of yellow to brown color indicates positive result for ammonia production (Cappuccino and Sherman, 2002). Hydrogen cyanide production Producton of hydrogen cyanide by actinobacterial culture was evaluated by adapting the method of Lorck (1948). Actionobacteria were streaked on ISP-2 medium amended with 4.4 glycine/l and whatman filter paper No. 1 dipped in 2% sodium carbonate in 0.5% picric acid for a minute was placed underneath the petri plates lids. Plates were sealed with parafilm and incubated at 28°C for 7–12 days. Orange to red color of filter was indicative of HCN production. Colorimetric ninhydrin assay for screening ACC utilizing actinobacteria Li et al. (2011) method was used. ISP-2 medium was inoculated with actinobacteria, incubated at 28°C with shaking 200 rpm for 7–12 days. After centrifugation of 2 ml of culture at 14,000 rpm for 5 min, supernatant was discarded, pellet washed with 1 ml of liquid DF- medium, later mixed with 2 ml of ACC substrate containing DF mineral medium. 2 ml of DF-ACC medium without inoculation served as control. All the samples were incubated for 4 days at 200 rpm. After incubation, 1 ml of each bacterial culture was centrifuged at 14,000 rpm for 5 min. 100 μl of each supernatant was shifted to another tube and was diluted with 1 ml of liquid DF medium. In 96-well PCR plate, 60 μl of each diluted supernatant was mixed 120 μl of ninhydrin reagent, covered with parafilm, and placed in boiling water bath for 30 min. DF medium was used as a blank. The resulted purple color depth was record for visual comparison of actinobacterial strains. 100 μl of the reaction mixture was transfer to the microtitre plate in triplicates and absorbance was taken at 570 nm with spectrophotometer. The actinobacterial isolate with visibly reduced color depth and less supernatant absorbance compared to DF-ACC medium without inoculation were considered as ACC utilizing actinobacterial strains. In-vivo screening for plant growth promotion activities Actinobacterial isolates were grown on ISP-2 broth at 28°C for 7–12 days with continuous shaking 200 rpm. Actinobacterial suspensions were prepared according to the method of Errakhi et al. (2007). Wheat seeds were surface sterilized by using the method of Khalid et al. (2004). Sterilized seeds were soaked for 1 h in the suspension and dried under laminar flow hood overnight. For control, seeds were dipped in distilled water only. Germination bioassay Germination assay was performed by using actinobacterial treated wheat (Triticum aestivum) seeds (three replicates, 5 seeds/plate) were placed in sterilized petri dishes covered with two sheets of filter papers, moistened with 10 ml of sterile distilled water. Water treated control seeds were used. All the petri dishes were incubated in Versatile Environmental Test Chamber (TEMI 850 WISE CUBE) with light intensity of 2000 lux for 16 h daily at 28°C. After 2 weeks, effect of actinobacterial cultures on root length and number of roots was observed. Pot experiment Actinobacterial treated eight wheat seeds (as described above) were sown to a depth of 1 cm in plastic pots (12 cm high × 10 cm diameter) filled with sterilized soil. Three replicates were used for actino-bacterial as well as for control treatments. For control, seeds were dipped in distilled water only. Pots were arranged in fully randomized complete block, under standard conditions, in plant growth room (22–26°C, 16 h light/8 h dark). All pots were watered daily with 10 ml of distilled autoclaved water to achieve moisture level sufficient for seed germination. After 25 days, five plants were removed carefully from the soil, washed with tap water to remove soil particles. Total 15 plants per treatment were considered for statistical analysis. Data was recorded for number of seeds germinated, root and shoot length, root and shoot fresh weight, root, and shoot dry weight (80°C, 24 h electric oven) and number of roots. Root colonization potential of inoculated Streptomyces was determined at every 8 days by using serial dilution plating technique on GYM agar and number of viable cells was recorded as colony forming units (CFU) as described in Somasegaran and Hoben (1994). Genomic DNA isolation, PCR amplification and sequencing of the 16s rRNA gene Total genomic DNA was isolated according to the CTAB method described by the Liu et al. (2000). Universal Primers 27F 5′ AGAGTTTGATCMTGGCTCAG 3 and 1492 R′ TACGGYTACCTTGTTACGACTT 3′ were used for the PCR amplification of 16S rRNA gene of the selected strains. Agarose gel electrophoresis (1%) was used for analyzing PCR product and remaining mixture was purified by using PCR Purification kit (FavorPrep™). Purified PCR products were sequenced commercially by GATC Biotech (Germany). The obtained gene sequences were compared with others in the Gen Bank databases using the NCBI Nucleotide BLAST at http://blast.ncbi.nlm.nih.gov/Blast.cgi. Sequences were submitted to NCBI GenBank data base and accession numbers were obtained. Statistical analysis For all experiments, the data were subjected to statistical analysis using software IBM SPSS Statistics version 21. Data were subjected to analysis of variance (ANOVA) and means separated using Duncan's multiple range test (P = 0.05). The correlation coefficients between bacterial auxin production and L-tryptophan concentrations as well as between bacterial growth traits and plant growth parameters were also calculated (P = 0.01 or P = 0.05). Results Taxonomic characteristics of the selected rhizopsheric actinomycetes Ninty eight actinomycetes were isolated from six different wheat and four different tomato rhizospheric soil samples. All of them were found to be gram positive filamentous rods. On the basis of color of aerial mycelium they were grouped into gray (TA-3, TB-4), yellow (WC-3, WD-3), green (WA-1), and orange (WA-3) color series. Diffusible pigment was produced by the isolate WC-3. All strains contained LL-diaminopimelic acid isomer in their cell wall. Physiologically, most of the actinomycetes isolates were able to utilize different sugars as the carbon source. All strains were able to utilize glucose, xylose, galactose, arabinose, mannose, and mannitol. Strain WA-1 and TA-3 were unable to utilize sucrose, while strain WC-3, and WD-3 were unable to utilize raffinose as their carbon source (Table 1). Most of them were able to grow at temperature 28 and 37°C. None of them were able to grow at temperature 4°C. Optimum temperature and pH was found to be 28°C and eight respectively (Table 2; Tables S1, S2, Figures S1, S2). Table 1 Morphological, biochemical, and physiological characteristics of the selected rhizospheric actinomycete strains. Characteristics Actinomycete isolates WA-1 WA-3 WC-3 WD-3 TA-3 TB-4 Colony diameter (cm) 0.6 0.3 0.3 0.4 0.5 0.2 Color of aerial mycelium Green Orange Light yellow Pale yellow Platinum gray Dusty gray Color of substrate mycelium Beige yellow Orange red Olive yellow Light brown Clay brown Clay brown Diffusible pigment − − Beige − − − Melanoid pigment Black Black − Black − Black CARBON SOURCE UTILIZATION Glucose + + + + + + Xylose − − − + + + Galactose + + + + + + Sucrose − + + + − + Mannose + + + + + + Arabinose + + − + + − Mannitol + + + + + + Raffinose + + − − + + HYDROLYSIS OF Tyrosine + + − + − + Xanthine + − + + − + Hypoxanthine + + + + − + Starch + + + + + + Urea + + + + + + Decomposition of oxalic acid + − − + + + Degradation of esculin + − − + − + The symbol, + representing the positive reaction/presence of growth while symbol, − represents the negative reaction/absence of growth. Table 2 Temperature (°C), pH and NaCl (%) tolerance of the selected rhizospheric actinomycete strains. Strains Temperature tolerance pH tolerance NaCl tolerance 10°C 28°C 37°C 45°C 5 6 7 8 0% 4% 7% 10% 13% WA-1 + + + + − − + + + + + + − WA-3 − + + + − + + + + + + − − WC-3 − + + − − + + + + − − − − WD-3 + + + + − + + + + + + + − TA-3 − + + − − − + + + + − − − TB-4 − + + − − + + + + + + − − The symbol, + represents the positive reaction/presence of growth while symbol, − represents the negative reaction/absence of growth. IAA production by selected actinomycetes Qualitative analysis of culture supernatant of selected actinomycete isolates revealed production of variable amounts of IAA in the absence and presence of different concentrations of tryptophan (100–500 μg/ml). In the absence of L-tryptophan, IAA production was not observed. With increasing concentration of L-tryptophan, the IAA production was increased. For instance, isolate Streptomyces sp. WA-1(r = 0.979, P = 0.01), S. nobilis WA-3 (r = 0.942, P = 0.01), S. kunmingensis WC-3 (r = 0.995, P = 0.01), S. mutabilis WD-3 (r = 0.932, P = 0.01), S. enissocaesilis TA-3 (r = 0.971, P = 0.01), and S. djakartensis TB-4 (r = 0.919, P = 0.05) showed significant positive correlation with increasing L-tryptophan concentrations. Isolate WA-3 produced highest amount of IAA followed by WC-3, TA-3, WD-3, WA-1, and TB-4. The most active IAA producer S. nobilis WA-3, S. Kunmingenesis WC-3 and S. enissocaesilis TA-3 produce 79.5, 79.23, and 69.26 μg/ml IAA respectively at 500 μg/ml L-tryptophan (Figure 1). Figure 1 Effect of different L-tryptophan concentrations (100, 200, 300, 400, and 500 μg/ml) on IAA production by actinomycetes. The results shown are representative of three repetitions of the experiment. Value “r” indicates highly significant positive correlation between L-tryptophan and bacterial IAA production. **P = 0.01, *P = 0.05. Phosphate solubilization Among the selected isolates, six Streptomyces were able to solubilize phosphate by producing clear zones around the colonies after 7 days of incubation. Streptomyces sp. WA-1 showed highest phosphate solubilization index of 2.33 followed by S. djakartensis TB-4 (2.27) > S. enissocaesilis TA-3 (2.25) > S. nobilis WA-3 (2.22) > S. mutabilis WD-3 (2.21) > S. kunmingensis WC-3 (2.20). Quantitatively, the highest concentration of soluble phosphate was produced by Streptomyces sp. WA-1 (72.13 mg/100 ml) and S. djakartensis TB-4 (70.36 mg/100 ml; Table 3). Table 3 Plant growth promoting traits of the selected rhizospheric actinomycetes under in-vitro conditions. Strains Phosphate solubilization ACCb deaminase concentration (mmol/l) Ammonia production HCNc production Siderophore production SIa Soluble P concentration (mg/100 ml) WA-1 2.33 72.13 ± 0.336 1.65 ± 0.42 d + + + WA-3 2.22 66.0 ± 0.40 1.29 ± 0.45 b + + + WC-3 2.20 61.46 ± 0.279 1.59 ± 0.38 c + + + WD-3 2.21 57.83 ± 0.32 1.9 ± 0.35 f + + + TA-3 2.25 62.9 ± 0.40 0.71 ± 0.16 a + + + TB-4 2.27 70.36 ± 0.33 1.81 ± 0.39 e + + + a SI, Solubilization index = the ratio of the total diameter (colony + halo zone) to the colony diameter. b ACC, 1-aminocyclopropane-1-carboxylate (ACC) deaminase. Different letters on bars indicate significant difference between treatments, using Duncan's multiple range test (P = 0.05). c HCN, Hydrogen cyanide. The symbol, + represents the positive reaction/presence of trait while symbol, − represents the negative reaction/absence of trait. Siderophore, ammonia, and HCN production Among the selected isolates, six rhizospheric actinomycetes were positive for the ammonia production. Siderophore production was detected in all six isolates on CAS agar media, forming clear orange halo zone around the colonies. (Figure 2). All the six isolates were positive for HCN production (Table 3). Streptomyces sp. WA-1 and S. djakartensis TB-4 displayed the highest amount of HCN production as depicted by a very deep red color on the filter paper (Table 3). Figure 2 Screening of siderophore producing actinomycete isolates using CAS agar plates after 7 days of growth at 28 ± 2°C. The arrows indicate the halo zone around the colonies of isolate WA-1 (Streptomyces sp.) and TB-4 (Streptomyces djakartensis). ACC deaminase activity by actinomycetes All strains showed variable concentration of ACC-deaminase in the culture media. Strain S. mutabilis WD-3 showed highest concentration of ACC-deaminase 1.9 mmol /l). On the other hand, strain S. djakartensis TB-4, Streptomyces sp. WA-1, S. kunmingensis WC-3, S. nobilis WA-3 produce 1.81, 1.65, 1.59, 1.29 mmol/l concentration of ACC-deaminase, respectively. Strain S. djarkartensis TA-3 produce the lowest concentration (0.71) of ACC deaminase (Table 3). Seed germination assay Inoculation of wheat seeds with selected actinomycetes stimulated root growth in majority of the strains (Figure 3). In case of root length, significant increases were recorded for S. djakartensis TB-4 (61%), S. Kunmingenesis WC-3 (47%) and S. mutabilis (42%). Increases in number of roots were also observed with S. mutabilis WD-3 (55%), Streptomyces sp. WA-1 (50%), and S. djakartensis TB-4 (41%). Figure 3 Effect of actinomycetes spore suspension on seed germination of Triticum aestivum. Bars represents mean ± SE of three replicates (15 plants). Different letters on bars indicate significant difference between treatments, using Duncan's multiple range test (P = 0.05). Root colonization potential Actinobacterial population size was estimated by plate count method on GYM agar at three (8, 16, 24 days) different time intervals. It was observed that all rhizospheric actinobacteria were able to colonize wheat plant roots and displayed persistence in the rhizosphere up to 25 days after inoculation (Figure 4). Maximum colonization was recorded between 8 and 16 days post inoculation. Actinobacterial isolate WA-3 showed maximum number of root colonizing colonies at all times as compared to other actinobacterial isolates. Figure 4 Population density of different actinobacterial isolates inoculated to wheat at different time intervals under axenic conditions. Plant growth promotion experiment with selected actinomycetes Wheat seeds inoculated with six selected actinomycetes showed variable growth and yield parameters after 3 weeks growth (Table 4). Significant increases in shoot length were observed with S. nobilis WA-3 (65%), S. djakartensis TB-4 (54%), and S. enissocaesilis TA-3 (53%) as compared to water treated control. Significant increase in root length was recorded with S. nobilis WA-3 (81%), S. djakartensis TB-4 (69%), and S. enissocaesilis TA-3 (51%) as compared to water treated control. Maximum increases in plant fresh weight were recorded with S.nobilis WA-3 (84%) and S. djakartensis TB-4 (75%), increased plant dry weight was recorded in case of S. nobilis WA-3(85%), S. djakartensis TB-4 (66%), and S enissocaesilis TA-3 (60%). In case of number of leaves, significant increases were recorded with S. nobilis WA-3 (27%), S. djakartensis TB-4 (27%), and S. enissocaesilis TA-3 (23%). Significant increase in case of number of roots were recorded in case of S. nobilis WA-3 (30%), S. djakartensis TB-4 (26%), and S. enissocaesilis TA-3 (22%) as compared to water treated control (Figure 5). Table 4 Effect of selected actinomycetes on the growth of Wheat (Triticum aestivum). Isolates Growth parameters of actinomycetes treated wheat plants Root length (cm) Shoot length (cm) No. of leaves No. of roots Seedling fresh weight (g) Seedling dry weight (g) Control 13.1 ± 1.4a 116.2 ± 0.7 a 3.13 ± 0.1 a 4.3 ± 0.6 a 11.4 ± 0.7 a 2.32 ± 0.1a WA-1 15.0 ± 10.9ab 21.7 ± 1.1b 3.4 ± 0.1ab 4.6 ± 0.3ab 13.9 ± 0.6bc 2.81 ± 0.2b WA-3 23.8 ± 1.2f 26.8 ± 0.3c 4 ± 0d 5.6 ± 0.2b 21.0 ± 0.5d 4.33 ± 0.1d WC-3 16.0 ± 0.6bc 22.0 ± 0.7b 3.46 ± 0.1ab 4.93 ± 0.2ab 14.3 ± 0.4b 3.08 ± 0.1b WD-3 18.4 ± 0.4cd 22.4 ± 1.0b 3.53 ± 0.1bc 5 ± 0.4ab 15.5 ± 0.7b 3.17 ± 0.1b TA-3 19.9 ± 0.4de 24.8 ± 1.0c 3.86 ± 0.09cd 5.26 ± 0.4ab 17.4 ± 0.5c 3.72 ± 0.06c TB-4 22.2 ± 0.6ef 25.0 ± 0.5c 4 ± 0d 5.46 ± 0.1ab 20.0 ± 0.5d 3.87 ± 0.2cd Control, non-inoculated. Data is the mean of 15 plants obtained from three replicates. ± Symbol indicate values for standard error of means. Different letters indicate significant difference between treatments, using Duncan's multiple range test (P = 0.05). Figure 5 Effect of seed treatment with actinomycetes spore suspensions on plant growth promotion of Triticum aestivum. (A) Seedling root and shoot length (cm) (B), seedling fresh and dry weight (g) (C) number of leaves and roots per plant. Evaluation was made 30 days after planting. Bars represents mean ± SE of three replicates (15 plants). Different letters on bars indicate significant difference between treatments, using Duncan's multiple range test (P = 0.05). Identification of selected actinomycetes strains by 16s rRNA gene sequencing Six of them including WA-1, WA-3, WC-3, WD-3, TA-3, and TB-4 were selected based on their ability to produce phytohormone IAA, solubilization of inorganic phosphates, ACC deaminase activity, siderophore, ammonia, and hydrogen cyanide production. For all the selected actinomycetes, single band PCR product of 1.5 kb length were achieved with the universal primers (Figure 6). The sequences of 16S rRNA gene were analyzed by comparison with sequences in GenBank through Nucleotide BLAST (http://www.ncbi.nlm.nih.gov/BLAST). After comparison, strains WA-1 showed 98% similarity with Streptomyces sp. while the strains WA-3, WC-3, WD-3, TA-3, and TB-4 showed 99% similarity with Streptomyces nobilis, Streptomyces kunmingenesis, Streptomyces mutabilis, Streptomyces enissocaesilis, Streptomyces djakartensis, respectively. The sequences from strains WA-1, WA-3, WC-3, WD-3, TA-3, and TB-4 have been deposited in the GenBank and accession numbers were obtained (Table 5). Figure 6 Gel electrophoresis of PCR products for detection of Actinomycetes isolates. Lane 1: DNA ladder (Fermentas, Germany); lane 2, 3, 4, 5, 6, and 7: positive samples of Streptomyces sp. (WA-1), S. nobilis (WA-3), S. kunmingenesis (WC-3), S. mutabilis (WD-3), Streptomyces enissocaesilis (TA-3), and S. djakartensis (TB-4) respectively (1.5 kb). Table 5 Genetic similarity of the selected strains with different species of the genus Streptomyces determined by 16S rRNA gene sequencing. Strains Source of isolation, Plant Identified as Similarity (%) Accessions WA-1 Rhizophere soil, Triticum aestivum Streptomyces sp. 98 KU750781 WA-3 Rhizophere soil, Triticum aestivum Streptomyces nobilis 99 KU750782 WC-3 Rhizophere soil, Triticum aestivum Streptomyces kunmingensis 99 KU750783 WD-3 Rhizophere soil, Triticum aestivum Streptomyces mutabilis 99 KU750784 TA-3 Rhizopshere soil, Solanum lycopersicum Streptomyces enissocaesilis 99 KU750780 TB-4 Rhizopshere soil, Solanum lycopersicum Streptomyces djakartensis 99 KU750779 Discussion In this study, a total of 98 actinomycetes strains were isolated from wheat and tomato rhizospheric soil. All of the isolates were screened for different PGP traits and 30 of them were found to be showing excellent PGP activities in initial screening. Out of these, six most promising actinobacterial strains belonging to the genus Streptomyces were investigated by in-vivo studies. The 16S rRNA gene sequence of these six selected actinomycetes strains including WA-1, WA-3, WC-3, WD-3, TA-3, and TB-4 showed maximum sequence similarity with members of the genus Streptomyces. Majority of the PGPR actinomycetes synthesize IAA which is responsible for increased number of adventitious roots which help plant to uptake a large volume of nutrients and absorb water, while increased root exudates in turn benefits the bacteria (El-Tarabily, 2008). In our study, the IAA production ranges between 10 and 79.5 μg/ml and Streptomyces nobilis strain WA-3 was detected as the best strain for the IAA production (79.5 ug/ml) that exceeds the level of previously reported work by Khamna et al. (2009). Abd-Alla et al. (2013) reported Streptomyces sp. CMU-MH021 that could produce 28.5 μg/ml of IAA. Several Streptomyces species, such as S. olivaceoviridis, S. rimosus, S. Rochei, S. griseoviridis, and S. lydicus have the ability to produce IAA and improve plant growth by increasing seed germination, root elongation and root dry weight (Mahadevan and Crawford, 1997). Soil acidification often resulted due to the growth of phosphate-solubilizing bacteria (PSB), which in turn resulted in phosphorus solubilization. Therefore, PSB are well known as solubilizers of inorganic phosphate (Verma et al., 2001). The maximum phosphate solubilization activity was detected in the strain Streptomyces sp. WA-1 (72.13 mg/100 ml). Hamdali et al. (2008) reported 83.3, 58.9, and 39 mg/100 ml phosphate solubilization by Streptomyces cavourensis, Streptomyces griseus, and Micromonospora aurantiaca, respectively. Almost all of the rhizospheric actinomycetes were also able to produce ammonia and hydrogen cyanide. Marques et al. (2010) recommended that bacteria can synthesize ammonia and supply nitrogen to the host plant. Additionally, over production of ammonia serve as a prompting factor for the virulence of opportunistic plant pathogens. Ammonia production also plays an important role HCN production play an essential role in suppression of plant disease. In this study, all the ammonia and HCN producing isolates belong to the genus Streptmyces. Similarly, Husen et al. (2011) reported Streptomyces sp. LSW05 strain as a potent HCN producer. Siderophore production is one more feature that stimulates plant growth by forming complex with iron form (Fe 3+) in the rhizosphere making iron unavailable to the phytopathogens. It is suggested by Tan et al. (2009) siderophore production that the production of siderophore is an important factor for phytopathogen antagonism and developing growth of the plant. In our study, we detected the 85.7% isolates positive for siderophore production. Similarly, Khamna et al. (2010) has revealed that Streptomyces CMU-SK 126 isolated from Curcuma mangga rhizospheric soil exhibited high amount of siderophore. ACC deaminase-producing bacteria have been known to promote plant growth by decreasing ethylene inhibition of various plant processes (Husen et al., 2011). They can increase root growth by lowering endogenous ACC levels (Glick, 2005). Plant roots must be able to perceive and recognize such elicitors in ways similar to the recognition of elicitors from plant pathogens. In fact, plant pathogens might interfere with the action of PGPR by being perceived by similar receptors (Husen, 2003). The effect of soil microbes on PGP including root development has been reported by Uphoff et al. (2009). In the present study, root elongation assay and pot experiment performed by using wheat seeds inoculated with PGP Streptomyces strains, significantly enhanced the plant growth by increasing plant root length, plant shoot length, dry weight, fresh weight, number of leaves, and number of roots over the un-inoculated control. The Streptomyces strains are extensively reported in the literature for its PGP potential (Nassar et al., 2003; El-Tarabily, 2008; Gopalakrishnan et al., 2011b). As hypothesized earlier, the mechanism by which the Streptomyces enhanced the morphological and yield parameters on both sorghum and rice could be their PGP traits including IAA and siderophore production (direct stimulation of PGP; Gopalakrishnan et al., 2011b). Conclusion The study revealed that these rhizospheric actinomycetes are potential microbial inoculants because of their intensified PGP activities such as IAA production, phosphate solubilization, siderophore, and HCN production and ACC deaminase production. The strains reported in this study are promising candidates to be developed as commercial biofertilizer formulation and can also be exploited for the production of various agroactive compounds like auxins etc. As such this is the first comprehensive report from Pakistan about the PGP traits and potential agricultural applications of actinomycetes. Author contributions SA: Isolation, biochemical characterization, identification and in-vitro and in-vivo screening of actinomycetes for multiple PGP traits/hormones. Did all the experimental work. IS: Supervise in the sampling of the rhizospheric soils. Teach how to isolate actinomycetes by using enrichment, help in DNA isolation and PCR purification technique. BA: Co-supervise in performing PGP experiments. Both in-vitro and in-vivo screening of actinomycetes. Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The financial support for this study by Higher Education Commission (HEC) of Pakistan, under indigenous Ph.D. fellowships program is gratefully acknowledged. Supplementary material The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb.2016.01334 Click here for additional data file. ==== Refs References Abd-Alla M. H. El-Sayed E. S. A. Rasmey A. H. M. (2013 ). Indole-3-acetic acid (IAA) production by Streptomyces atrovirens . J. Biol. Earth Sci . 3 , 82 –93 . Ahmad F. Ahmad A. I. Khan M. S. (2008 ). Screening of free-living rhizospheric bacteria for their multiple plant growth promoting activities . Microbiol. Res . 163 , 173 –181 . 10.1016/j.micres.2006.04.001 16735107 Akhtar M. J. Asghar H. N. Shahzad K. Arshad M. (2009 ). Role of plant growth promoting rhizobacteria applied in combination with compost and mineral fertilizers to improve growth and yield of wheat (Triticum Aestivum L.) . Pak. J. Bot . 41 , 381 –390 . Alexander D. B. Zuberer D. A. (1991 ). Use of Chrome Azurol S reagents to evaluate siderophore production by rhizosphere bacteria Biol . Fertility Soils. 12 , 39 –45 . 10.1007/BF00369386 Alexander M. (1977 ). Introduction to Soil Microbiology . 2nd Edn. , Malabar, FL : Krieger Publishing Company . Ali B. Sabri A. N. Ljung K. Hasnain S. (2009 ). Quantification of indole-3-acetic acid from plant associated Bacillus spp. and their phytostimulatory effect on Vigna radiata (L.) . World J. Microbiol. Biotechnol. 25 , 519 –526 . 10.1007/s11274-008-9918-9 Barriuso J. Ramos Solano B. Gutiérrez Mañero F. J. (2008 ). Protection against pathogen and salt stress by four plant growth-promoting rhizobacteria isolated from Pinus sp. on Arabidopsis thaliana . Biol. Control. 98 , 666 –672 . 10.1094/PHYTO-98-6-0666 18944290 Becker B. Lechevalier M. P. Gordon R. E. Lechevalier H. A. (1964 ). Rapid differentiation between Nocardia and Streptomyces by paper chromatography of whole-cell hydrolysates . Appl. Microbiol. 12 , 421 –423 . 14215972 Cappuccino J. G. Sherman N. (2002 ). Microbiology: A Laboratory Manual , 6th Edn. California City, CA : Pearson Education . Cowan S. T. (1974 ). Cowan and Steel's Manual for the Identification of Medical Bacteria , 2nd Edn London : Cambridge University Press . Deepa C. K. Dastager S. G. Pandey A. (2010 ). Isolation and characterization of plant growth promoting bacteria from non-rhizospheric soil and their effect on cowpea (Vigna unguiculata (L.) Walp.) seedling growth . World J. Microbiol. Biotechnol. 26 , 1233 –1240 . 10.1007/s11274-009-0293-y 24026928 Doumbou C. L. Hamby-Salove M. K. Crawford D. L. Beaulieu C. (2002 ). Actinomycetes, promising tools to control plant diseases and to promote plant growth . Phytoprotection 82 , 85 –102 . 10.7202/706219ar Edi Premono M. Moawad A. M. Vlek P. L. G. (1996 ). Effect of phosphate-solubilizing Pseudomonas putida on the growth of maize and its survival in the rhizosphere. Indones . J. Crop Sci . 11 , 13 –23 . El-Tarabily K. A. (2008 ). Promotion of tomato (Lycopersicon esculentun Mill.) plant growth by rhizosphere competent 1-aminocyclopropane-1-carboxylic acid deaminase–producing Streptomycete actinomycetes . Plant Soil. 308 , 161 –174 . 10.1007/s11104-008-9616-2 Errakhi R. Bouteau F. Lebrihi A. Barakate M. (2007 ). Evidences of biological control capacities of Streptomyces spp. against Sclerotium rolfsii responsible for damping-off disease in sugar beet (Beta vulgaris L.) . World J. Microbiol. Biotechnol . 23 , 1503 –1509 . 10.1007/s11274-007-9394-7 Fageria N. K. Baligar V. C. (1997 ). Integrated plant nutrient management for sustainable crop production—An over . Inter. J. Trop. Agri . 15 , 7 –18 . Franco-Correa M. Quintana A. Duque C. Suarez C. Rodrıguez M. X. Barea J. M. (2010 ). Evaluation of actinomycete strains for key traits related with plant growth promotion and mycorrhiza helping activities . Appl. Soil Ecol. 45 , 209 –217 . 10.1016/j.apsoil.2010.04.007 Gaur A. C. (1990 ). Phosphate Solubilizing Microorganisms as Biofertilizer . New Delhi : Omega Scientific Publishers . Glick B. R. (2005 ). Modulation of plant ethylene levels by bacterial enzyme ACC deaminase . FEMS Microbiol. Lett . 251 , 1 –7 . 10.1016/j.femsle.2005.07.030 16099604 Gopalakrishnan S. Pande S. Sharma M. Humayun P. Kiran B. K. Sandeep D. (2011b ). Evaluation of actinomycete isolates obtained from herbal vermicompost for biological control of Fusarium wilt of chickpea . Crop Protect. 30 , 1070 –1078 . 10.1016/j.cropro.2011.03.006 Gordon R. E. Barnett D. A. Handerhan J. E. Pang C. H. (1974 ). Nocardia coeliaca, Nocardia autotrophica, and the nocardin strain . Int. J. Syst. Bacteriol. 24 , 54 –63 . 10.1099/00207713-24-1-54 Gordon R. E. Smith M. M. (1953 ). Rapidly growing, acid fast bacteria . J. Bacteriol. 66 , 41 –48 . 13069464 Hamdali H. Hafidi M. Virolle M. J. Ouhdouch Y. (2008 ). Growth promotion and protection against damping-off of wheat by two rock phosphate solubilizing actinomycetes in a P-deficient soil under greenhouse conditions . Appl. Soil Ecol. 40 , 510 –517 . 10.1016/j.apsoil.2008.08.001 Husen E. (2003 ). Screening of soil bacteria for plant growth promotion activities in vitro . Indo. J. Agric. Sci. 4 , 27 –31 . Husen E. Wahyudi A. T. Suwanto A. Giyanto (2011 ). Growth enhancement and disease reduction of soybean by 1-aminocyclopropane-1-carboxylate deaminase-producing Pseudomonas . Am. J. Appl. Sci. 8 , 1073 –1080 . 10.3844/ajassp.2011.1073.1080 Jeon J. S. Lee S. S. Kim H. Y. Ahn T. S. Song H. G. (2003 ). Plant growth promotion in soil by some inoculated microorganisms . J. Microbiol. 41 , 271 –276 . Khalid A. Arshad M. Zahir Z. A. (2004 ). Screening plant growth-promoting rhizobacteria for improving growth and yield of wheat . J. Appl. Microbiol. 96 , 473 –480 . 10.1046/j.1365-2672.2003.02161.x 14962127 Khamna S. Yokota A. Lumyong S. (2009 ). Actinomycetes isolated from medicinal plant rhizosphere soils: diversity and screening of antifungal compounds, indole-3-acetic acid and siderophore production . World J. Microbiol. Biotechnol . 25 , 649 –655 . 10.1007/s11274-008-9933-x Khamna S. Yokota A. Peberdy J. F. Lumyong S. (2010 ). Indole-3-acetic acid production by Streptomyces sp. isolated from some Thai medicinal plant rhizosphere soils . Eurasia J. Biosci. 4 , 23 –32 . 10.5053/ejobios.2010.4.0.4 Khan A. G. (2005 ). Role of soil microbes in the rhizospheres of plants growing on trace metal contaminated soils in phytoremediation . J. Trace Elem. Med. Biol. 18 , 355 –364 . 10.1016/j.jtemb.2005.02.006 16028497 King J. E. (1932 ). The colorimetric determination of phosphorus . Biochem. J. 26 , 292 –297 . 16744823 Kumar A. Maurya B. R. Raghuwanshi R. (2014 ). Isolation and characterization of PGPR and their effect on growth, yield and nutrient content in wheat (Triticum aestivum L.) . Biocatal. Agric. Biotechnol. 3 , 121 –128 . 10.3389/fmicb.2015.00198 Li J. Zhao G. Z. Huang H. Y. Zhu W. Y. Lee J. C. Xu L. H. . (2011 ). Nonomuraea endophytica sp. nov., an endophytic actinomycete isolated from Artemisia annua L. Int. J. Syst. Evol. Microbiol. 61 , 757 –761 . 10.1099/ijs.0.022558-0 20435751 Liu D. Coloe S. Baird R. Pedersen J. (2000 ). Rapid mini-preparation of fungal DNA for PCR . J. Clin. Microbiol. 38 :471 . 10681211 Lorck H. (1948 ). Production of hydrocyanic acid by bacteria . Physiol. Plant. 1 , 142 –146 . 10.1111/j.1399-3054.1948.tb07118.x Lucy M. Reed E. Glick B. R. (2004 ). Application of free living plant-promoting rhizobacteria . Antonie van Leeuwenhoek. 86 , 1 –25 . 10.1023/B:ANTO.0000024903.10757.6e 15103234 Mahadevan B. Crawford D. L. (1997 ). Properties of the chitinase of the antifungal biocontrol agent Streptomyces lydicusWYEC108 . Enzyme Microb. Technol. 20 , 489 –493 . Majeed A. Abbasi M. K. Hameed S. Imran A. Rahim N. (2015 ). Isolation and characterization of plant growth-promoting rhizobacteria from wheat rhizosphere and their effect on plant growth promotion . Front. Microbiol. 6 :198 . 10.3389/fmicb.2015.00198 25852661 Marques A. P. G. C. Pires C. Moreira H. Rangel A. O. S. S. Castro P. M. L. (2010 ). Assessment of the plant growth promotion abilities of six bacterial isolates using Zea mays as indicator plant . Soil Biol. Biochem. 42 , 1229 –1235 . 10.1016/j.soilbio.2010.04.014 Mehnaz S. Baig D. N. Lazarovits G. (2010 ). Genetic and phenotypic diversity of plant growth promoting rhizobacteria isolated from sugarcane plants growing in Pakistan . J. Microbiol. Biotechnol. 20 , 1614 –1623 . 21193815 Minuto A. Spadaro D. Garibaldi A. Gullino M. L. (2006 ). Control of soilborne pathogens of tomato using a commercial formulation of Streptomyces griseoviridis and solarization . Crop. Prot. 25 , 468 –475 . 10.1016/j.cropro.2005.08.001 Nassar A. H. El-Tarabily K. A. Sivasithamparam K. (2003 ). Growth promotion of bean (Phaseolus vulgaris L.) by a polyamine producing isolate of Streptomyces griseoluteus . Plant Growth Reg. 40 , 97 –106 . 10.1023/A:1024233303526 Nitsch B. Kutzner H. J. (1969 ). Egg-yolk agar as a diagnostic medium for Streptomyces . Cell Mol. Life Sci. 25 , 220 –221 . 10.1007/BF01899136 Oskay M. (2009 ). Comparison of Streptomyces diversity between agricultural and non- agricultural soils by using various culture media . Sci. Res. Essay. 10 , 997 –1005 . Pridham T. G. Lyons A. J. (1969 ). Progress in clarification of the taxonomic and nomenclatural status of some problem actinomycetes . Dev. Ind. Microbiol. 10 , 183 –221 . Seipke R. F. Barke J. Brearley C. (2011 ). A single Streptomyces symbiont makes multiple antifungals to support the fungus farming ant Acromyrmex octospinosus , PLoS ONE 6 :e22028 . 10.1371/journal.pone.0022028 21857911 Seong C. H. Choi J. H. Baik K. S. (2001 ). An improved selective isolation of rare actinomycetes from forest soil . J. Microbiol. 17 , 23 –39 . Shirling E. B. Gottlieb D. (1966 ). Methods for characterization of Streptomyces species . Int. J. Syst. Bacteriol. 16 , 313 –340 . 10.1099/00207713-16-3-313 Somasegaran P. Hoben H. J. (1994 ). Handbook for Rhizobia. Methods in Legume–Rhizobium Technology. Heidelberg, NY : Springer . Taechowisan T. Peberdy J. F. Lumyong S. (2003 ). Chitinase production by endophytic Streptomyces aureofaciens CMUAc 130 and its antagonism against phytopathogenic fungi . Ann. Microbiol . 53 , 447 –461 . Tan H. Deng Z. Cao L. (2009 ). Isolation and characterization of actinomycetes from healthy goat faeces , Lett. Appl. Microbiol . 49 , 248 –253 . 10.1111/j.1472-765X.2009.02649.x 19552775 Tang W. Y. Borner J. (1979 ). Enzymes involved in synthesis and breakdown of indoleacetic acid , in Modern Methods of Plant Analysis , Vol. 7 , eds Paech K. Tracey M. V. (Gohingen; Heidelberg : Springer Verlag ), 238 –241 . Tunio S. D. Korejo M. N. Jarwar A. D. Waggan M. R. (2006 ). Studies on indegenious and exotic weed competition in wheat . Pak. J. Agri. Biol. 5 , 1 –8 . Uphoff N. Anas I. Rupela O. P. Thakur A. K. Thyagarajan T. M. (2009 ). Learning about positive plant-microbial interactions from the system of rice intensification (SRI) . Aspect. Appl. Biol. 98 , 29 –54 . Verma K. Kukreja D. V. Pathak S. Suneja N. Narula (2001 ). In vitro production of plant growth regulators (PGRs) by Azorobacter chroococcum . Indian J. Microbiol. 41 , 305 –307 . Vessey J. K. (2003 ). Plant growth promoting rhizobacteria as biofertilizers . Plant Soil. 255 , 571 –586 . 10.1023/A:1026037216893 Wang X. J. Zhang J. Wang J. D. Qian P. T. Liu C. X. Xiang W. S. (2013 ). Novel Novel cyclopentenone derivatives produced by a rare actinobacterial strain Actinoalloteichus nanshanensis sp. nov. NEAU 119 . Nat. Prod. Res. 27 , 1863 –1869 . 10.1080/14786419.2013.771349 23432091 Zhang J. Liu J. Meng L. Ma Z. Tang X. Cao Y. (2012 ). Isolation and characterization of plant growth-promoting rhizobacteria from wheat roots by wheat germ agglutinin labeled with fluorescein isothiocyanate . J. Microbiol. 50 , 191 –198 . 10.1007/s12275-012-1472-3 22538646

PMC005xxxxxx/PMC5002419.txt

==== Front Front PharmacolFront PharmacolFront. Pharmacol.Frontiers in Pharmacology1663-9812Frontiers Media S.A. 10.3389/fphar.2016.00279PharmacologyPerspectiveEvaluation of Two Commercially Available Cannabidiol Formulations for Use in Electronic Cigarettes Peace Michelle R. 1*Butler Karen E. 1Wolf Carl E. 12Poklis Justin L. 3Poklis Alphonse 1231Department of Forensic Science, Virginia Commonwealth University, Richmond, VAUSA2Department of Pathology, Virginia Commonwealth University, Richmond, VAUSA3Department of Pharmacology and Toxicology, Virginia Commonwealth University, Richmond, VAUSAEdited by: Rukiyah Van Dross-Anderson, Brody School of Medicine at East Carolina University, USA Reviewed by: Alexander Oksche, Mundipharma Research, Germany; Allison Danell, East Carolina University, USA *Correspondence: Michelle R. Peace, mrpeace@vcu.eduThis article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology 29 8 2016 2016 7 27916 5 2016 12 8 2016 Copyright © 2016 Peace, Butler, Wolf, Poklis and Poklis.2016Peace, Butler, Wolf, Poklis and PoklisThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.Since 24 states and the District of Columbia have legalized marijuana in some form, suppliers of legal marijuana have developed Cannabis sativa products for use in electronic cigarettes (e-cigarettes). Personal battery powered vaporizers, or e-cigarettes, were developed to deliver a nicotine vapor such that smokers could simulate smoking tobacco without the inherent pathology of inhaled tobacco smoke. The liquid formulations used in these devices are comprised of an active ingredient such as nicotine mixed with vegetable glycerin (VG) and/or propylene glycol (PG) and flavorings. A significant active ingredient of C. sativa, cannabidiol (CBD), has been purported to have anti-convulsant, anti-nociceptive, and anti-psychotic properties. These properties have potential medical therapies such as intervention of addictive behaviors, treatments for epilepsy, management of pain for cancer patients, and treatments for schizophrenia. However, CBD extracted from C. sativa remains a DEA Schedule I drug since it has not been approved by the FDA for medical purposes. Two commercially available e-cigarette liquid formulations reported to contain 3.3 mg/mL of CBD as the active ingredient were evaluated. These products are not regulated by the FDA in manufacturing or in labeling of the products and were found to contain 6.5 and 7.6 mg/mL of CBD in VG and PG with a variety of flavoring agents. Presently, while labeled as to content, the quality control of manufacturers and the relative safety of these products is uncertain. cannabidiolelectronic cigarettesvapinge-liquidsDART-MSHPLC-MS/MSNational Institute of Justice10.13039/1000052892014-R2-CX-K010National Institutes of Health10.13039/100000002P30DA033934 ==== Body Introduction Electronic cigarettes (e-cigarettes) were developed as an alternative method for nicotine delivery. Their utility and popularity have transformed them into a general drug-delivery device. They are inexpensive, easy to use, and have some public perception as a healthy alternative. They work by either drawing negative pressure through the mouthpiece or depressing a button to activate a battery that heats a coil, containing a wick saturated with a formulation known as the e-liquid. The e-liquids are made of some ratio of propylene glycol (PG) and vegetable glycerin (VG) and/or a pharmaceutical and/or herbal remedy plus, potentially, a flavoring agent. When the e-cigarette is activated, the e-liquid is vaporized, followed by rapid condensation into an aerosol that the user inhales (Breland et al., 2016; Peace et al., 2016a). On May 5, 2016, the Food and Drug Administration (FDA) announced the extension of their authority to regulate all tobacco products, including e-cigarettes (U.S. Food and Drug Administration [FDA], 2016). A significant reason to do so was to address the quality assurance of e-cigarette products, from the devices to the nicotine-based e-liquids contained within. The e-liquid formulations have been found to vary significantly from the labeled content around the world (Etter et al., 2013; Goniewicz et al., 2013; Kavvalakis et al., 2015; Pagano et al., 2015; Peace et al., 2016a). Since the legalization of marijuana in some form in 24 states and the U.S. District of Columbia, e-liquids containing cannabinoids have emerged in the market place. As with nicotine e-liquid concentrations, the measured concentration of Δ9-tetrahydrocannabinol (THC) in a commercially available product was found to contain significantly different THC concentration than was labeled (Peace et al., 2016b). Cannabidiol (CBD) has been purported to have anti-convulsant, anti-nociceptive, and anti-psychotic properties (Brenneisen, 2007; Bhattacharyya et al., 2010). These properties have potential medical therapies such as intervention of addictive behaviors, treatments for epilepsy, management of pain for cancer patients, and treatments for schizophrenia (Johnson et al., 2010; Fischer et al., 2015; Friedman and Devinsky, 2015; Manseau and Goff, 2015). According to the Drug Enforcement Agency (DEA), CBD is a Schedule I substance as defined by the Controlled Substances Act (CSA). Recently, the DEA made it easier for scientists conducting FDA-approved studies to acquire CBD (United States Drug Enforcement Administration [DEA], 2016). Despite the ease of regulation for these research purposes, the CSA still disallows the addition of CBD to products for medicinal benefit since the FDA has not approved it for medical intervention. In 2015 and 2016, the FDA issued warning letters to companies marketing an unapproved drug in their products for therapeutic benefit (United States Food and Drug Administration [FDA], 2015a,b). Some companies selling products containing CBD continue to claim medicinal value for their products. However, some post the FDA disclaimer citing that their products are “not intended for the diagnosis, cure, mitigation, treatment, or prevention of a disease” according to the Federal Food, Drug, and Cosmetic Act and that they do not “distribute any products that are in violation of the U.S. Controlled Substances Act (Cloud 9 Hemp, 2015; Isodiol, 2016). Aside from volatile organic compounds and other potential degradation products generated by an e-cigarette during aerosolization which may pose health concerns (Flora et al., 2016), condensation aerosols are known to be useful and effective drug delivery systems. If research on CBD demonstrates acceptable therapeutic utility and thereby removed from Schedule I, the inhalation of CBD through an aerosol produced by an e-cigarette may be advantageous over traditional smoking methods and ingestion. THC enriched e-liquid vaporized in an e-cigarette has been demonstrated to be an effective route of administration for cannabinoids (Varlet et al., 2016). Of major import is that these cannabinoid infused e-liquids be subjected to manufacturing standards to ensure safety and quality of product. Presented is the evaluation of two commercially available e-liquids labeled to contain 3.3 mg/mL CBD in PG and VG with flavorings. The vendor claims that a hemp strain with the highest CBD potency was used in the manufacture of their products (Cloud 9 Hemp, 2015). The products were presumptively evaluated using Direct Analysis in Real Time ion source attached to a time of flight mass spectrometer (DART-MS) for cannabinoids, flavorants, and other possible constituents. Cannabinoids were quantitated by high-performance liquid chromatography–tandem mass spectrometry (HPLC-MS/MS). Alcohols were analyzed by headspace gas chromatography with flame ionization detector (HS-GC-FID). The aerosol produced from the e-liquids with an e-cigarette was analyzed by solid phase microextraction gas chromatograph (SPME-GC/MS). Materials Two e-liquids, Cloud 9 Hemp Easy Rider and Yellow Brick Road purported to be infused with CBD, were submitted to the lab for analysis. These products contained no information as to lot number or date of production (Figure 1). All tubing, glassware, and fritted gas dispersion tubes were purchased from Colonial Scientific (Richmond, VA, USA). HPLC-grade methanol used for all dilutions, stock and working solutions was purchased from Pharmco-Aaper (Brookfield, CT, USA). Polyethylene glycol (PEG) with an average molecular mass of 600 Da was used for DART-MS calibration and obtained from ULTRA Inc (North Kingstown, RI, USA). United States Pharmacopiea (USP)-grade PG and VG were obtained from Wizard Labs (Altamonte Springs, FL, USA). Nitrogen and helium gases were acquired from Praxair and Airgas (Richmond, VA, USA). Certified ACS Ammonium acetate, formic acid, HPLC-grade methanol and de-ionized (DI) water, optima grade acetone, ethanol, methanol, n-propanol, and isopropanol were purchased from Fisher Scientific (Hanover Park, IL, USA). Medical-grade nitrogen and helium were purchased from National Welders Supply Company (Richmond, VA, USA). CBD primary reference standard was purchased from Cerilliant (Round Rock, TX, USA). The e-cigarette was a KangerTech Aerotank clearomizer (v2) attached to an eGo-V v2 variable voltage battery, purchased from 101vape.com (Carlsbad, CA, USA). The single coil was wrapped in non-contact configuration with 34 gauge Nichrome wire to 1.8 Ω and a 2 mm diameter silica string was used as a wick. The flow meter was purchased from Cole Palmer (Vernon Hills, IL, USA). Seven micrometers polydimethylsiloxane (PDMS) SPME fibers were purchased from Supelco (Bellefonte, PA, USA). FIGURE 1 (A) Easy Rider and Yellow Brick Road CBD e-liquids with (B) the FDA statement regarding the uses of the e-liquid, and (C) the listed ingredients of the e-liquids. Methods Screening of Cannabidiol E-Liquids by DART-MS A previously published method employing a JEOL JMS T100LC Accu-TOF mass spectrometry controlled by Mass Center software version 1.3.4 m (JEOL Inc., Tokyo, Japan) was used to screen both CBD infused e-liquids for the presence of cannabinoids, flavoring agents, and other possible components (Poklis et al., 2015a,b). Briefly, a capillary tube was dipped five times into an aliquot of each e-liquid prior to introduction into the helium stream of a DART-MS. Each sample was analyzed in positive-ion mode with a helium stream temperature of 300°C. The flow rate was 2.3 L/min with a discharge electrode needle voltage at 150 V and grid electrode at 250 V. The ion guide peak voltage was 400 V, reflectron voltage was 900 V, orifice 2 was set to 5 V, and the ring lens was set to 3 V with orifice 1 operated in function switching mode at 20, 60, or 90 V with a single data file created for all three voltages. The range of masses measured was from 40 to 1,100 Da. Each e-liquid was analyzed five separate times to ensure reproducibility of the results. The data was analyzed by the creation of averaged, background subtracted, centroided mass spectra that was calibrated using the PEG 600. Identification of CBD was made when the exact mass was detected within 5 mDa of its calculated monoisotopic mass (M+H)+ and by its fragmentation pattern in function switching mode. PG and VG were confirmed by known standard analysis. All other compounds detected in the e-liquid were identified using a NIST 11.0 library. Volatile Analysis of Cannabidiol E-Liquids Headspace GC-FID A validated analytical method routinely used for the analysis of clinical and forensic samples for ethanol, acetone, isopropanol, and methanol was employed using a Tekmar HT3 headspace sampler attached to a Shimadzu 2014 GC-FID. The chromatographic separation was performed on a RTX-BAC1 (30 m × 0.32 mm id × 1.80 μM column (Restek Corp, Bellefonte, PA, USA). The headspace oven and transfer line temperatures were set to 160°C with a standby flow rate of 200 mL/min. The platen sample temperature was 80°C with the mixer on. The sample equilibrium time was 3.5 min with a sample injection time of 0.5 min. The GC has an oven temperature of 50°C and the injection temperature of 200°C run in split mode with a 1:20 ratio. The column flow rate was 6.85 mL/min with purge flow of 0.5 mL/min, and the detector temperature was 225°C. Calibrators, controls and the CBD e-liquids were diluted with the internal standard, 234 mg/L n-propanol in water, at 1:10 ratio prior to analysis. The limit of detection (LOD) for all volatiles was 100 mg/L with a determined linear range of 100–8,000 mg/L for the assay. Quantification of Cannabidiol E-Liquids by HPLC-MS/MS Quantification of CBD was performed using a modification of a previously validated method (Poklis et al., 2010) on an Applied Biosystems 3200 Q Trap with a turbo V source for TurboionSpray (Applied Biosystems, Foster City, CA, USA) tandem mass spectrometer (MS/MS) attached to a Shimadzu SCL HPLC system (Shimadzu Corp., Kyoto, Japan) with a Zorbax eclipse XDBC18 column 4.6 mm × 75 mm, 3.5 μ (Agilent Technologies, Santa Clara, CA, USA) controlled by Analyst 1.4.2 software. The mass spectrometer detector was operated in scan mode with a mass range of 50–700 Da. The injection volume was 10 μL. The mobile phase was 10:90 (v:v) methanol:DI water with a pump flow rate of 0.5 mL/min. The ion spray voltage was 5,000 V with ion source gases 1 and 2 having flow rates of 60 mL/min. The source temperature was set at 650°C and the curtain gas flow rate was 30 mL/min. The following transition ions (m/z) were monitored with their corresponding collection energies of 29 eV: CBD: 315 > 193, 315 > 259 and CBD-d3: 318 > 196, 318 > 261. The total run time was 8.0 min. Each e-liquid was diluted 1:10,000 and 1:100,000 with methanol. A CBD calibration curve was prepared in duplicate with a range of 10–100 ng/mL. Quality controls specimens of 10, 30, 300, and 750 ng/mL of CBD were prepared in triplicate. A linear regression of the ratio of the peak area counts of cannabinoids and the deuterated internal standards versus concentration was used to construct the calibration curves with a coefficient of determination (r2) of 0.9995. The transition ions’ relative abundances for the identified cannabinoids were ±20% of target, relative to the calibrators. All quality control specimens were within 15% of their expected target values. Analysis of the Aerosol Generated by Cannabidiol E-Liquids The CBD e-liquids were mixed by hand for 20 s before adding them to a clean KangerTech e-cigarette. The aerosol produced by the e-cigarette was trapped with common glassware. Briefly, two Erlenmeyer flasks were connected in tandem to a vacuum with a flow rate of 2.3 L/min. DI water was added to each trap and a gas dispersion tube bubbled the aerosol into the water. Glass wool was placed in between the two traps to contain the aerosol in the first trap. A 7 μm PDMS SPME fiber was inserted through a stopper in the first trap. The fiber was introduced into the trap while the e-cigarette was activated and the aerosol filled the trap. The e-cigarette was activated for 4 s, aerosolizing 7–10 μL of the e-liquid, and the SPME fiber was held in the trap for 5 min, after which the fiber was retracted. The SPME fiber was inserted into the injection port of an Agilent GC-MS 6890N/5973 Mass Selective Detector (Agilent, Santa Clara, CA, USA) with an HP-5MS column 30 m × 0.25 mm id × 0.25 μm (Agilent, Santa Clara, CA, USA). The injection port was set to 315°C and the run was made in splitless mode with a 15 min thermal desorption time. The initial temperature was set to 120°C, with a ramp of 300°C at 10°C/min, and then a hold time of 12 min, for a total run time of 30 min. Five separate samplings were collected for each e-liquid. The fibers were thermally cleaned between runs following manufacturer specifications to ensure no carryover occurred between samples. A 100 ng/mL CBD reference standard was also analyzed on the GC-MS and the combination of retention time and mass spectrum were used to identify CBD in the aerosolized e-liquids. Flavoring agents were identified using a NIST 11.0 library. Results Cloud 9 Hemp Yellow Brick Road and Easy Rider e-liquids were determined to contain 7.6 and 6.5 mg/L CBD, respectively. Both e-liquids were found to contain PG and VG (65:35 v:v). Yellow Brick Road and Easy Rider e-liquids were also determined to contain 3,600 and 6,600 mg/L ethanol, respectively. Yellow Brick Road was determined to contain the following flavoring agents: peach lactone, y-decalactone, and vanillin, and Easy Rider contained ethyl maltol, ethyl vanillin, vanillin, y-nonalactone, benzaldehyde PG acetal, dimethyl anthranilate, propenylguaethol, 6-methylcoumarin, and heliotropin PG acetal (Table 1). The analysis of the aerosol generated using Yellow Brick Road and Easy Rider e-liquids in the KangerTech e-cigarette resulted in the detection of the CBD, PG, VG, and flavoring agents found in the e-liquids (Supplementary Figure S1). Table 1 Compounds detected in the Easy Rider and Yellow Brick Road e-liquids by DART-MS and SPME-GC/MS analysis. Compound Formula Monoisotopic MW (g/mol) DART-MS [M+H]+ SPME-GC/MS RT (minute) Easy Rider Ethyl maltol C7H8O3 140.047 141.061 3:07 Benzaldehyde propylene glycol acetal C10H12O2 164.084 ND 3:45 Wine ether C11H22O2 186.162 187.174 3:44 γ-Nonalactone C9H16O2 156.115 157.123 4:28 Vanillin C8H8O3 152.047 153.060 4:50 Dimethyl anthranilate C9H11NO2 165.079 166.091 4.95 Ethyl vanillin C9H10O3 166.063 ND 5:24 Propenylguaethol C11H14O2 178.099 179.112 6:05 6-Methylcoumarin C10H8O2 160.052 161.069* 6:34 Heliotropin propylene glycol acetal C11H12O4 208.074 209.074 7:14 Benzyl benzeneacetate C15H14O2 226.099 ND 9:06 Cannabidiol C21H30O2 314.225 315.232 14:39 Yellow Brick Road Isomenthol acetate C12H22O2 198.162 ND 3.82 Vanillin C8H8O3 152.047 153.062 4:50 Peach lactone C11H20O2 184.146 185.151 5:29 γ-Decalactone C10H18O2 170.131 171.141 6:35 Cannabidiol C21H30O2 314.225 315.232 14:39 MW, molecular weight; ND, not detected by DART-MS.*Confirmation by GC-MS. Low abundance in DART-MS prohibited identification.Discussion Electronic cigarettes have become a popular means for using pharmaceuticals other than nicotine. This study characterized commercially available e-liquids that were advertised as containing CBD. DART-MS was used to assess the presence of PG, VG, CBD, and a flavor profile for each of the e-liquids. It was also used to evaluate the e-liquids for any other potential constituents. Volatile analysis was used to detect and quantitate the presence of ethanol within the samples. The CBD detected using the DART-MS method was then quantitated using HPLC-MS/MS, and SPME-GC/MS was used to identify compounds aerosolized by the e-cigarette. Only a single cannabinoid, CBD, was detected by the DART-MS, HPLC-MS/MS, and SPME-GC/MS methods. SPME has historically been used in order to characterize the smoke of traditional cigarettes (Watson et al., 2004). One such study, performed by Bao et al. (2010), used HS-SPME to determine the concentration of free-base nicotine in cigarette smoke. Another study performed by Ye (2008) used SPME-GC/MS in order to analyze any volatile organic compounds found from cigarette smoke from 10 different types of traditional cigarettes. Wu et al. (2002) were also capable of detecting the abundant alkaloids from cigarette smoke from 14 different countries. The present study sought to use a similar concept in order to analyze condensation aerosol of e-liquids containing CBD in order to characterize the aerosol produced by the e-liquids. The e-cigarette aerosolized the CBD as it was successfully extracted from the aerosol using SPME, then thermally desorbed by GC/MS. The final chromatographic profile of the aerosol included PG, VG, CBD, and flavoring agents. No carbonyl compounds were detected in the aerosol produced by the e-liquid using the SPME-GC/MS method presented. These compound have been detected by other investigators in aerosols by other methodologies (Bekki et al., 2014; Kosmider et al., 2014; Flora et al., 2016; Jensen et al., 2015). While there is no current literature regarding trends among the quality control of CBD containing e-liquids, there have been several involving quantitative studies of nicotine-based e-liquids. Two studies by Goniewicz et al. (2013, 2014) involved the analysis of the concentration of nicotine in cartridges and e-liquids from the UK, Poland, and the United States while Kavvalakis et al. (2015) evaluated products available on the Greek market. Etter et al. (2013), Lisko et al. (2015), and Trehy et al. (2011) analyzed e-liquids and cartridges for flavoring agents, nicotine, and potential nicotine impurities. Pagano et al. (2015) evaluated products purchased in the United States, as did this research group (Peace et al., 2016a). These studies all noted that there was a difference in quantity from the advertised concentration of the active pharmaceutical from the actual concentration, some to be significant. The previous study by Peace et al. (2016b) found that 30% of the e-liquids purchased in the United States had greater than a 20% difference from the labeled concentration of nicotine. This research group also found that a commercially available product, labeled as 69.1% THC and 1.0% CBD, actually contained 42.6% THC (w/v) and 0.5% CBD (w/v). They also found that the THC e-liquid contained four additional cannabinoids and 13 terpenes, indicating that the product was created by extracting Cannabis sativa (Peace et al., 2016b). The CBD e-liquids evaluated in this study both contained over twice the advertised concentrations of CBD and also ethanol. Since the e-liquids in this study did not contain any terpenes nor were there any other cannabinoids detected, it could be deduced that the CBD was not produced by an extraction of marijuana. The detection of the unlabeled ethanol in these products may have been used as a natural flavorant or as a solvent, however, the reason for the ethanol as an ingredient cannot be fully ascertained. The flavorants detected by DART-MS and/or SPME-GC/MS were only included on the product labels as “Natural and Artificial Flavorings.” Lack of regulation provides opportunities for products to be developed without the oversight for the quantity, quality, and safety of products. While they may not necessarily be harmful to the user, it could be important for a consumer, who is trying to monitor their CBD use for medical intervention, to know. Conclusion The e-liquids analyzed in this study both screened positive for CBD by DART-MS and were confirmed to have CBD present by HPLC-MS/MS. SPME was effective in extracting CBD and the other components produced by e-cigarette to confirm aerosolization. The contents of the e-liquids were labeled to contain 3.3 mg/mL of CBD. They were found to contain 6.5 and 7.6 mg/mL of CBD. The analysis of these two different products illustrates the potential quality control issues that can occur in an unregulated industry. Even with uncertain safety of e-cigarettes and e-liquids, the e-cigarette used was shown to generate an aerosol containing CBD using the commercially available e-liquids. Vaping CBD would eliminate harmful combustion products of smoking, while still employing the advantages of inhalation over ingestion. As research continues to understand the effects of CBD and its potential therapeutic properties the e-cigarettes many prove a useful CBD drug delivery system. Author Contributions MP oversees all operations, coordinates experiments and collaborators, and evaluates data to decide quality, experiments, and next directions. KB conducted all experiments on the DART-MS and SPME-GC-MS, significant manuscript preparation. CW conducted the ethanol concentration experiments by GC-MS. JP conducted the CBD experiments by HPLC-MS/MS, significant manuscript preparation. AP coordinated all experiments for ethanol and CBD quantitation, evaluate data, manuscript preparation. Conflict of Interest Statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer AD and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review. Funding. This project was supported by Award No. 2014-R2-CX-K010, awarded by the National Institute of Justice, Office of Justice Programs, U.S. Department of Justice and the National Institutes of Health Award No P30DA033934. The opinions, findings, and conclusions or recommendations expressed in this publication/program/exhibition are those of the author(s) and do not necessarily reflect those of the Department of Justice. Supplementary Material The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fphar.2016.00279 TABLE S1 The Total Ion Chromatographs of SPME-GC/MS analysis from the condensation aerosol of the Easy Rider and Yellow Brick Road e-liquids. Click here for additional data file. Click here for additional data file. ==== Refs References Bao M. Joza P. Ricket W. S. Lauterback J. H. (2010 ). An improved headspace solid-phase microextraction method for the analysis of free-base nicotine in particulate phase of mainstream cigarette smoke. Anal. Chim. Acta 663 49 –54 . 10.1016/j.aca.2010.01.036 20172096 Bekki K. Uchiyama S. Ohta K. Inaba Y. Nakagome H. Kunugita N. (2014 ). Carbonyl compounds generated from electronic cigarettes. Int. J. Environ. Res. Public Health 11 11192 –11200 . 10.3390/ijerph111111192 25353061 Bhattacharyya S. Morrison P. D. Fusar-Poli P. Martin-Santos R. Borgwardt S. Winton-Brown T. (2010 ). Opposite effects of delta-9-tetrahydrocannabinol and cannabidiol on human brain function and psychopathology. Neuropsychopharmacology 35 764 –774 . 10.1038/npp.2009.184 19924114 Breland A. Soule E. Lopez A. Ramôa C. El-Hellani A. Eissenberg T. (2016 ). Electronic cigarettes: what are they, and what do they do? Ann. N. Y. Acad Sci. 10.1111/nyas.12977 [Epub ahead of print] . Brenneisen R. (2007 ). “Chemistry and analysis of phytocannabinoids and other cannabis constituents ,” in Marijuana and the Cannabinoids ed. Elsohly M. (Totowa, NJ : Human Press ) 17 –51 . Cloud 9 Hemp (2015 ). Cloud 9 Hemp,” Cloud9hemp. com, Last Modified. Available at: https://cloud9hemp.com/ (accessed May 10 2016 ). Etter J. F. Zather E. Svensson S. (2013 ). Analysis of refill liquids for electronic cigarettes. Addiction 108 1671 –1679 . 10.1111/add.12235 23701634 Fischer B. Kuganesan S. Gallassi A. Malcher-Lopes R. van den Brink W. Wood E. (2015 ). Addressing the stimulant treatment gap: a call to investigate the therapeutic benefits potential of cannabinoids for crack-cocaine use. Int. J. Drug Policy 26 1177 –1182 . 10.1016/j.drugpo.2015.09.005 26500166 Flora J. W. Meruva N. Huang C. B. Wilkinson C. T. Ballentine R. Smith D. C. (2016 ). Characterization of potential impurities and degradation products in electronic cigarette formulations and aerosols. Regul. Toxicol. Pharmacol. 74 1 –11 . 10.1016/j.yrtph.2015.11.009 26617410 Friedman D. Devinsky O. (2015 ). Cannabinoids in the treatment of epilepsy. N. Engl. J. Med. 373 1048 –1058 . 10.1056/NEJMra1407304 26352816 Goniewicz M. L. Hajek P. McRobbie H. (2014 ). Nicotine content of electronic cigarettes, its release in vapour and its consistency across batches: regulatory implications. Addiction 109 500 –507 . 10.1111/add.12410 24345184 Goniewicz M. L. Kuma T. Gawron M. Knysak J. Kosmider L. (2013 ). Nicotine levels in electronic cigarettes. Nicotine Tob. Res. 15 158 –166 . 10.1093/ntr/nts103 22529223 Isodiol . (2016 ). “Disclaimer,” Isodiol.com, Last Modified. Available at: http://www.isodiol.com/#!disclaimer/oztke Jensen R. P. Luo W. Pankow J. F. Strongin R. M. Peyton D. H. (2015 ). Hidden formaldehyde in e-cigarette aerosols. N. Engl. J. Med. 372 392 –394 . 10.1056/NEJMc1413069 25607446 Johnson J. R. Burnell-Nugent M. Lossignol D. Ganae-Motan E. D. Potts R. Fallon M. T. (2010 ). Multicenter, double-blind, randomized, placebo-controlled, parallel-group study of the efficacy, safety, and tolerability of THC:CBD extract and THC extract in patients with intractable cancer-related pain. J. Pain Symptom Manage. 39 167 –179 . 10.1016/j.jpainsymman.2009.06.008 19896326 Kavvalakis M. P. Stivaktakis P. D. Tzatzarakis M. N. Kouretas D. Liesivuori J. Alegakis A. K. (2015 ). Multicomponent analysis of replacement liquids of electronic cigarettes using chromatographic techniques. J. Anal. Toxicol. 39 262 –269 . 10.1093/jat/bkv002 25681325 Kosmider L. Sobczak A. Fik M. Knysak J. Zaciera M. Kurek J. (2014 ). Carbonyl compounds in electronic cigarette vapors: effects of nicotine solvent and battery output voltage. Nicotine Tob. Res. 16 1319 –1326 . 10.1093/ntr/ntu078 24832759 Lisko J. G. Tran H. Stanfill S. B. Blount B. C. Watson C. H. (2015 ). Chemical composition and evaluation of nicotine, tobacco alkaloids, ph, and selected flavors in e-cigarette cartridges and refill solutions. Nicotine Tob. Res. 17 1270 –1278 . 10.1093/ntr/ntu279 25636907 Manseau M. W. Goff D. C. (2015 ). Cannabinoids and Schizophrenia: risks and therapeutic potential. Neurotherapeutics 12 816 –824 . 10.1007/s13311-015-0382-6 26311150 Pagano T. DiFrancesco A. G. Smith S. B. George J. Wink G. Rahman I. (2015 ). Determination of nicotine content and delivery in disposable electronic cigarettes available in the united states by gas chromatography-mass spectrometry. Nicotine Tob. Res. 18 700 –707 . 10.1093/ntr/ntv120 26045251 Peace M. R. Baird T. R. Smith N. Wolf C. E. Poklis J. L. Poklis A. (2016a ). Concentration of nicotine and glycols in 27 electronic cigarette formulations. J. Anal. Toxicol. 40 403 –407 . 10.1093/jat/bkw037 27165804 Peace M. R. Stone J. W. Poklis J. L. Turner J. B. M. Poklis A. (2016b ). Analysis of a commerical marijuana e-cigarette formulation. J. Anal. Toxicol. 40 374 –378 . 10.1093/jat/bkw021 27059691 Poklis J. L. Raso S. A. Alford K. N. Poklis A. Peace M. R. (2015a ). Analysis of 25I-NBOMe, 25B-NBOMe, 25C-NBOMe and Other Dimethoxyphenyl-N-[(2-Methoxyphenyl) Methyl]Ethanamine derivatives on blotter paper. J. Anal. Toxicol. 39 617 –623 . 10.1093/jat/bkv073 26378135 Poklis J. L. Thompson C. C. Long K. A. Lichtman A. H. Poklis A. (2010 ). Disposition of cannabichromene, cannabidiol, and Δ?-tetrahydrocannabinol and its metabolites in mouse brain following marijuana inhalation determined by high-performance liquid chromatography-tandem mass spectrometry. J. Anal. Toxicol. 8 516 –520 . 10.1093/jat/34.8.516 21258613 Poklis J. L. Wolf C. E. ElJordi O. I. Liu K. Zhang S. Poklis A. (2015b ). Analysis of the first- and second-generation raving dragon novelty bath salts containing methylone and pentedrone. J. Forensic Sci. 60 S234 –S240 . 10.1111/1556-4029.12629 25470207 Trehy M. L. Ye W. Hadwiger M. E. Moore T. W. Allgire J. F. Woodruff J. T. (2011 ). Analysis of electronic cigarette cartridges, refill solutions, and smoke for nicotine and nicotine related impurities. J. Liq. Chromatogr. Relat. Technol. 34 1442 –1458 . 10.1093/ntr/ntu197 U.S. Food and Drug Administration [FDA] (2016 ). “FDA Takes Significant Steps to Protect Americans From Dangers of Tobacco Through New Regulation,” fda.gov, Last Modified May 10 2016. Available at: http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm499234.htm [accessed May 12 2016 ]. United States Drug Enforcement Administration [DEA] (2016 ). “DEA Eases Requirements for FDA-Approved Clinical Trials on Cannabidiol,” dea.gov, last modified December 23 2015. Available at: http://www.dea.gov/divisions/hq/2015/hq122315.shtml [accessed May 12 2016] . United States Food and Drug Administration (2015a ). Gas Chromatography - Mass Spectrometry (GC-MS) Screening Procedure for the Presence of Diethylene Glycol and Ethylene Glycol in Toothpaste. Available at: http://www.fda.gov/Food/FoodScienceResearch/LaboratoryMethods/ucm113209.htm [accessed October 16 2014] . United States Food and Drug Administration (2015b ). Warning Letters and Test Results,” fda.gov, last Modified February 9 2015. Available at: http://www.fda.gov/NewsEvents/PublicHealthFocus/ucm435591.htm [accessed May 12 2016] . Varlet V. Concha-Lozano N. Berthet A. Plateel G. Fav rat B. De Cesare M. (2016 ). Drug vaping applied to cannabis: is “Cannavaping” a therapeutic alternative to marijuana? Nature 6 25599 10.1038/srep25599 Watson C. H. Trommel J. S. Ashley D. L. (2004 ). Solid-phase microextraction-based approach to determine free-base nicotine in trapped mainstream cigarette smoke total particulate matter. J. Agric. Food Chem. 52 7240 –7245 . 10.1021/jf049455o 15563201 Wu W. Ashley D. L. Watson C. H. (2002 ). Determination of nicotine and other minor alkaloids in international cigarettes by solid-phase microextraction and gas chromatography/mass spectrometry. Anal. Chem. 74 4878 –4884 . 10.1021/ac020291p 12380807 Ye Q. (2008 ). Development of solid-phase microextraction followed by gas chromatography-mass spectrometry for rapid analysis of volatile organic chemicals in mainstream cigarette smoke. J. Chromatography. A 1213 239 –244 . 10.1016/j.chroma.2008.10.063

PMC005xxxxxx/PMC5002420.txt

==== Front Front MicrobiolFront MicrobiolFront. Microbiol.Frontiers in Microbiology1664-302XFrontiers Media S.A. 10.3389/fmicb.2016.01330MicrobiologyOriginal ResearchComparative Analysis of Carbon Monoxide Tolerance among Thermoanaerobacter Species Alves Joana I. 12*Alves M. Madalena 1Plugge Caroline M. 2Stams Alfons J. M. 12Sousa Diana Z. 121Centre of Biological Engineering, University of MinhoBraga, Portugal2Laboratory of Microbiology, Wageningen UniversityWageningen, NetherlandsEdited by: Eric Altermann, AgResearch, New Zealand Reviewed by: Jose M. Bruno-Barcena, North Carolina State University, USA; Christopher L. Hemme, University of Rhode Island, USA *Correspondence: Joana I. Alves joana.alves@deb.uminho.ptThis article was submitted to Microbiotechnology, Ecotoxicology and Bioremediation, a section of the journal Frontiers in Microbiology 29 8 2016 2016 7 133031 5 2016 11 8 2016 Copyright © 2016 Alves, Alves, Plugge, Stams and Sousa.2016Alves, Alves, Plugge, Stams and SousaThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.An anaerobic thermophilic strain (strain PCO) was isolated from a syngas-converting enrichment culture. Syngas components cannot be used by strain PCO, but the new strain is very tolerant to carbon monoxide (pCO = 1.7 × 105 Pa, 100% CO). 16S rRNA gene analysis and DNA-DNA hybridization revealed that strain PCO is a strain of Thermoanaerobacter thermohydrosulfuricus. The physiology of strain PCO and other Thermoanaerobacter species was compared, focusing on their tolerance to carbon monoxide. T. thermohydrosulfuricus, T. brockii subsp. finnii, T. pseudethanolicus, and T. wiegelii were exposed to increased CO concentrations in the headspace, while growth, glucose consumption and product formation were monitored. Remarkably, glucose conversion rates by Thermoanaerobacter species were not affected by CO. All the tested strains fermented glucose to mainly lactate, ethanol, acetate, and hydrogen, but final product concentrations differed. In the presence of CO, ethanol production was generally less affected, but H2 production decreased with increasing CO partial pressure. This study highlights the CO resistance of Thermoanaerobacter species. syngascarbon monoxideThermoanaerobactersugar fermentationethanolhydrogen ==== Body Introduction Thermophiles thrive at austere and unusual conditions and their evolutionary significance and biotechnological potential have triggered microbiological research over the last decades (Turner et al., 2007; Wagner and Wiegel, 2008; Yoneda et al., 2015). There is a continuous biotechnological interest in highly thermostable enzymes, which make this kind of organisms very attractive. Thermophilic bacteria of the class Clostridia, such as members of the genera Clostridium, Thermoanaerobacter, Thermoanaerobacterium, and Caldicellulosiruptor, are currently used as biocatalysts for the production of biofuels or other chemicals of interest (Hemme et al., 2010; Carere et al., 2012). Specifically members of the Thermoanaerobacter genus are utilized to produce ethanol and hydrogen (H2) from a variety of saccharides (Jessen and Orlygsson, 2012). Within thermophiles, an organism from Thermoanaerobacter genus—T. ethanolicus—is one of the most well-studied ethanol-producing bacteria (Wiegel and Ljungdahl, 1981; Lacis and Lawford, 1991). A less common substrate, carbon monoxide (CO), is used by T. thermohydrosulfuricus subsp. carboxydovorans and T. kivui. T. thermohydrosulfuricus subsp. carboxydovorans can grow with CO as sole electron donor (25% in the headspace), producing H2 and CO2 (Balk et al., 2009). T. thermohydrosulfuricus shares 99% similarity of the 16S rRNA gene sequence and over 70% DNA-DNA hybridization with T. thermohydrosulfuricus subsp. carboxydovorans, but only the latter one can use CO. Growth of the homoacetogenic T. kivui with CO diluted with CO2/N2 or CO2/H2 was described by Kevbrina et al. (1996). Recently, Weghoff and Müller (2016) reported the ability of T. kivui to grow on only CO (100% in the headspace), producing acetate and hydrogen. Carboxydotrophic metabolism in Thermoanaerobacter species is normally not assessed, and it is not known if they can endure CO or even adapt to grow on CO, as recently reported for T. kivui (Weghoff and Müller, 2016). In this work we isolated Thermoanaerobacter thermohydrosulfuricus strain PCO from a thermophilic syngas-converting enrichment, but this strain appears unable to oxidize CO. The main objectives of this work were (1) to characterize and determine the CO tolerance of Thermoanaerobacter thermohydrosulfuricus strain PCO, and (2) to compare the effect of CO on growth, glucose consumption and product formation of strain PCO and of four close relative species from the Thermoanaerobacter genus. Materials and methods Enrichments and isolation Suspended sludge from a thermophilic anaerobic municipal solid waste digester (Barcelona, Spain) was used as inoculum for starting up syngas-converting enrichments. Microbial cultures were enriched with synthetic syngas (mixture of 60% CO, 10% CO2, and 30% H2, total pressure 1.7 × 105 Pa) as sole carbon and energy source (Alves et al., 2013). Isolation of strain PCO was done using soft agar (1.5%, w/v) incubations and liquid medium serial dilutions, with 20 mM pyruvate as sole substrate. Sodium pyruvate was added to the medium from a 1M filter-sterilized stock solution. A phosphate-buffered mineral medium was used, containing (per liter): Na2HPO4, 1.63 g; NaH2PO4, 1.02 g; resazurin, 0.5 g; NH4Cl, 0.3 g; CaCl2·2H2O, 0.11 g; MgCl2·6H2O, 0.10 g; NaCl, 0.3 g; 1 mL of acid and alkaline trace element stock each, and 0.2 ml of vitamin stock. Trace elements and vitamins were prepared as described previously (Stams et al., 1993). Before inoculation, medium was reduced with sodium sulfide (0.8 mM final concentration). Bottles were incubated in the dark at 55°C while shaken at 100 rpm (liquid cultures) or standing (soft-agar cultures). Colonies were picked from soft-agar incubations, inoculated in fresh liquid medium containing pyruvate (20 mM). Cultures were further purified by subsequent serial dilutions alternating with soft-agar colony picking. Purity of the culture was checked by microscopic examination after growth with different substrates (Olympus CX41, Tokyo, Japan). Direct sequencing of the 16S rRNA gene and denaturing gradient gel electrophoresis (DGGE) were also applied to check the genetic purity of the culture. DNA isolation, PCR and DGGE Genomic DNA from strain PCO was extracted using the FastDNA SPIN kit for soil (MP Biomedicals, Solon, OH), according to the manufacturer's instructions. The 16S rRNA gene was directly amplified from genomic DNA by PCR, using the primer set 027F/1492R (Nübel et al., 1996) and the following PCR program: pre-denaturation, 2 min at 95°C; 30 cycles of denaturation, 30 s at 95°C, annealing, 40 s at 52°C, and elongation, 90 s at 72°C; and post-elongation, 5 min, at 72°C. For DGGE analysis, the 16S rRNA gene was partially amplified from genomic DNA with primer set U968GC-f/L1401-r (Lane, 1991; Muyzer et al., 1993). The thermocycling program used for PCR-DGGE amplification was: pre-denaturation, 5 min at 95°C; 35 cycles of denaturation, 30 s at 95°C, annealing, 40 s at 56°C, and elongation, 90 s at 72°C; and post-elongation, 5 min at 72°C. DGGE was performed using a DCode system (Bio-Rad, Hercules, CA). Gels contained 8% (wt/vol) polyacrylamide (37.5:1 acrylamide/bis-acrylamide) and a linear denaturing gradient of 30–60%, with 100% of denaturant corresponding to 7 M urea and 40% (vol/vol) formamide. Electrophoresis was performed for 16 h at 85 V and 60°C in a 0.5x Tris-Acetate–EDTA buffer. DGGE gels were stained with silver nitrate (Sanguinetti et al., 1994). Sequencing and phylogenetic analysis PCR products obtained from 16S rRNA gene amplification were purified using the PCR Clean Up kit NucleoSpin Extract II (Macherey-Nagel, Düren, Germany) and sequenced directly at Eurofins MWG Operon (Ebersberg, Germany). Partial sequences were assembled using the alignment editor BioEdit v7.0.9 software package (Hall, 1999). Similarity searches for the 16S rRNA gene sequence derived from strain PCO were performed using the NCBI BLAST search program within the GenBank database (Altschul et al., 1990). The 16S rRNA gene sequence of Thermoanaerobacter strain PCO is available in the DDBL/EMBL/GenBank databases under the accession number HF586422. Characterization of strain PCO and cultivation of Thermoanaerobacter strains Unless otherwise stated, all the physiological tests of strain PCO and its close relatives (T. thermohydrosulfuricus, T. brockii subsp. finnii, T. pseudethanolicus, and T. wiegelii) were performed using a bicarbonate-buffered mineral salt medium (Stams et al., 1993). Type strains of Thermoanaerobacter thermohydrosulfuricus (DSM 527T), T. brockii subsp. finnii (DSM 3389T), T. pseudethanolicus (DSM 2355T), and T. wiegelii (DSM 10319T) were obtained from the Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ; German Collection of Microorganisms and Cell Cultures, Braunschweig, Germany). Growth of strain PCO was tested with the following substrates (at a concentration of 20 mM unless indicated otherwise): acetate, arabinose, cellobiose, cellulose (5 g L−1), ethanol, formate, fructose, galactose, glucose, glycerol, glycine, lactate, lactose, maltose, mannitol, mannose, methanol, pectin (5 g L−1), propionate, pyruvate, raffinose, ribose, sorbitol, starch (5 g L−1), sucrose, trehalose, xylan (5 g L−1), xylose, yeast extract (5 g L−1), CO (from 20 to 100% CO, 1.7 × 105 Pa), and H2/CO2 (80/20%, 1.7 × 105 Pa). Utilization of different electron acceptors (elemental sulfur, AQDS, sulfate, sulfite, thiosulfate, nitrate, and nitrite) by strain PCO was done using pyruvate (20 mM) as electron donor, while glucose (20 mM) was for tests with T. thermohydrosulfuricus, T. brockii subsp. finnii, T. pseudethanolicus, and T. wiegelii. Pyruvate (20 mM) was used to test the optimum growth temperature (range 20–85°C) and pH (range 5.7–8.0) of strain PCO. All the assays were done in duplicate. Cell growth was determined by measuring optical density at 600 nm with a spectrophotometer (U-1500 Hitachi, Tokyo, Japan). Cell morphology of strain PCO was examined by phase contrast microscopy (Leica DM 2000, Wetzlar, Germany). Cells from active cultures of strain PCO were stained using standard Gram staining techniques. The DNA–DNA hybridization analysis and the G+C content of the DNA were determined by the identification service of the Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ; German Collection of Microorganisms and Cell Cultures, Braunschweig, Germany). Carbon monoxide tolerance tests Strain PCO, Thermoanaerobacter thermohydrosulfuricus, T. brockii subsp. finnii, T. pseudethanolicus, and T. wiegelii were tested for CO tolerance. All cultures were incubated with 0, 50, and 100% CO (pCO/P, where “pCO” is the CO partial pressure and “P” the total gas pressure). Additionally, strain PCO was incubated with 25 and 75% CO. Initial total pressure was 1.7 × 105 Pa in all the assays; N2 was used to pressurize the headspace for CO percentages lower than 100%. The tests were performed using an anaerobic phosphate-buffered mineral salt medium and glucose (20 mM) was used as carbon and energy source. Bottles were incubated in the dark, at 55°C and shaken at 100 rpm. All the assays were done in duplicate. Growth of the strains was determined by measuring optical density increase at 600 nm with a spectrophotometer (U-1500 Hitachi, Tokyo, Japan). The statistical significance of the differences detected in glucose conversion rates and end products production was evaluated using single factor analysis of variances (ANOVA). Analytical methods Soluble substrates and intermediates (sugars, organic acids, and alcohols) were measured using a HPLC Thermo Electron equipment with a Shodex SH1821 column and equipped with a RI detector. The mobile phase used was sulfuric acid (0.01 N) at a flow rate of 0.6 mL min−1. Column temperature was set at 60°C. Inorganic anions were analyzed by chromatography using a HPLC Dionex system, equipped with an Ionpac AS22 column, and ED40 electrochemical detector. Column temperature and pressure varied between 35–40°C and 130 × 105–160 × 105 Pa. Gaseous compounds (CO, CO2, H2) were analyzed by gas chromatography on a GC-2014 Shimadzu with a thermal conductivity detector. CO2 was analyzed with a CP Poraplot Q column (25 m length, 0.53 mm internal diameter; film thickness, 20 μm). Helium was used as carrier gas at a flow rate of 15 mL min−1, and the temperatures in the injector, column, and detector were 60, 33, and 130°C. CO and H2 were analyzed with a Molsieve 13X column (2 m length, 3 mm internal diameter). Argon was used as carrier gas at a flow rate of 50 mL min−1, and temperatures in the injector, column, and detector were 80, 100, and 130°C. Results and discussion Physiological characterization of strain PCO and comparison with closely related Thermoanaerobacter species Strain PCO was isolated from the thermophilic syngas-converting enrichment described by Alves et al. (2013). Isolation was performed using pyruvate as sole carbon and energy source. Strain PCO has a G+C content of the DNA of 34.5 mol % and shares 98% identity with the 16S rRNA gene of Thermoanaerobacter thermohydrosulfuricus (the 16S rRNA gene sequence of strain PCO is available in the DDBL/EMBL/GenBank databases under the accession number HF586422). DNA-DNA hybridization between the two strains was 100%, confirming that strain PCO is a strain of Thermoanaerobacter thermohydrosulfuricus. Strain PCO was not a predominant microorganism in the syngas enriched culture, as shown by the molecular characterization (Alves et al., 2013). However, its presence indicates that it can grow on metabolic byproducts and/or dead cells at high CO concentrations. Strain PCO formed terminal round endospores, which is a characteristic of Thermoanaerobacter species (Wiegel and Ljungdahl, 1981; Lee et al., 1993, 2007; Kim et al., 2001; Balk et al., 2009; Shaw et al., 2010). Cells of strain PCO are straight rods and normally occur singly (Figure 1). Strain PCO had an optimum growth temperature of 70°C; no growth was detected below 37°C or above 75°C. The optimum pH for growth was between 6.5 and 7.5. Strain PCO is a very versatile organism that can utilize a range of different substrates, such as: arabinose, cellobiose, cellulose, fructose, galactose, glucose, lactose, maltose, mannitol, mannose, pectin, pyruvate, raffinose, ribose, sorbitol, starch, sucrose, trehalose, xylan, xylose, and yeast extract. No growth occurred with acetate, ethanol, formate, glycerol, glycine, lactate, methanol, propionate, CO (from 20 to 100% CO, total pressure 1.7 × 105 Pa), and H2/CO2 (80/20%, total pressure 1.7 × 105 Pa). The main products detected and quantified from glucose fermentation by strain PCO were lactate, ethanol, acetate, and H2 (Figure 2) which are typically formed from glucose by most of the Thermoanaerobacter species (Wiegel and Ljungdahl, 1981; Lee et al., 1993; Kim et al., 2001; Lee et al., 2007; Balk et al., 2009; Shaw et al., 2010). Strain PCO is able to reduce elemental sulfur and AQDS, but sulfate, sulfite, thiosulfate, nitrate, and nitrite could not serve as electron acceptors. The comparison between the morphological, biochemical and physiological characteristics of strain PCO and its close relatives is presented in Table 1. All of them can use thiosulfate as electron acceptor, but strain PCO cannot. Even though strain PCO and T. thermohydrosulfuricus are the same species, strain PCO can be differentiated because of its ability to grow and ferment cellulose and reduce AQDS. Figure 1 Phase-contrast micrographs showing morphology of cells of strain PCO. The arrows indicate vegetative and sporulating cells. Bar, 5 μm. Figure 2 Glucose conversion and products formation by strain PCO over time. The results represent the average of duplicate experiments. Table 1 Physiological and biochemical characteristics of (1) strain PCO and its phylogenetic related species: (2) Thermoanaerobacter thermohydrosulfuricus (DSM 527T), (3) T. brockii subsp. finnii (DSM 3389T), (4) T. pseudethanolicus (DSM 2355T), and (5) T. wiegelii (DSM 10319T). Characteristics 1* 2 3 4 5 Growth pH (optimum) 6.5–7.5 6.9–7.5 6.5–6.8 nr 6.8 Growth temperature (optimum) (°C) 70 67–69 65 65 65–68 Spore formation + + + + + Gram reaction Negative Variable Variable Variable Negative DNA G+C content (mol %) 34.5 37.6 32 34.4 35.6 SUBSTRATE UTILIZATION Arabinose + ± −* − − Carboxymethylcellulose ± + −* −* +* Cellobiose + + + + + Cellulose + − ±* −* +* Acetate − − −* −* −* Fructose + + + −* + Galactose + + + +* + Glucose + + + + + Lactose + +* + −* + Maltose + + + + + Mannose + + + ±* + Raffinose + +* +* +* + Ribose + + + + − Sucrose + + + + + Trehalose + + +* +* +* Xylose + + + + + Starch + + +* + + Pectin + + ±* −* + Peptone ± + −* ±* +* Xylan + + +* +* +* Yeast extract + + −* −* +* Pyruvate + + + + − Ethanol − −* −* −* − Glycerol − − ±* − + Mannitol + ± + − + Methanol − −* −* −* −* Sorbitol + +* ±* ±* +* CO − − −* −* −* H2/CO2 − −* −* − −* Formate − −* −* −* −* Glycine − ±* −* −* ±* Lactate − − −* −* − Propionate − −* −* −* −* Succinate − −* −* −* − ELECTRON ACCEPTORS AQDS + −* +* +* +* Elemental sulfur + +* +* +* +* Nitrate − + −* −* −* Nitrite − − −* −* −* Sulfate − − −* −* −* Sulfite − + −* −* +* Thiosulfate − + + + +* Data are from Cayol et al. (1995), Lee et al. (1993), Onyenwoke et al. (2007), Cook et al. (1996) and this study. nr, not reported. Symbols: +, utilized; ±, poorly utilized; −, not utilized; * , data from this study. Carbon monoxide tolerance of strain PCO and related Thermoanaerobacter sp. When strain PCO was cultured with 20 mM of glucose and subjected to different CO concentrations in the headspace (0, 25, 50, 75, and 100%, total pressure 1.7 × 105 Pa) no significant differences were found in glucose consumption rates (Table 2). Carbon monoxide concentrations of 75% or lower did not affect substantially product formation from glucose (Figures 3A, 4A), except for H2. Hydrogen production by strain PCO decreased significantly (P = 0.0001) from 3.94 ± 0.55 to 1.52 ± 0.13 mM when cultures were incubated with 0 and 25% CO in the headspace, respectively (Figure 3A). When testing different CO partial pressures, from 25 to 100%, a decrease in H2 production is observed, with H2 production of < 0.02 mmol L−1 at 100% CO. The other end products from glucose conversion were only affected when strain PCO was exposed to 100% CO. There was a significant decrease on the final production of lactate (P = 0.01) and ethanol (P = 0.004) only when comparing cultures grown with 75 and 100% CO (Figure 4A). These results could be explained due to the fact that this strain had been isolated from a syngas-converting culture, that was in contact with high CO concentrations for over 1 year (Alves et al., 2013), which might have increased the tolerance of strain PCO to CO. There are only two Thermoanaerobacter species able to use CO: T. thermohydrosulfuricus subsp. carboxydovorans and T. kivui (Balk et al., 2009; Weghoff and Müller, 2016); for other Thermoanaerobacter species the ability to convert CO was never reported. The ability to utilize CO as carbon and energy source and/or the ability to tolerate CO by T. thermohydrosulfuricus, T. brockii subsp. finnii, T. pseudethanolicus, and T. wiegelii were tested in this study. None of the tested Thermoanaerobacter species could utilize CO (Table 1), but all species could grow and completely convert glucose in the presence of 0, 50, or 100% of CO in the headspace (Table 3, Figures 3, 4). This suggests that CO tolerance is a characteristic present among Thermoanaerobacter genus, and not only a property of strain PCO. Carbon recovery at the end of the incubations was nearly 100% for all the growth tests (Table 3). All the tested strains produced the same identified and quantified end products from glucose fermentation, i.e., hydrogen, lactate, acetate, and ethanol, as expected from previous reports (Kim et al., 2001; Lee et al., 2007; Balk et al., 2009; Shaw et al., 2010). Nevertheless, final product concentrations (Figures 3, 4) varied for the different tested strains. Ethanol production was in general less affected by the presence of different CO concentrations. A significant decrease in ethanol production was only observed in the presence of 100% CO and just for two of the five microorganisms tested: strain PCO (P = 0.004) and T. wiegelii (P = 0.002). H2 production by all the tested strains decreased significantly, even in the presence of low CO concentrations. In the presence of 100% CO, H2 production by the Thermoanaerobacter species tested was decreased by 75–95%. Acetate concentration decreased with increasing CO percentage (reduction between 25 and 50%); the exception was strain PCO for which the final acetate concentration was significantly higher in the presence of high CO percentage [acetate final concentration in cultures with 0 and 100% CO were 5.4 ± 0.6 and 8.3 ± 0.5 mM, respectively (P = 0.0003)]. These results show the metabolic changes in versatile Thermoanaerobacter species upon addition of CO, resulting in a general decrease in H2 production. Although none of the tested Thermoanaerobacter strains could convert CO, all were able to withstand CO. The lower H2 production suggests that CO is inhibiting the hydrogenases of Thermoanaerobacter species. Hydrogenases catalyze the oxidation of hydrogen or the reduction of protons. From recent genomic studies, it was confirmed that [Fe-Fe]-hydrogenases, responsible for hydrogen production, are present and well-conserved in all of the Thermoanaerobacter species. However, another class of hydrogenases, [Ni-Fe]-hydrogenases, is present only in T. thermohydrosulfuricus and T. wiegelii (Verbeke et al., 2013; Bhattacharya et al., 2015). [Fe-Fe] or iron-only hydrogenases are known to be more sensitive to CO than [Ni-Fe]-hydrogenases (Diender et al., 2015), which corroborate the results obtained regarding to the effect of CO on hydrogen production from glucose conversion by Thermoanaerobacter species. Hydrogenases were shown to be specifically inhibited by carbon monoxide, since CO binds at the active site of the enzyme (Purec et al., 1962; Gutiérrez-Sánchez et al., 2010; Baffert et al., 2011; Matsumoto et al., 2011; Bertsch and Müller, 2015). Genomic analysis of T. kivui revealed the presence of genes encoding for carbon monoxide dehydrogenases (CODH) and Ech-hydrogenases complexes (which are responsible for hydrogen production by carboxydotrophic organisms; Hess et al., 2014), although its ability to convert CO was only reported very recently after adaption to increasing concentrations of CO (Weghoff and Müller, 2016). Proper adaption to CO may be the key for achieving CO conversion by Thermoanaerobacter species which contain the necessary genomic machinery. Therefore, the high tolerance to CO and the potential of some microorganisms for CO utilization, make the members of Thermoanaerobacter genus important for the biotechnological use of syngas/industrial CO-rich gases. Thermophilic microorganisms including members of Thermoanaerobacter genus are interesting catalysts for production of biofuels (Carere et al., 2012; Verbeke et al., 2013; Hess et al., 2014; Bhattacharya et al., 2015; Sant'Anna et al., 2015). From this perspective the present study is important as CO is a way to steer the formation of fermentation products. Further research is needed to get a better insight into how at a molecular level carbon monoxide affects product formation in Thermoanaerobacter species. Table 2 Glucose conversion rates by strain PCO under different CO partial pressures. CO (%) (Ptotal = 1.7 × 105 Pa) mM glucose consumed · day−1 0 2.91 ± 0.66 25 2.89 ± 0.54 50 2.90 ± 0.46 75 2.78 ± 0.51 100 2.35 ± 0.18 Figure 3 Glucose conversion, production of H2 from glucose and growth curves over time by (A) strain PCO, (B) Thermoanaerobacter thermohydrosulfuricus (DSM 527T), (C) T. brockii subsp. finnii (DSM 3389T), (D) T. pseudethanolicus (DSM 2355T), and (E) T. wiegelii (DSM 10319T), in incubations with different CO concentration in the gas phase (0, 25, 50, 75, or 100% CO). Plotted are the average data of duplicate experiments. The values of optical density were plotted vs. time on a logarithmic scale. Figure 4 Production of ethanol and organic acids (lactate and acetate) from glucose, over time, by (A) strain PCO, (B) Thermoanaerobacter thermohydrosulfuricus (DSM 527T), (C) T. brockii subsp. finnii (DSM 3389T), (D) T. pseudethanolicus (DSM 2355T), and (E) T. wiegelii (DSM 10319T), in incubations with different CO concentration in the gas phase (0, 25, 50, 75, or 100% CO). Plotted are the average data of duplicate experiments. Table 3 Effect of CO partial pressure on glucose conversion and carbon recovery (%) by strain PCO, Thermoanaerobacter thermohydrosulfuricus (DSM 527T), T. brockii subsp. finnii (DSM 3389T), T. pseudethanolicus (DSM 2355T), and T. wiegelii (DSM 10319T). Strains CO (%) Glucose consumed (mM) Product yielda CO2 producedb(mmol L−1medium) Carbon recoveryc(%) mol H2/mol glucose mol acetate/mol glucose mol lactate/mol glucose mol ethanol/mol glucose Strain PCO 0 17.9 ± 1.80 0.22 ± 0.04 0.29 ± 0.04 0.92 ± 0.11 0.39 ± 0.07 12.2 ± 1.2 88.1 ± 9.4 25 18.0 ± 0.80 0.08 ± 0.01 0.33 ± 0.03 0.93 ± 0.05 0.43 ± 0.02 13.7 ± 0.5 91.2 ± 4.4 50 18.3 ± 1.24 0.05 ± 0.01 0.35 ± 0.06 0.85 ± 0.14 0.33 ± 0.04 12.3 ± 1.1 82.2 ± 8.8 75 17.7 ± 1.54 0.06 ± 0.01 0.34 ± 0.06 0.93 ± 0.12 0.36 ± 0.07 12.3 ± 1.4 87.8 ± 9.3 100 18.1 ± 0.22 <0.02 0.46 ± 0.03 0.78 ± 0.05 0.21 ± 0.02 12.2 ± 0.7 77.1 ± 2.9 T. thermohydrosulfuricus 0 17.8 ± 1.00 0.61 ± 0.06 0.38 ± 0.06 0.64 ± 0.09 0.86 ± 0.14 22.2 ± 2.6 102.3 ± 9.0 50 15.7 ± 0.90 0.40 ± 0.08 0.28 ± 0.04 0.94 ± 0.14 0.58 ± 0.08 13.5 ± 1.3 100.9 ± 9.4 100 15.2 ± 0.30 0.18 ± 0.01 0.20 ± 0.01 1.21 ± 0.05 0.37 ± 0.06 8.7 ± 1.0 100.0 ± 3.8 T. brockii subsp. finnii 0 14.2 ± 1.80 0.37 ± 0.05 0.48 ± 0.07 0.85 ± 0.11 0.66 ± 0.19 16.1 ± 2.6 113.7 ± 15.9 50 14.9 ± 0.90 0.02 ± 0.002 0.32 ± 0.04 0.89 ± 0.06 0.70 ± 0.13 15.3 ± 1.9 112.4 ± 8.2 100 15.7 ± 0.40 <0.02 0.30 ± 0.05 0.84 ± 0.04 0.66 ± 0.18 15.1 ± 2.9 105.6 ± 7.7 T. pseudethanolicus 0 16.6 ± 0.50 0.22 ± 0.02 0.40 ± 0.02 0.72 ± 0.03 0.63 ± 0.22 17.0 ± 3.6 100.7 ± 8.7 50 17.1 ± 0.60 <0.02 0.26 ± 0.07 0.70 ± 0.06 0.68 ± 0.10 16.1 ± 2.0 91.4 ± 6.1 100 16.0 ± 1.60 <0.02 0.29 ± 0.04 0.85 ± 009 0.61 ± 0.12 14.6 ± 1.6 97.5 ± 10.7 T. wiegelii 0 17.7 ± 1.80 0.90 ± 0.17 0.39 ± 0.108 0.22 ± 0.04 1.45 ± 0.22 32.6 ± 3.2 124.9 ± 14.3 50 18.5 ± 1.20 0.71 ± 0.12 0.32 ± 0.03 0.26 ± 0.02 1.43 ± 0.13 32.3 ± 1.9 112.8 ± 8.0 100 17.1 ± 1.70 0.05 ± 0.02 0.25 ± 0.08 0.91 ± 0.21 0.66 ± 0.26 15.5 ± 4.5 99.0 ± 16.7 a Product yield = (CP, tf − CP, t0)/(CS, t0 − CS, tf); where, CP is product concentration and CS is glucose concentration measure at time zero (t0) and at the end of the assay (tf). Note: final product concentration was calculated as an average of the plateau of production curves. b Estimated CO2 production considering that 1 mol of CO2 is produced for each mol of ethanol or acetate formed [a maximum deviation of 20% was obtained when estimating total CO2 concentrations from measured CO2 concentration in the headspace summed with the correspondent calculated dissolved CO2 (using the Henry law)]. c Carbon recovery (CR) = Σ (nP, tf − nP, t0)/(nS, tf − nS, t0) + nbiomass; where n is the number of carbon moles in products (P − acetate, lactate, ethanol, and CO2) and in glucose (S) and nbiomass is the estimated mol of carbon used for biomass growth. For the calculation of nbiomass, cell density was analyzed photometrically by optical density at 600 nm (OD) and converted to cell dry weight per liter using experimentally determined conversion factors (OD/(gdry−weight/L)) for each of the tested strains: 0.78 (strain PCO), 0.89 (T. thermohydrosulfuricus), 1.36 (T. brockii subsp. finnii), 1.04 (T. pseudethanolicus), and 0.86 (T. wiegelii). Author contributions JA planned and performed the experiments, data interpretation and wrote the manuscript. MA assisted in the design of the study, participated in data interpretation as well as revisions of the final manuscript. CP assisted in the design of the study, participated in data interpretation as well as revisions of the final manuscript. AS assisted in the design of the study, participated in data interpretation as well as revisions of the final manuscript. DS conceived the study, participated in the planning and coordination of the study, and revised the manuscript. All authors read and gave approval for publication of the manuscript. Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. This study was supported by the Portuguese Foundation for Science and Technology (FCT) under the scope of the strategic funding of UID/BIO/04469/2013 unit and COMPETE 2020 (POCI-01-0145-FEDER-006684). The financial support from FCT and European Social Fund (POPH-QREN) through post-doc grant SFRH/BPD/104837/2014 given to JA is gratefully acknowledged. AS and DS are supported by an advanced ERC grant (project 323009) and a Gravitation grant (project 024.002.002) of the Netherlands Ministry of Education, Culture and Science and the Netherlands Science Foundation (NWO). ==== Refs References Altschul S. F. Gish W. Miller W. Myers E. W. Lipman D. J. (1990 ). Basic local alignment search tool . J. Mol. Biol . 215 , 403 –410 . 10.1016/S0022-2836(05)80360-2 2231712 Alves J. I. Stams A. J. Plugge C. M. Alves M. M. Sousa D. Z. (2013 ). Enrichment of anaerobic syngas-converting bacteria from thermophilic bioreactor sludge . FEMS Microbiol. Ecol . 86 , 590 –597 . 10.1111/1574-6941.12185 23899025 Baffert C. Bertini L. Lautier T. Greco C. Sybirna K. Ezanno P. . (2011 ). CO disrupts the reduced H-cluster of FeFe hydrogenase. A combined DFT and protein film voltammetry study . J. Am. Chem. Soc . 133 , 2096 –2099 . 10.1021/ja110627b 21271703 Balk M. Heilig H. G. van Eekert M. H. Stams A. J. M. Rijpstra I. C. Sinninghe-Damste J. S. . (2009 ). Isolation and characterization of a new CO-utilizing strain, Thermoanaerobacter thermohydrosulfuricus subsp. carboxydovorans, isolated from a geothermal spring in Turkey . Extremophiles 13 , 885 –894 . 10.1007/s00792-009-0276-9 19701714 Bertsch J. Müller V. (2015 ). Bioenergetic constraints for conversion of syngas to biofuels in acetogenic bacteria . Biotechnol. Biofuels 8 :210 . 10.1186/s13068-015-0393-x 26692897 Bhattacharya P. Barnebey A. Zemla M. Goodwin L. Auer M. Yannone S. M. (2015 ). Complete genome sequence of the chromate-reducing bacterium Thermoanaerobacter thermohydrosulfuricus strain BSB-33 . Stand. Genomic Sci . 10 :74 . 10.1186/s40793-015-0028-7 26445627 Carere C. R. Rydzak T. Verbeke T. J. Cicek N. Levin D. B. Sparling R. (2012 ). Linking genome content to biofuel production yields: a meta-analysis of major catabolic pathways among select H2 and ethanol-producing bacteria . BMC Microbiol . 12 :295 . 10.1186/1471-2180-12-295 23249097 Cayol J. L. Ollivier B. Patel B. K. Ravot G. Magot M. Ageron E. . (1995 ). Description of Thermoanaerobacter brockii subsp. lactiethylicus subsp. nov., isolated from a deep subsurface French oil well, a proposal to reclassify Thermoanaerobacter finnii as Thermoanaerobacter brockii subsp. finnii comb. nov., and an emended description of Thermoanaerobacter brockii . Int. J. Syst. Bacteriol. 45 , 783 –789 . 10.1099/00207713-45-4-783 7547300 Cook G. M. Rainey F. A. Patel B. K. Morgan H. W. (1996 ). Characterization of a new obligately anaerobic thermophile, Thermoanaerobacter wiegelii sp . nov. Int. J. Syst. Bacteriol . 46 , 123 –127 . 10.1099/00207713-46-1-123 8573487 Diender M. Stams A. J. M. Sousa D. Z. (2015 ). Pathways and bioenergetics of anaerobic carbon monoxide fermentation . Front. Microbiol . 6 :1275 . 10.3389/fmicb.2015.01275 26635746 Gutiérrez-Sánchez C. Rudiger O. Fernandez V. M. De Lacey A. L. Marques M. Pereira I. A. (2010 ). Interaction of the active site of the Ni-Fe-Se hydrogenase from Desulfovibrio vulgaris Hildenborough with carbon monoxide and oxygen inhibitors . J. Biol. Inorg. Chem . 15 , 1285 –1292 . 10.1007/s00775-010-0686-2 20669037 Hall T. A. (1999 ). BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT . Nucleic Acids Symp. Ser. 41 , 95 –98 . Hemme C. L. Mouttaki H. Lee Y. J. Zhang G. Goodwin L. Lucas S. . (2010 ). Sequencing of multiple clostridial genomes related to biomass conversion and biofuel production . J. Bacteriol . 192 , 6494 –6496 . 10.1128/JB.01064-10 20889752 Hess V. Poehlein A. Weghoff M. C. Daniel R. Müller V. (2014 ). A genome-guided analysis of energy conservation in the thermophilic, cytochrome-free acetogenic bacterium Thermoanaerobacter kivui . BMC Genomics 15 :1139 . 10.1186/1471-2164-15-1139 25523312 Jessen J. E. Orlygsson J. (2012 ). Production of ethanol from sugars and lignocellulosic biomass by Thermoanaerobacter J1 isolated from a hot spring in Iceland . J. Biomed. Biotechnol . 2012 , 186982 . 10.1155/2012/186982 23118498 Kevbrina M. V. Ryabokon A. M. Pusheva M. A. (1996 ). Acetate formation from CO-containing gas mixtures by free and immobilized cells of the Thermophilic homoacetogenic bacterium Thermoanaerobacter kivui . Microbiology 65 , 753 –757 . Kim B. C. Grote R. Lee D. W. Antranikian G. Pyun Y. R. (2001 ). Thermoanaerobacter yonseiensis sp. nov., a novel extremely thermophilic, xylose-utilizing bacterium that grows at up to 85 degrees C . Int. J. Syst. Evol. Microbiol. 51 , 1539 –1548 . 10.1099/00207713-51-4-1539 11491356 Lacis L. S. Lawford H. G. (1991 ). Thermoanaerobacter ethanolicus growth and product yield from elevated levels of xylose or glucose in continuous cultures. Appl. Environ . Microbiol . 57 , 579 –585 . 16348422 Lane D. J. (1991 ). Nucleic Acid Techniques in Bacterial Systematics . Chichester : John Wiley & Sons . Lee Y.-E. Jain M. K. Lee C. Lowe S. E. Zeikus J. G. (1993 ). Taxonomic distinction of saccharolytic thermophilic anaerobes: description of Thermoanaerobacterium xylanolyticum gen. nov., sp. nov., and Thermoanaerobacterium saccharolyticum gen. nov., sp. nov.; Reclassification of Thermoanaerobium brockii, Clostridium thermosulfurogenes, and Clostridium thermohydrosulfuricum E100-69 as Thermoanaerobacter brockii comb. nov., Thermoanaerobacterium thermosulfurigenes comb. nov., and Thermoanaerobacter thermohydrosulfuricus comb. nov., respectively; and transfer of Clostridium thermohydrosulfuricum 39E to Thermoanaerobacter ethanolicus . Int. J. Syst. Bacteriol. 43 , 41 –51 . 10.1099/00207713-43-1-41 Lee Y. J. Dashti M. Prange A. Rainey F. A. Rohde M. Whitman W. B. . (2007 ). Thermoanaerobacter sulfurigignens sp. nov., an anaerobic thermophilic bacterium that reduces 1 M thiosulfate to elemental sulfur and tolerates 90 mM sulfite . Int. J. Syst. Evol. Microbiol. 57 , 1429 –1434 . 10.1099/ijs.0.64748-0 17625170 Matsumoto T. Kabe R. Nonaka K. Ando T. Yoon K. S. Nakai H. . (2011 ). Model study of CO inhibition of [NiFe]hydrogenase . Inorg. Chem . 50 , 8902 –8906 . 10.1021/ic200965t 21853978 Muyzer G. de Waal E. C. Uitterlinden A. G. (1993 ). Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA . Appl. Environ. Microbiol. 59 , 695 –700 . 7683183 Nübel U. Engelen B. Felske A. Snaidr J. Wieshuber A. Amann R. I. . (1996 ). Sequence heterogeneities of genes encoding 16S rRNAs in Paenibacillus polymyxa detected by temperature gradient gel electrophoresis . J. Bacteriol. 178 , 5636 –5643 . 8824607 Onyenwoke R. U. Kevbrin V. V. Lysenko A. M. Wiegel J. (2007 ). Thermoanaerobacter pseudethanolicus sp. nov., a thermophilic heterotrophic anaerobe from Yellowstone National Park . Int. J. Syst. Evol. Microbiol. 57 , 2191 –2193 . 10.1099/ijs.0.65051-0 17911280 Purec L. Krasna A. I. Rittenberg D. (1962 ). The inhibition of hydrogenase by carbon monoxide and the reversal of this inhibition by light . Biochemistry 1 , 270 –275 . 10.1021/bi00908a013 14489243 Sanguinetti C. J. Dias Neto E. Simpson A. J. (1994 ). Rapid silver staining and recovery of PCR products separated on polyacrylamide gels . Biotechniques 17 , 914 –921 . 7840973 Sant'Anna F. H. Lebedinsky A. V. Sokolova T. G. Robb F. T. Gonzalez J. M. (2015 ). Analysis of three genomes within the thermophilic bacterial species Caldanaerobacter subterraneus with a focus on carbon monoxide dehydrogenase evolution and hydrolase diversity . BMC Genomics 16 :757 . 10.1186/s12864-015-1955-9 26446804 Shaw A. J. Hogsett D. A. Lynd L. R. (2010 ). Natural competence in Thermoanaerobacter and Thermoanaerobacterium species . Appl. Environ. Microbiol . 76 , 4713 –4719 . 10.1128/AEM.00402-10 20472726 Stams A. J. M. van Dijk J. B. Dijkema C. Plugge C. M. (1993 ). Growth of syntrophic propionate-oxidizing bacteria with fumarate in the absence of methanogenic bacteria. Appl. Environ . Microbiol . 59 , 1114 –1119 . Turner P. Mamo G. Karlsson E. N. (2007 ). Potential and utilization of thermophiles and thermostable enzymes in biorefining . Microb. Cell Fact . 6 :9 . 10.1186/1475-2859-6-9 17359551 Verbeke T. J. Zhang X. Henrissat B. Spicer V. Rydzak T. Krokhin O. V. . (2013 ). Genomic evaluation of Thermoanaerobacter spp. for the construction of designer co-cultures to improve lignocellulosic biofuel production . PLoS ONE 8 :e59362 . 10.1371/journal.pone.0059362 23555660 Wagner I. D. Wiegel J. (2008 ). Diversity of thermophilic anaerobes . Ann. N. Y. Acad. Sci . 1125 , 1 –43 . 10.1196/annals.1419.029 18378585 Weghoff M. C. Müller V. (2016 ). CO metabolism in the thermophilic acetogen Thermoanaerobacter kivui . Appl. Environ. Microbiol . 82 , 2312 –2319 . 10.1128/AEM.00122-16 26850300 Wiegel J. Ljungdahl L. G. (1981 ). Thermoanaerobacter ethanolicus gen. nov., spec. nov., a new extreme thermophilic anaerobic bacterium . Arch. Microbiol. 128 , 343 –348 . 10.1007/BF00405910 Yoneda Y. Kano S. I. Yoshida T. Ikeda E. Fukuyama Y. Omae K. . (2015 ). Detection of anaerobic carbon monoxide-oxidizing thermophiles in hydrothermal environments . FEMS Microbiol. Ecol. 91 :fiv093 . 10.1093/femsec/fiv093 26223231

PMC005xxxxxx/PMC5002421.txt

==== Front Front Plant SciFront Plant SciFront. Plant Sci.Frontiers in Plant Science1664-462XFrontiers Media S.A. 10.3389/fpls.2016.01300Plant ScienceOriginal ResearchTranscriptional Regulation of Tetrapyrrole Biosynthetic Genes Explains Abscisic Acid-Induced Heme Accumulation in the Unicellular Red Alga Cyanidioschyzon merolae Kobayashi Yuki 1Tanaka Kan 12*1Laboratory for Chemistry and Life Science, Institute of Innovative Research, Tokyo Institute of Technology, YokohamaJapan2Core Research for Evolutional Science and Technology, Japan Science and Technology Agency, SaitamaJapanEdited by: Henri Batoko, Université catholique de Louvain, Belgium Reviewed by: Shan Lu, Nanjing University, China; Martha Gledhill, GEOMAR Helmholtz-Zentrum für Ozeanforschung, Germany *Correspondence: Kan Tanaka, kntanaka@res.titech.ac.jpThis article was submitted to Plant Physiology, a section of the journal Frontiers in Plant Science 29 8 2016 2016 7 130028 5 2016 15 8 2016 Copyright © 2016 Kobayashi and Tanaka.2016Kobayashi and TanakaThis is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.Abscisic acid (ABA), a pivotal phytohormone that is synthesized in response to abiotic stresses and other environmental changes, induces various physiological responses. Heme, in its unbound form, has a positive signaling role in cell-cycle initiation in Cyanidioschyzon merolae. ABA induces heme accumulation, but also prevents cell-cycle initiation through the titration of the unbound heme by inducing the heme scavenging protein tryptophan-rich sensory protein-related protein O. In this study, we analyzed the accumulation of tetrapyrrole biosynthetic gene transcripts after the addition of ABA to the medium and found that transcripts of a ferrochelatase and a magnesium-chelatase subunit increased, while other examined transcripts decreased. Under the same conditions, the heme and magnesium-protoporphyrin IX contents increased, while the protoporphyrin IX content decreased. Thus, ABA may regulate the intracellular heme and other tetrapyrrole contents through the transcriptional regulation of biosynthetic genes. abscisic acidCyanidioschyzon merolaehemetetrapyrroletranscriptional regulationTSPOJapan Society for the Promotion of Science10.13039/5011000016912137001523120505242480615K145391327435015621958 ==== Body Introduction Abscisic acid (ABA) is a phytohormone of land plants that is involved in many aspects of plant physiology. ABA induces stress tolerance under various stressful conditions, such as drought and high salt (Skriver and Mundy, 1990; Tuteja, 2007). ABA induces the growth of roots, as well as dormancy in buds and seeds, and in leaf organs it induces the stomatal closure, which helps the plant to preserve water during droughts. It is widely believed that these responses are mostly mediated by transcriptional activation, and an underlying mechanism involving the specific receptors pyrabactin resistance 1 (PYR1)/PYR1-like/regulatory components of ABA receptors, SNF1-related kinases 2 and Protein phosphatase 2C has been clarified (Ma et al., 2009; Park et al., 2009; Hartung, 2010; Hauser et al., 2011). In addition, ABA also positively regulates heme biosynthesis in Arabidopsis (Vanhee and Batoko, 2011; Vanhee et al., 2011). Exogenous ABA addition induces the transient increase of intracellular unbound heme, which is required to activate ABA-8′-hydroxylase, an ABA degradation enzyme to prevent continuous signaling events (Vanhee and Batoko, 2011; Vanhee et al., 2011). At the same time, ABA induces a heme-scavenger protein, tryptophan-rich sensory protein-related protein O (TSPO), to quench excess unbound heme that may cause oxidative damage to the cell (Vanhee and Batoko, 2011; Vanhee et al., 2011). Abscisic acid has also been found in eukaryotic algae and cyanobacteria. Thus, ABA signaling could be very ancient, but the physiological significance has been poorly understood. In a previous study, we analyzed ABA function in the unicellular red alga Cyanidioschyzon merolae, which has available complete nuclear, mitochondrial, and chloroplast genome sequences and molecular genetic tools (Kuroiwa, 1998; Matsuzaki et al., 2004; Nozaki et al., 2007; Kobayashi et al., 2010). In Arabidopsis and other land plants, ABA can be synthesized from the common precursor zeaxanthin by three enzymes, zeaxanthin epoxidase (ZEP), 9-cis-epoxycarotenoid dioxygenase (NCED), and short-chain dehydrogenase/reductase (SDR). The C. merolae genome was found to encode orthologous proteins for NCED and SDR, and the presence of zeaxanthin was previously confirmed in C. merolae (Cunningham et al., 2007). Therefore, we checked for endogenous ABA from C. merolae culture, and found that C. merolae accumulates ABA under the salt stressed condition (Kobayashi et al., 2016). As the NCED knock out strain could not accumulate ABA, the common ABA biosynthetic pathway as in land plants was confirmed in C. merolae. While no orthologous gene encoding ZEP was specified from the C. merolae genome, it was reported that an Arabidopsis plant lacking the functional ZEP still accumulates ABA (Barrero et al., 2005). Thus, there is possibly another type enzyme that directs the reaction, which could be common with the C. merolae enzyme. The addition of exogenous ABA induced a block in the cell-cycle G1/S transition and the homologous TSPO gene’s expression. Because TSPO scavenges the intracellular unbound heme in Arabidopsis, we wondered whether this unbound heme was required for the cell-cycle initiation, and found that the inhibitory effect by ABA was canceled by the addition of exogenous heme. Thus, ABA prevents the cell-cycle G1/S transition through the reduction of intracellular unbound heme accumulation in C. merolae (Kobayashi et al., 2016). As the interconnection among ABA, TSPO, and heme was likely conserved between primitive algae and land plants, the regulatory scheme is likely an ancient trait of ABA signaling conserved during plant evolution. However, the underlying mechanism of ABA-induced heme accumulation has not been clearly elucidated. In this study, we examined the expression of tetrapyrrole biosynthetic genes and the cellular contents of tetrapyrrole intermediates, and hypothesize that ABA affects the tetrapyrrole contents through transcriptional control. Materials and Methods Materials and Culture Conditions Cells of C. merolae 10D were cultured and their growth synchronized as described previously (Kobayashi et al., 2010). Quantitative PCR Cells were cultured under constant light or synchronizing conditions, with or without ABA (10 μM). Total RNA was extracted from C. merolae cells as described previously (Kobayashi et al., 2011). First-strand synthesis of cDNA was performed using 5 μg RNA and random primers with ReverTra Ace (Toyobo, Osaka, Japan), and the abundance of the each transcript was quantified using real-time PCR. Real-time PCR was performed as described previously (Kobayashi et al., 2016), using the primers shown in Supplementary Table S1. Measurement of Tetrapyrrole Molecules Protoporphyrin IX (ProtoIX), magnesium-protoporphyrin IX (Mg-ProtoIX), and chlorophyll-a were measured by high-performance liquid chromatography (HPLC) as described previously (Zapata et al., 2000), with minor modifications. Synchronized cells were homogenized in 80% acetone and centrifuged at 10,000 ×g for 5 min. The supernatant was mixed with water to a final concentration of 75% before the HPLC analysis. According to the method of Zapata et al. (2000), pigments were separated on a reversed-phase column, Symmetry C8 (150 mm × 4.6 mm; Waters, Milford, MA, USA) using a Nexera X2 HPLC system (Shimadzu, Kyoto, Japan). Mg-ProtoIX was detected with an excitation wavelength at 417 nm and emission at 600 nm. ProtoIX was detected with an excitation wavelength at 400 nm and emission at 635 nm. Chlorophyll-a was detected by measuring the absorbance at 410 nm. Standard curves were made using authentic standards. Northern Blot Analysis Total RNA preparation and northern blot analyses were performed as described previously (Kobayashi et al., 2011). Gene-specific probes for northern blot analyses were generated with specific primers (Supplementary Table S1) and C. merolae genomic DNA as the template. Results and Discussion Genes Involved in Heme Homeostasis in C. merolae Because most known ABA responses are mediated by the transcriptional regulation of nuclear genes, we examined whether this was also the case for the ABA induced-heme increase. In C. merolae, the tetrapyrrole biosynthetic pathway was previously analyzed and localization of each enzyme to specific cell compartments was proposed (Watanabe et al., 2013) (Figure 1). Here, we focused on 13 enzymes, glutamyl-tRNA reductase (HemA; CMJ054C), glutamate-1-semialdehyde 2, 1- aminomutase (HemL; CMP285C), aminolevulinic acid dehydrase (HemB; CMD104C), porphobilinogen deaminase (CME132C), uroporphyrinogen III synthase (HemD; CML040C), uroporphyrinogen decarboxylase (HemE; CME194C, CMP083C), coproporphyrinogen III oxidase (HemF; CMO136C), oxygen-independent coproporphyrinogen III oxidase (HemN; CMR445C), protoporphyrinogen IX oxidase (HemY; CMB025C), ferrochelatase (FeCh; CMS035C), heme oxygenase (HO; CMH209C), and, three Mg-chelatases, ChlH (CMB093C), ChlD (CMM270C), and ChlI (CMV024C), that were considered to be involved in the heme homeostasis. These enzymes are encoded by the nuclear genome, except for one of the three Mg-chelatase subunits, ChlI. FIGURE 1 ABA affected the accumulation of heme-related gene transcripts. The tetrapyrrole biosynthetic pathway is schematically represented. C. merolae cells were dark adapted for 16 h, and the transcript accumulation for each gene was monitored after illumination by quantitative PCR in the absence or presence of ABA. Cells were sampled at the indicated times, and ABA (10 μM) was added 30 min before the light illumination. Data represent the average of three independent experiments (n = 3 SD). Boxes indicate tetrapyrrole intermediates, and the localization of each enzyme is indicated as in the mitochondrion (MT) or chloroplast (CP) as predicted (Watanabe et al., 2013). The light-dependent increase in the transcript levels were repressed (blue) or activated (red) by ABA addition. Asterisks indicate genes whose transcripts were not detected by the quantitative PCR analysis. Accumulation of the Heme-Related Gene Transcripts Was Affected by ABA We first incubated C. merolae cells under dark conditions, and the accumulations of heme-related gene transcripts were examined after illumination by quantitative PCR in the absence or the presence of ABA. We could not detect transcripts for two genes, CME132C and CMP083C, probably because of the limited transcript abundance. For the other genes examined, the detected transcripts increased in response to light while the relative increase and the time course profiles differed for each gene (Figure 1). However, the effect of ABA was divided into two categories: ABA repressively affected the light-dependent increase of most gene transcripts, while conversely activating the accumulation of FeCh and ChlH gene transcripts. As FeCh is the enzyme directly responsible for heme biosynthesis, the activation by ABA may result in the increased heme accumulation. HO is an enzyme involved in heme degradation (Shan et al., 2004; Shekhawat and Verma, 2010), and thus the decrease of HO transcripts in the presence of ABA is also consistent with heme accumulation. However, the relevance of the ChlH transcript increase, which encodes for a subunit of Mg-chelatase for the chlorophyll biosynthesis branch, is not clear. As we initially only examined the ChlH transcripts, we further checked for the expression of other Mg-chelatase subunit genes, ChlD and ChlI, by northern blot analysis, and confirmed that their transcripts’ accumulation was also activated by ABA (Supplementary Figure S1). ProtoIX and Mg-ProtoIX Pools Were also Affected by ABA If the transcripts levels explain the change in the heme content, then not only would the increase in heme be expected in the presence of ABA, but also the decrease in ProtoIX and the increase in Mg-ProtoIX. To test this hypothesis, we determined the cellular contents of heme, ProtoIX and Mg-ProtoIX, as well as chlorophyll-a as a control, under the same conditions as in Figure 1. As shown in Figure 2, the ABA-induced decrease in ProtoIX and increases in heme and Mg-ProtoIX were observed, which was consistent with the gene expression changes. The depletion of ProtoIX, the direct precursor of these compounds, explains the transient nature of the heme and Mg-ProtoIX increases. The chlorophyll-a content was not affected by ABA probably because the cellular content was rather abundant as compared with other tetrapyrrole intermediates and not easily affected by the change in the biosynthesis rate. FIGURE 2 Cellular Contents of Mg-ProtoIX, Heme, ProtoIX, and Chlorophyll-a. Changes in accumulation of (A) Mg-ProtoIX, (C) ProtoIX, and, (D) Chlorophyll-a in the absence (gray) or presence (orange) of ABA are shown. The sampled conditions were as in Figure 1. (B) Changes in heme (heme a and heme b) accumulation (from Kobayashi et al., 2016) are also shown compared with previously obtained results. Each data point represents the average of three independent experiments (n = 3 SD). The x-axis indicates the amounts in 1 mL of culture medium (OD750 = 1). TSPO May Scavenge the ABA-Induced Unbound Mg-ProtoIX In a previous study, we showed that a transient increase in ProtoIX or Mg-ProtoIX induces nuclear DNA replication (Kobayashi et al., 2009). This could have indicated that ABA induces nuclear DNA replication, but this was not the case, and ABA inhibited cell-cycle initiation (Kobayashi et al., 2016). The total heme content was increased by ABA, but the unbound heme that is required for cell-cycle initiation was decreased through the induction of the heme scavenger protein TSPO (Kobayashi et al., 2016). TSPO has an affinity not only to heme but also for dicarboxylic porphyrins, including Mg-ProtoIX (Papadopoulos et al., 2006; Vanhee and Batoko, 2011). Thus, even when the total Mg-ProtoIX content increased, its unbound form, required for the signaling events, was likely quenched by TSPO and not available for cell-cycle activation. Physiological Significance of the Heme Accumulation It was previously shown that salt stress induced ABA accumulation in C. merolae (Kobayashi et al., 2016). While ABA is the phytohormone that transmits the stress signal to other cells in land plants, ABA in C. merolae is likely an intracellular signaling molecule since ABA rapidly loses its activity in the sulfur acidic culture medium (Kobayashi et al., 2016). As the signaling molecule, it is important to prevent the continuous ABA signaling, and thus the degradation enzyme ABA-8′-hydroxylase should be activated subsequently. In addition, salt stress induces generation of oxygen radicals, which needs to be scavenged by enzymes such as catalase and peroxidase. These enzymes require heme as the prosthetic group, and therefore it is reasonable that salt stress-induced ABA activates the heme accumulation. The excess unbound heme itself may cause the radical formation, which requires quenching by the ABA-induced TSPO. Evolutionary Implications The interrelationship among ABA, heme, and TSPO was first shown in Arabidopsis (Vanhee and Batoko, 2011; Vanhee et al., 2011), and a similar scheme was subsequently revealed in C. merolae (Kobayashi et al., 2016). Thus, this regulatory scheme was likely conserved during plant evolution. Presently, two issues remain to be addressed from mechanical and evolutionary points of view. First, although heme activates the cell-cycle G1/S progression in C. merolae, the presence of a similar mechanism has not been found in land plants. Future studies on the underlying molecular mechanism in C. merolae would clarify the significance for plant cell-cycle control. Second, an increase in the heme content was observed in C. merolae, as well as in Arabidopsis, but almost no information on the underlying mechanism in Arabidopsis has been provided. ABA-induced changes in the Arabidopsis transcriptome have been reported previously (Matsui et al., 2008), but it is difficult to determine the relevance to the heme content because the experimental time scale was rather different. Whether the expression of tetrapyrrole biosynthetic genes is regulated as in C. merolae needs to be clarified. The elucidation and comparison of these whole schemes in C. merolae and Arabidopsis would provide insights on the evolution of plant cell physiology. Author Contributions YK performed the experiments and contributed to the writing of the manuscript. KT designed and had overall responsibility for the study, and wrote the manuscript. Conflict of Interest Statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Funding. This study was supported by the Ministry of Education, Culture, Sports, Science and Technology/Japan Society for the Promotion of Science KAKENHI (Grant Nos. 21370015, 23120505, 2424806, and 15K14539 to KT, and 13274350 and 15621958 to YK). Supplementary Material The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2016.01300 Click here for additional data file. Click here for additional data file. ==== Refs References Barrero J. M. Piqueras P. González-Guzmán M. Serrano R. Rodríguez P. L. Ponce M. R. (2005 ). A mutational analysis of the ABA1 gene of Arabidopsis thaliana highlights the involvement of ABA in vegetative development. J. Exp. Bot. 56 2071 –2083 . 10.1093/jxb/eri206 15983017 Cunningham F. X. Jr.Lee H. Gantt E. (2007 ). Carotenoid biosynthesis in the primitive red alga Cyanidioschyzon merolae. Eukaryot. Cell 6 533 –545 . 10.1128/EC.00265-06 17085635 Hartung W. (2010 ). The evolution of abscisic acid (ABA) and ABA function in lower plants, fungi and lichen. Funct. Plant Biol. 37 806 –812 . 10.1071/FP10058 Hauser F. Waadt R. Schroeder J. I. (2011 ). Evolution of abscisic acid synthesis and signaling mechanisms. Curr. Biol. 21 R346 –R355 . 10.1016/j.cub.2011.03.015 21549957 Kobayashi Y. Ando H. Hanaoka M. Tanaka K. (2016 ). Abscisic acid participates in the control of cell-cycle initiation through heme homeostasis in the unicellular red alga Cyanidioschyzon merolae. Plant Cell Physiol. 57 953 –960 . 10.1093/pcp/pcw054 27044672 Kobayashi Y. Imamura S. Hanaoka S. Tanaka K. (2011 ). A tetrapyrrole-regulated ubiquitin ligase controls algal nuclear DNA replication. Nat. Cell Biol. 13 483 –487 . 10.1038/ncb2203 21378982 Kobayashi Y. Kanesaki Y. Tanaka A. Kuroiwa H. Kuroiwa T. Tanaka K. (2009 ). Tetrapyrrole signal as a cell-cycle coordinator from organelle to nuclear DNA replication in plant cells. Proc. Natl. Acad. Sci. U.S.A. 106 803 –807 . 10.1073/pnas.0804270105 19141634 Kobayashi Y. Ohnuma M. Kuroiwa T. Tanaka K. Hanaoka M. (2010 ). The basics of cultivation and molecular genetic analysis of the unicellular red alga Cyanidioschyzon merolae. Endocytobiosis Cell Res. 20 53 –61 . Kuroiwa T. (1998 ). The primitive red algae: Cyanidium caldarium and Cyanidioschyzon merolae as model system for investigating the dividing apparatus of mitochondria and plastids. Bioessays 20 344 –354 . 10.1002/(SICI)1521-1878(199804)20:4<344::AID-BIES11>3.0.CO;2-2 Ma Y. Szostkiewicz I. Korte A. Moes D. Yang Y. Christmann A. (2009 ). Regulators of PP2C phosphatase activity function as abscisic acid sensors. Science 324 1064 –1068 . 10.1126/science.1172408 19407143 Matsui A. Ishida J. Morosawa T. Mochizuki Y. Kaminuma E. Endo T. A. (2008 ). Arabidopsis transcriptome analysis under drought, cold, high-salinity and ABA treatment conditions using a tiling array. Plant Cell Physiol. 8 1135 –1149 . 10.1093/pcp/pcn101 18625610 Matsuzaki M. Misumi O. Shin-I T. Maruyama S. Takahara M. Miyagishima S. (2004 ). Genome sequence of the ultrasmall unicellular red alga Cyanidioschyzon merolae 10D. Nature 428 653 –657 . 10.1038/nature02398 15071595 Nozaki H. Takano H. Misumi O. Terasawa K. Matsuzaki M. Maruyama S. (2007 ). A 100%-complete sequence reveals unusually simple genomic features in the hot spring red alga Cyanidioschyzon merolae. BMC Biol. 5 :28 10.1186/1741-7007-5-28 Papadopoulos V. Baraldi M. Guilarte T. R. Knudsen T. B. Lacapère J. J. Lindemann P. (2006 ). Translocator protein (18kDa): new nomenclature for the peripheral-type benzodiazepine receptor based on its structure and molecular function. Trends Pharmacol. Sci. 27 402 –409 .16822554 Park S. Y. Fung P. Nishimura N. Jensen D. R. Fujii H. Zhao Y. (2009 ). Abscisic acid inhibits type 2C protein phosphatases via the PYR/PYL family of start proteins. Science 324 1068 –1071 . 10.1126/science.1173041 19407142 Shan Y. Lambrecht R. W. Ghaziani T. Donohue S. E. Bonkovsky H. L. (2004 ). Role of Bach-1 in regulation of heme oxygenase-1 in human liver cells: insights from studies with small interfering RNAS. J. Biol. Chem. 279 51769 –51774 . 10.1074/jbc.M409463200 15465821 Shekhawat G. S. Verma K. (2010 ). Haem oxygenase (HO): an overlooked enzyme of plant metabolism and defence. J. Exp. Bot. 61 2255 –2270 . 10.1093/jxb/erq074 20378668 Skriver K. Mundy J. (1990 ). Gene expression in response to abiscisic acid and osmotic stress. Plant Cell 2 503 –512 . 10.2307/3869112 2152172 Tuteja N. (2007 ). Mechanisms of high salinity tolerance in plants. Methods Enzymol. 428 419 –438 . 10.1016/S0076-6879(07)28024-3 17875432 Vanhee C. Batoko H. (2011 ). Arabidopsis TSPO and porphyrins metabolism. Plant Signal. Behav. 6 1383 –1385 . 10.4161/psb.6.9.16477 21852759 Vanhee C. Zapotoczny G. Masquelier D. Ghislain M. Batoko H. (2011 ). The Arabidopsis multistress regulator TSPO is a heme binding membrane protein and a potential scavenger of porphyrins via an autophagy-dependent degradation mechanism. Plant Cell 23 785 –805 . 10.1105/tpc.110.081570 21317376 Watanabe S. Hanaoka M. Ohba Y. Ono Y. Ohnuma M. Yoshikawa H. (2013 ). Mitochondrial localization of ferrochelatase in a red alga Cyanidioschyzon merolae. Plant Cell Physiol. 54 1289 –1295 . 10.1093/pcp/pct077 23700350 Zapata M. Rodriguez F. Grarrido J. L. (2000 ). Separation of chlorophylls and carotenoids from marine phytoplankton: a new HPLC method using a reversed phase C8 column and pyridine-containing mobile phases. Mar. Ecol. Prog. Ser. 195 29 –45 .